Configuring Hardware and Performing...

Configuring Hardware and Performing Measurement

1

IC-CAP 2012.01January 2012



2

© Agilent Technologies, Inc. 2000-20113501 Stevens Creek Blvd., Santa Clara, CA 95052 USANo part of this documentation may be reproduced in any form or by any means (includingelectronic storage and retrieval or translation into a foreign language) without prioragreement and written consent from Agilent Technologies, Inc. as governed by UnitedStates and international copyright laws.

Acknowledgments

UNIX ® is a registered trademark of the Open Group.MS-DOS ®, Windows ®, and MS Windows ® are U.S. registered trademarks of MicrosoftCorporation.Pentium ® is a U.S. registered trademark of Intel Corporation.PostScript® is a trademark of Adobe Systems Incorporated.Java™ is a U.S. trademark of Sun Microsystems, Inc.Mentor Graphics is a trademark of Mentor Graphics Corporation in the U.S. and othercountries.Qt Version 4.6Qt NoticeThe Qt code was modified. Used by permission.Qt CopyrightQt Version 4.6, Copyright (c) 2009 by Nokia Corporation. All Rights Reserved.Qt License Your use or distribution of Qt or any modified version of Qt implies that youagree to this License. This library is free software; you can redistribute it and/or modify itunder the terms of the GNU Lesser General Public License as published by the FreeSoftware Foundation; either version 2.1 of the License, or (at your option) any laterversion. This library is distributed in the hope that it will be useful, but WITHOUT ANYWARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR APARTICULAR PURPOSE. See the GNU Lesser General Public License for more details. Youshould have received a copy of the GNU Lesser General Public License along with thislibrary; if not, write to the Free Software Foundation, Inc., 51 Franklin St, Fifth Floor,Boston, MA 02110-1301 USA Permission is hereby granted to use or copy this programunder the terms of the GNU LGPL, provided that the Copyright, this License, and theAvailability of the original version is retained on all copies. User documentation of anycode that uses this code or any modified version of this code must cite the Copyright, thisLicense, the Availability note, and "Used by permission." Permission to modify the codeand to distribute modified code is granted, provided the Copyright, this License, and theAvailability note are retained, and a notice that the code was modified is included.Qt Availability http://www.qtsoftware.com/downloadsPatches Applied to Qt can be found in the installation at:$HPEESOF_DIR/prod/licenses/thirdparty/qt/patches.You may also contact Brian Buchanan at Agilent Inc. at [email protected] formore information. For details see:http://bmaster.soco.agilent.com/mw/Qt_License_Information

Errata The IC-CAP product may contain references to "HP" or "HPEESOF" such as in filenames and directory names. The business entity formerly known as "HP EEsof" is now partof Agilent Technologies and is known as "Agilent EEsof." To avoid broken functionality andto maintain backward compatibility for our customers, we did not change all the namesand labels that contain "HP" or "HPEESOF" references.

Warranty The material contained in this documentation is provided "as is", and is subjectto being changed, without notice, in future editions. Further, to the maximum extentpermitted by applicable law, Agilent disclaims all warranties, either express or implied,with regard to this manual and any information contained herein, including but not limitedto the implied warranties of merchantability and fitness for a particular purpose. Agilentshall not be liable for errors or for incidental or consequential damages in connection withthe furnishing, use, or performance of this document or of any information containedherein. Should Agilent and the user have a separate written agreement with warrantyterms covering the material in this document that conflict with these terms, the warrantyterms in the separate agreement shall control.

Technology Licenses The hardware and/or software described in this document arefurnished under a license and may be used or copied only in accordance with the terms ofsuch license.

Restricted Rights Legend U.S. Government Restricted Rights. Software and technicaldata rights granted to the federal government include only those rights customarilyprovided to end user customers. Agilent provides this customary commercial license inSoftware and technical data pursuant to FAR 12.211 (Technical Data) and 12.212(Computer Software) and, for the Department of Defense, DFARS 252.227-7015(Technical Data - Commercial Items) and DFARS 227.7202-3 (Rights in CommercialComputer Software or Computer Software Documentation).

http://www.qtsoftware.com/downloads

http://www.qtsoftware.com/downloads

http://bmaster.soco.agilent.com/mw/Qt_License_Information

http://bmaster.soco.agilent.com/mw/Qt_License_Information


3

Supported Instruments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 DC Analyzers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Capacitance-Voltage meters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Network Analyzers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Oscilloscopes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Pulse Generators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 Dynamic Signal Analyzers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

Drivers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 Adding an Instrument Driver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 Prober Drivers in IC-CAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Matrix Drivers in IC-CAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 Driver Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Handling Signals and Exceptions in Prober and Matrix Drivers . . . . . . . . . . . . . . . . . . . . . . . 80

Performing a Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 A Sample Measurement Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 Sweep Modes and Input/Output Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 Repetitive Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 Fast Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 GPIB Analyzer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 icedil Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 Configured Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

Using IC-CAP with an Agilent 85122A Precision Modeling System . . . . . . . . . . . . . . . . . . . . . 99 Using IC-CAP with an Agilent 85123A Device Modeling System . . . . . . . . . . . . . . . . . . . . . . . 108


4

Supported InstrumentsThis section discusses the instruments supported by IC-CAP and describes the options foreach instrument. The instruments are divided into following basic groups:

DC analyzers (measurement)Capacitance Voltage meters (measurement)Network Analyzers (measurement)Oscilloscopes (measurement)Pulse Generators (measurement)Dynamic Signal Analyzers (measurement)


5

DC AnalyzersDC analyzers source and monitor voltages and currents and return data representing DCcharacteristics. IC-CAP supports the following DC analyzers:

HP 4071A Semiconductor Parametric Tester (measurement)HP 4140 pA Meter and DC Voltage Source (measurement)HP 4141 DC Source and Monitor (measurement)HP and Agilent 4142 Modular DC Source and Monitor (measurement)HP 4145 Semiconductor Parameter Analyzer (measurement)HP and Agilent 4155 Semiconductor Parameter Analyzer (measurement)HP and Agilent 4156 Precision Semiconductor Parameter Analyzer (measurement)Agilent E5260 Series Parametric Measurement Solutions (measurement)Agilent E5270 Series Parametric Measurement Solutions (measurement)Agilent B1500A Semiconductor Device Analyzer (measurement)Agilent B1505A Power Device Analyzer and Curve Tracer (measurement)Agilent B2900 Precision Source Measure Unit (measurement)

CautionIC-CAP does not restrict bias magnitude. When using a DC analyzer as a bias source for other instrumentssuch as capacitance-voltage meters or network analyzers, check the limit on external bias voltage orcurrent for each instrument. Excessive voltage or current may damage other instruments.

HP 4071A Semiconductor Parametric Tester

The HP 4071A IC-CAP driver enables you to control the HP 4071A SemiconductorParametric Tester from within IC-CAP.

NoteIC-CAP requires the Agilent 4070 System Software (also referred to as TIS), version B.02.00, or higher, todrive the Agilent 4071 Semiconductor Parametric Tester. The Agilent 4071 Semiconductor ParametricTester is only supported on the HP-UX 11i platform. For assistance using the Agilent 4070 SystemSoftware (TIS), please contact your local Agilent Instrument Support Team.

GPIB Interface

The HP 4071A does not have a GPIB interface available by which you can controlmeasurements. However, in keeping within the IC-CAP framework, an interface is requiredby the hardware manager in IC-CAP. The interface choices for the HP 4071 are limited totis_offline, and tis_online. tis_offline runs the HP 4071 driver in a mode that does notrequire that the HP 4071 system be connected. tis_online runs the HP4071 driver in amode that communicates with the HP 4071 system when one is available. You can add aninterface in the Hardware Setup window using Tools > Hardware Setup in the IC-CAP/Main window, then click on Rebuild to set up the tester.

IC-CAP will invoke the hp4070 executable if it is not already running or is shutdown duringan IC-CAP function. Therefore, in the window where you start IC-CAP, you must set thePATH environment variable to the directory where the hp4070 executable is located. Thetypical installation directory for the hp4070 executable is /opt/hp4070/bin.

Pin Connections

The HP 4071A switch matrix is controlled by the values entered for each of the Pinsoptions in the Instrument Options Table. You can view the instrument options in the Modelwindow after setting up the HP 4071 hardware, and creating an input for a setup.Highlight the setup name, then click on the Instrument Options tab. The values for thePins option describes which PORT is connected to the available test head pins. Generally,each SMU has the same options implemented in the driver. One exception is that theGuard Pins option available for SMU1 and SMU2 are not available for SMU3. See theavailable instrument options in HP 4071A Options.

The following table shows examples of valid entries for Pins and the resulting connections:

Valid Pins Field Entry Resulting Pin Connections

10 10

1,5,7,9 1, 5, 7, 9

2,4-7,9 2, 4, 5, 6, 7, 9

2,4-7,9,0 Not connected

35,5,2-4 35, 5, 2, 3, 4

12-16 12, 13, 14, 15,16

Notice that valid entries include a series of numbers separated by commas, and a range ofnumbers using a dash. A 0 appearing anywhere in a Pins field disconnects the PORT fromthe switch matrix. This is an easy way to disconnect the PORT without having to erase thepin numbers. The Pins field also can be left blank.

If the Pins field is left blank, then IC-CAP will search for a pre-defined IC-CAP variable.The string value of the pre-defined IC-CAP variable becomes the Pins entry for thecorresponding PORT. You can view the pre-defined IC-CAP variables by clicking on theModel Variables tab in the Model window.

You may use these pre-defined IC-CAP variables in PEL programs and Macros, whichenables you to programmatically change the pin assignments of each PORT. The followingprogram listing is a PEL macro snippet that manipulates pin assignments. Though pinvalues for variables SMU1-4 are pre-defined, you can see that the variables are beingassigned new values before an iccap_func statement is executed.

n = 1

while (n <= 17)

HP4070_SMU1 = n

HP4070_SMU2 = n + 1

HP4070_SMU3 = n + 2

HP4070_SMU4 = n + 3

print "SMU1=", HP4070_SMU1




iccap_func("/Test1/SMU/sweepOrder1", "measure")

n = n + 4

end while

Prober Functions

The HP 4071A driver incorporates the TIS prober control functions as IC-CAP PELfunctions. The TIS prober functions are described briefly in this section. Thetis_prober_init() function is described in detail because its arguments differ slightly fromthe TIS function prober_init(). The remaining functions have the same arguments as theirTIS counterparts. Consult the TIS Function Reference for complete descriptions of allprober commands. All prober functions return 0 when successful, and -1 when they fail.


6

tis_prober_init (selectCode, busAddress, ProberType, InterfaceName)

selectCode - Integer value, range 0 and 7-31. This is the GPIB select code.Setting selectCode and busAddress to 0 retrieves the GPIB select code and busaddress from PCONFIG file.

busAddress - Integer value, range 0-30. This is the GPIB bus address. SettingselectCode and busAddress to 0 retrieves the GPIB select code and bus addressfrom PCONFIG file.

ProberType - String value, 30 characters max. String that specifies the type ofprober. See TIS Function Reference for prober types.

InterfaceName - String value. This is the interface name, either TIS_OFFLINE orTIS_ONLINE.

tis_p_home ()

Used for loading a wafer onto the chuck and moving it to the home position.

tis_p_up ()

Moves the chuck of the wafer prober up.

tis_p_down ()

Lowers the chuck of the wafer prober.

tis_p_scale (xIndex, yIndex)

Defines the X & Y stepping dimensions that are used by the tis_p_move andtis_p_imove functions.

tis_p_move (xCoordinate, yCoordinate)

Moves the chuck to an absolute position.

tis_p_imove (xDisplacement, yDisplacement)

Moves the chuck a relative increment from its current position.

tis_p_orig (xCoordinate, yCoordinate)

Defines the current X & Y position of the chuck. Must be called before calling thetis_p_move or tis_p_imove functions.

tis_p_pos (xPosition, yPosition)

Returns the current X & Y position of the chuck.

tis_p_ink (inkCode)

Calls the inker function of the prober if it is supported.

tis_prober_reset ()

Sends a device clear command to the prober.

tis_prober_status (isRemote, onWafer, lastWafer)

Sends a query to the prober to obtain the Remote/Local control state and theedge sensor contact state. The prober should be initialized with tis_prober_initbefore this function.

tis_prober_get_name (proberModeName)

Sends query to prober to read name of current mode.

tis_prober_get_ba (proberBusAddress)

Sends query to prober to read its bus address.

tis_prober_read_sysconfig (proberType, scba)

Sends query to prober to read its complete interface address includinginstrument type, select code, and bus address.

The following PEL macro example uses the prober functions. For the prober used in thisexample, notice that the operator must manually place the prober into AUTO PROBE modewhile the program is actively querying the prober and it is in remote mode. Also noticethat isRemote, _isOnWafer, and isLastWafer must be parameters that appear in a variablelist such as Model Variables.

status = -1

busAddress = 0

selectCode = 0

proberType = "EG4080X"

interfaceName = "TIS_ONLINE"

stepSizeX = 500

stepSizeY = 300

isRemote = 0

isOnWafer = 0

isLastWafer = 0

dum = 1

! Prober Commands return 0 for success, -1 for failure

dum = tis_prober_reset()

status=tis_prober_init(selectCode,busAddress,proberType,inte

rfaceName)

if (status == 0) then

status = tis_p_scale(stepSizeX, stepSizeY)

print "status =", status

end if


status = tis_prober_status(isRemote, isOnWafer, isLastWafer)

print "status =", status

print "isRemote =", isRemote

end if



7

linput "Align the wafer. Press OK, then press [AUTO PROBE]",

ans

! EG4080X MUST be actively querying bus when AUTO PROBE is

commenced

while (isRemote == 0)

dum2 = tis_prober_status (isRemote, isOnWafer, isLastWafer)

end while

print "isRemote =", isRemote

if (isRemote ==1) then

status = 0

end if

end if


dum = tis_p_orig(5.0,5.0)

n = 1

while (n < 5)

dum = tis_p_move(n,n)

n = n + 1

end while

end if

Instrument Options for the HP 4071A

The following table describes the HP 4071A options and their default values.

HP 4071A Options

Option Description

Use UserSweep

Yes = use user mode sweep. No = use system mode, when all required conditions are met.Default = No.

Hold Time Time to allow for DC settling before starting internal or user sweep. Maximum 655 seconds.Default = 0.

Delay Time Time the instrument waits before taking a measurement at each step of an internal or usersweep. Maximum 65 seconds. Default = 100 msec.

Fast ADCIntegrationMode

Sets the integration mode for fast A/D converter to 0 = Manual, 1 = Short, 2 = Medium, 3 =Long. Default = 2.

Fast ADCIntegrationValue

Sets the integration time in Power Line Cycles (PLC) or number of samples to average forintegration. Allowed values depend on setting for Fast ADC Integration Mode: If IntegrationMode = 0 or 1, samples = 0, 1 to 4096. Default = 1. If Integration Mode = 2, values areignored, time is fixed to 1 PLC. If Integration Mode = 3, time = 0, 1 to 100 PLC. Default = 16.If 0 is entered as the value, the default value is used.

Use SmartFast ADC IntegMode

Yes/No, default = No. Specifying Yes will use Smart mode integration for fast A/D converterfor current measurements. Fast ADC Integ Mode and Fast ADC Integ Value will still be usedfor voltage measurements.


8

Smart FastADC IntegValue

Sets the integration time in Power Line Cycles (PLC) for integration on current measurementswhen Use Smart Fast ADC Integ Mode is Yes. Values can be 0, or 1 to 100 PLC. If 0 is enteredas the value, the default of 16 PLC will be used.

Slow ADCIntegrationMode

Sets the integration mode for high-resolution (slow) A/D converter to 0 = Manual, 1 = Short,2 = Medium, 3 = Long. Default = 2.

Slow ADCIntegrationValue

Sets the integration time in Power Line Cycles (PLC) or number of samples to average forintegration. Allowed values depend on setting for Slow ADC Integration Mode: If IntegrationMode = 0, time = 0, 80E-6 to 20E-3 seconds, or 1 to 100 PLC. Default = 240E-6 If IntegrationMode = 1, time = 0, 80E-6 to 20E-3 seconds. Default = 480E-6. If Integration Mode = 2,values are ignored, time is fixed to 1 PLC. If Integration Mode = 3, time = 0, 1 to 100 PLC.Default = 16. If 0 is entered as the value, the default value is used.

Slow ADCAuto Zero On

Sets SMU auto zero function to 0 = Off or 1= On. When turned on, the offset error is canceledat each measurement. Default = last valid value.

Use SmartSlow ADCInteg Mode

Yes/No, default = No. Specifying Yes will use Smart mode integration for high-resolution(slow) A/D converter for current measurements. Slow ADC Integ Mode and Slow ADC IntegValue will still be used for voltage measurements.

Smart FastADC IntegValue

Sets the integration time in Power Line Cycles (PLC) for integration on current measurementswhen Use Smart Slow ADC Integ Mode is Yes. Values can be 0, or 1 to 100 PLC. If 0 isentered as the value, the default of 16 PLC will be used.

Ground OpenGuardTerminals

Connects guard terminals of unused measurement pins to circuit common. 0 = Disconnectsterminals, any other value connects them. Default = 0.

Pins Sets the PORT that is connected to the test head pins.

Guard Pins Sets the pins to use for guard terminal. Available only for SMU1 and 2.

Fast/SlowADC

Selects ADC. F = high speed (fast), S = high resolution (slow).

Port Filter On Sets the SMU output filter mode, 0 = Off, 1 = On. Higher speed measurement is used whenfilter is off. Overshoot voltage or current is reduced when filter is on. Default = 0.

V Range (0.0= Auto)

Sets the SMU voltage range. For MPSMU, allowed range is -100 to 100, with recommendedrange of 0, 2, 20, 40, 100. For HPSMU, allowed range is -200 to 200, with recommendedrange of 0, 2, 20, 40, 100, 200. Default = 0 (auto range).

I Range (0.0= Auto)

Sets the SMU current range. For MPSMU, allowed range is -0.1 to 0.1, with recommendedrange of 0, 1E-9, 1E-8, 1E-7, 1E-6, 1E-5, 1E-4, 1E-3, 1E-2, 1E-1. For HPSMU, allowed rangeis -1 to 1, with recommended range of 0, 1E-9, 1E-8, 1E-7, 1E-6, 1E-5, 1E-4, 1E-3, 1E-2, 1E-1, 1. Default = 0 (auto range).

PowerCompliance

Sets SMU power compliance in Watts. Allowed range for MPSMU is 0, 0.001 to 2. Allowedrange for HPSMU is 0, 0.001 to 14.

Pulse Mode On Sets pulse mode. NO = off, YES = on.

Pulse Base Sets level of waveform's base for pulsed spot measurements. For MPSMU, allowed range is -0.1 to 0.1. For HPSMU, allowed range is -1 to 1. See Pulse Base, Width, and Period in PulsedSpot Measurements for pulse waveform characteristics.

Pulse Width Sets width of pulse for pulsed spot measurements. Allowed range is 0.0005 to 2.0000seconds. Default = 0.005. See Pulse Base, Width, and Period in Pulsed Spot Measurements forpulse waveform characteristics.

Pulse Period Sets period of pulse for pulsed spot measurements. Allowed range is 0, 0.0050 to 5.0000seconds. Default = 0.2. See Pulse Base, Width, and Period in Pulsed Spot Measurements forpulse waveform characteristics.

Perform Cal? TRUE = IC-CAP invokes calibration routine if a calibration is needed. FALSE = IC-CAP does notinvoke calibration routine if a calibration is needed.

Cal Type Sets the type of calibration routine to perform. Values are OPEN, SHORT, BOTH. BOTHinvokes the OPEN and SHORT calibration routines.

High Pin High voltage pin connection.

Low Pin Low voltage pin connection.

Guard Pins Guard pin connection.

Integ Time Sets the CMU measurement's integration time. Allowed values are 1 = Short, 2 = Medium, 3= Long.

Hold Time Sets the sweep hold time for C-G-V measurement by the CMU. Allowed range is 0 to 650.000seconds. Default = 0.

Delay Time Sets the sweep delay time for C-G-V measurement by the CMU. Allowed range is 0 to 650.000seconds. Default = 0.

Freq Sets the CMU measurement frequency. Allowed values are 1E+3, 1E+4, 1E+5, 1E+6 Hz.

Signal Level Sets the CMU measurement's test signal level. Allowed range is 0 to 2.0 volts (standard), and0 to 20.0 volts (option 001). Default = last valid setting, or 0.03.

High Pins High voltage pin connection.

Low Pins Low voltage pin connection.

Auto Zero On Sets auto zero mode for DVM. 0 = disable, 1 = enable. Default = last valid setting.

Integ Time Sets integration time for DVM. Allowed range is 0, 0.5E-6 to 999999.9E-6 seconds; 1 to 10PLC and 10 to 100 PLC. If set to 0, integration time is set to default value. Default = 0.5E-6.

The following figure is a diagram of the pulse waveform used in pulsed spotmeasurements showing Pulse Base, Pulse Width, and Pulse Period.

Pulse Base, Width, and Period in Pulsed Spot Measurements

HP 4140 pA Meter/DC Voltage Source

The HP 4140 is equipped with 2 DC voltage source units and 1 low current measurementunit. The units take measurements in either the internal system or user sweep mode.

IC-CAP assigns the following names to the units:

VA DC Voltage Source Unit. VA supports internal linear sweeps using step or ramp sweep mode. This unitcan also be used in user sweep mode.

VB DC Voltage Source Unit. VB only sources a constant voltage. If VB is assigned to the main sweep, usersweep mode is required.

LCU pA Current Monitor Unit.

The HP 4140 driver is an example of a driver created using the Open MeasurementInterface. The driver's source code can be found in the files user_meas2.h anduser_meas2.C in the directory $ICCAP_ROOT/src. For information, refer to Prober(measurement) and Matrix Drivers (measurement).

To recognize which data delimiter (CR/LF or Comma) is used, IC-CAP performs a spot I


9

measurement only when an HP 4140 is first accessed (when the Measure command isissued). When the data delimiter is changed, choose Rebuild in the Hardware Setupwindow so that IC-CAP will note the change.

With a ramp sweep, measured current I can be translated into quasi-static C by the

following equation. Use a transform to perform this calculation.

The following table describes the HP 4140 options and their default values, whereapplicable.

HP 4140 Options

Option Description

Use UserSweep

Yes = use user sweep. No = use the instrument's internal sweep. Default = No

Hold Time Time the instrument waits before starting an internal or user sweep. This option directly controlsthe instrument firmware, and overrides similar delay/hold options set in other instrument driversrunning on the same test system. The range is 0.1 to 1999 seconds in 100 msec steps. Default =0.1

Delay Time Time the instrument waits before taking a measurement at each step of an internal or usersweep. This option directly controls the instrument firmware, and overrides similar delay/holdoptions set in other instrument drivers running on the same test system. The range is 0.01 to100 seconds in 10 msec steps. Default = 0.01 seconds

Integ Time Instrument integration time: S (short), M (medium), or L (long). Default = L

Range Specifies the measurement range. 0 is auto range; 1 is range hold; 2 to 12 denotes a currentrange of 1E-2 to1 E-12. For a faster ramp rate, use a fixed range. Default = 0

Use RampSweep

Yes = use ramp sweep. No = use step sweep. With a ramp sweep, both start and stop values areexpanded by 1 point to have the same number of measurement points with a step sweep.Default = No

Ramp Rate The dV/dt value of a ramp sweep. Minimum is 0.001V/s; maximum is 1V/s. Default = 0.5

InitCommand

This command field is used to set the instrument to a mode not supported by the option table.This command is sent at the end of instrument initialization for each measurement. Normal Cescape characters such as \n (new line) are available. Default = none

HP 4141 DC Source/Monitor

The HP 4141 is equipped with 4 stimulus/measurement units (SMU), 2 programmablevoltage source units (VS), 2 voltage monitor units (VM) and 1 non-programmable groundunit. Use a 16059A Adaptor when measuring a device with a 16058A Test Fixture.


SMUn Stimulus/Measurement Unit n (1, 2, 3, 4)

VSn Voltage Source Unit n (1, 2)

VMn Voltage Monitor Unit n (1, 2)


HP 4141 Options

Option Description

Use User Sweep Yes = use user sweep. No = use the instrument's internal sweep. Default = No

Hold Time Time the instrument waits before starting internal or user sweep. This option directlycontrols the instrument firmware, and overrides similar delay/hold options set in otherinstrument drivers running on the same test system. Range is 0 to 650 seconds in 10 msecsteps. Default = 0

Delay Time Time the instrument waits before taking a measurement at each step of an internal or usersweep. This option directly controls the instrument firmware, and overrides similardelay/hold options set in other instrument drivers running on the same test system. Rangeis 0 to 6.5 seconds in 1 msec steps. Default = 0

Integ Time Instrument integration time; set to S (short), M (medium), or L (long). Default = S

FormatParameter (BDCommand)

0=ASCII Data Output, 1=Binary Data Output.Default = 1

Init Command Command field to set the instrument to a mode not supported by the option table. Thiscommand is sent at the end of instrument initialization for each measurement. Normal Cescape characters such as \n (new line) are available. Default = none

HP/Agilent 4142 Modular DC Source/Monitor

The 4142 contains 8 configurable plug-in slots for:

High-power stimulus/measurement units (HPSMU)Medium-power stimulus/measurement units (MPSMU)High current unit (HCU), high voltage unit (HVU)Voltage source units (VS)Voltage monitor units (VM)Analog Feedback units (AFU-not supported by IC-CAP)

The 4142 ground unit (GND) provides a means for connecting device terminals to aground reference and can sink current up to 1.6A. This ground unit cannot beprogrammed or monitored.

Unit names are dependent on the slot they occupy. An SMU (except MPSMU) uses 2 slotsin the mainframe; the value of slot number n is the higher of the 2 slots. IC-CAP assignsthe following names to the units:

MPSMUn Medium Powered Stimulus/Measurement Unit in slot n

HPSMUn High Powered Stimulus/Measurement Unit in slot n

HCUn High Current Stimulus/Measurement Unit in slot n

HVUn High Voltage Stimulus/Measurement Unit in slot n

VSmn Voltage Source Unit m (1 or 2) in slot n

VMmn Voltage Monitor Unit m (1 or 2) in slot n

The 4142 has a total maximum power consumption of 32W for HPSMU, MPSMU, HCU, HVUand VS/VM. If a measurement is performed and the 32W limit is exceeded, themeasurement will not be attempted and IC-CAP will issue an error message. Powerconsumed by the VS/VM unit (HP/Agilent 41424A) is 2.2W at the 20V range and 0.88W atthe 40V range. When using SMUs to source either voltage or current, refer to the Agilent4142 Operation Manual for the actual SMU power calculations.

NoteTo save power, IC-CAP disconnects output switches of unused HCUs and HVUs when they are not usedwith the current Setup.

In the user and the internal system mode, voltage and current pulsed measurements aresupported. Quasi-pulsed spot measurement is not supported by IC-CAP. For information


10

on how to set up a pulsed measurement, refer to the Pulse entries in HP/Agilent 4l42Options.

HCU and 2-channel pulsed measurements are supported with ROM version 3.0 and later;HVU is supported with version 4.0 and later; Module Selector requires version 4.1.

SMU

Current-forced SMUs of the same type can be connected in parallel to increase the outputcurrent. Use SYNC sweep if you want double current at each sweep point. System Sweepcan be used for 2 HPSMUs; however, User Sweep must be used for 2 HCUs. To avoid awarning message, set the system variable PARALLEL_INPUT_UNITS_OK to True.

HCU

An HCU can force up to 10A with 10V in the pulse mode only. Its pulse base is fixed tozero and it cannot force a constant value. Both 1- and 2-channel measurements aresupported with an HCU.

1-Channel Pulse Because an HCU can force only a pulse, an HCU can be used withoutplacing its name in the pulse unit field in the Instrument Options folder. This is called animplicit pulse channel and its pulse width and period are taken from the InstrumentOptions folder. The pulse base is always set to zero for an implicit pulse channel (HCU).The pulse width and pulse period of an HCU have a different specification from other units.The pulse width must be 0.1 to 1 msec; the pulse period must be 10 to 500 msec; thepulse duty must be 10 percent or less when its output or compliance current is 1A or less,and must be 1 percent or less when its current is more than 1A.

If an HCU is specified as the pulse unit explicitly in the Instrument Options folder, this iscalled an explicit pulse channel and the pulse base in the Instrument Options folder mustbe set to zero.

2-Channel Pulse When 2 pulsed channels are used, the primary channel must be anHCU; the secondary channel can be an HCU, SMU, or VS-it cannot be an HVU. Forinformation on the 2-channel configuration, refer to the following table.

2-Channel Options

Channel Primary Secondary

Pulse Unit HCU only HCU/SMU/VS

Pulse Width 0.1 to 0.8 msec; from Instrument Options folder approximately 1 msec

Pulse Period from Instrument Options folder from Instrument Optionsfolder

Pulse Base 0 only from Instrument Optionsfolder

Declared implicit from Instrument Optionsfolder

HVU

An HVU can force up to 1000V with 10 mA in either the constant or the pulsed mode. Thisunit has the same specification about the pulse width, pulse period, and pulse duty asother SMUs.

An HVU is a unipolar source that requires the output polarity be set before you set itsoutput value. An internal sweep from the minus-to-plus or from the plus-to-minus regionis impossible; set the Use User Sweep option to Yes, if such a sweep range is necessary.

To perform the self test and calibration, the INTLK switch must be closed for an HVU. Atthe start and end of each measurement, IC-CAP instructs all used units to force zero forsafety reasons. The shock hazard lamp of the HP/Agilent 16088B test fixture remains onafter each measurement because the output switch of the used HVU has been closed toforce zero.

VM

A differential voltage measurement of a VM unit is supported by supplying a commandstring to the Init Command field in the Instrument Options folder. If a VM unit is in slot 8,add the command string "VM 8,2;" to the Init Command field. This sets the VM unit at slot8 to a differential mode where it measures the differential voltage of VM18 versus VM28.Then add an output for VM18 (not VM28) to the Setup. When simulating this differentialmode VM, VM18 should correspond to the + Node to have the same polarity betweenmeasurement and simulation.

The following table describes the HP/Agilent 4142 options and their default values, whereapplicable.

HP/Agilent 4l42 Options


11

Option Description

Use UserSweep

Yes = use user mode sweep. No = use system mode, when all required conditions are met.Default = No

Hold Time Time to allow for DC settling before starting internal or user sweep. This option directlycontrols the instrument firmware, and overrides similar delay/hold options set in otherinstrument drivers running on the same test system. Maximum 655 seconds. Default = 0

Delay Time Time the instrument waits before taking a measurement at each step of an internal or usersweep. This option directly controls the instrument firmware, and overrides similar delay/holdoptions set in other instrument drivers running on the same test system. Maximum 65seconds. Default = 100 msec

Integ Time Instrument's integration time; can be set to S (short), M (medium), or L (long). Default = S

Range Specifies the measurement range. 0 specifies auto range. Applies to all SMUs in this 4142.Refer to the Agilent 4142 Operation Manual for definitions of other ranges. Default = 0

SMU Filters ON Yes = filters ON. No = filters OFF. Applies to all SMUs in this 4142. A pulsed unit isautomatically set to filter off. Default = Yes

Pulse Unit Enter name of a pulsed unit when taking pulsed measurements.

Pulse Base Enter value of pulse base.

Pulse Width Enter value of pulse width.

Pulse Period Enter value of pulse period.

Module Control Enter SMU, HCU, or HVU for module selection with option 300. For user relays, enter an exactargument for the ERC command (for example, 2,1,0). When blank, no unit is connected bythe module selector. Refer to the 4142 GPIB Command Reference Manual for the ERCcommand.

Init Command Command field used to set the instrument to a mode not supported by the option table.Command is sent at the end of instrument initialization for each measurement. Normal Cescape characters such as \n (new line) are available. Default = none

PowerCompliance †

Specify power compliance in Watts with 1mW resolution. Specifying 0 (zero) disables powercompliance mode (default).

† Supported for internal sweep mode only (USE USER SWEEP = NO) and DC onlymeasurement setups.This option applies to SMUs only. The allowable range of power compliance depends onthe sweep source (SMU type) and is not monitored by IC-CAP. Refer to instrument'sdocumentation for more details.IC-CAP requires rectangular datasets, thus when a power compliance is specified, theinstrument concludes the measurement at the power compliance limit, but IC-CAP fills thedatasets with the last point measured below power compliance. HP 4145 Semiconductor Parameter Analyzer

The HP 4145 is equipped with the following units:

Four programmable stimulus/measurement units (SMU)Two programmable voltage source units (VS)Two voltage monitor units (VM)

Time-domain measurement is not supported by IC-CAP.

NoteA user-defined function may cause an error E07 in the HP 4145 when the function refers to non-existingsource names. Clear any user-defined functions in the HP 4145 before making a measurement with IC-CAP.


SMUn Stimulus/Measurement Unit n (1, 2, 3, 4)

VSn Voltage Source Unit n (1, 2)

VMn Voltage Monitor Unit n (1, 2)

To recognize which data delimiter (CR/LF or Comma) is used, IC-CAP performs a 2-pointVM measurement only when an HP 4145 is first accessed (when the Measure command isissued). When the data delimiter is changed, choose Rebuild in the Hardware Setupwindow so that IC-CAP will note the change.

NoteThe HP 4145 performs an internal logarithmic sweep only if the number of points per decade is 10, 25 or50; otherwise IC-CAP will force the measurement into User Sweep. If a Setup contains only a single Inputwith a sweep order of 1, IC-CAP will force the measurement into User Sweep.

HP 4145 requires its test fixture lid be closed in User Sweep mode for safety reasons,even though output is low. A Shorting Connector (P/N 04145-61623) can be used tobypass this lid closure check.

NoteThe HP 4145 offers the internal secondary sweep capability known as VAR2. However, the internal SYNCsweep always depends on the primary sweep source VAR1. When a secondary SYNC sweep is desired, useUser Sweep.

NoteAlways fill the Node Name field of each Input in a Setup because the HP 4145 needs a channel namegenerated from a Node Name. The channel names must be unique within a Setup for the HP 4145 internalsweep mode.


HP 4145 Options

Option Description

Use UserSweep


Hold Time Time the instrument waits before starting an internal or user sweep. This option directly controlsthe instrument firmware, and overrides similar delay/hold options set in other instrument driversrunning on the same test system. Range is 0 to 650 sec in 10 msec steps. Default = 0

Delay Time Time the instrument waits before taking a measurement at each step of an internal or usersweep. This option directly controls the instrument firmware, and overrides similar delay/holdoptions set in other instrument drivers running on the same test system. The range is 0 to 6.5sec in 1 msec steps. Default = 0


InitCommand

This command field is used to set the instrument to a mode not supported by the option table.This command is sent at the end of instrument initialization for each measurement. Normal Cescape characters such as \n (new line) are available. Default = none

HP/Agilent 4155 Semiconductor Parameter Analyzer

The HP/Agilent 4155 is equipped with the following units:

Four programmable medium power stimulus/measurement units (MPSMU)Two programmable voltage source units (VS)Two voltage monitor units (VM)


12


MPSMUn Medium Power Stimulus/Measurement Unit n (1, 2, 3, 4)

VSUx Voltage Source Unit n (1, 2)

VMUx Voltage Monitor Unit n (1, 2)

The HP 41501A is an optional SMU and pulse generator expander box that can beattached to and controlled by the 4155. The HP 41501A can be equipped with a highpower stimulus/measurement unit (HPSMU), medium power stimulus/measurement units(MPSMU), and pulse generator units (PGU) (IC-CAP does not support PGUs). Theavailability and combination of these units depends on the expander box option.

NoteWhen making pulsed mode measurements, if you specify an SMU as the unit for an Output, and there isno corresponding SMU unit for an Input, compliance errors will result. The same problem occurs if youspecify Voltage Monitor units. To prevent this from happening, you should define a compliance value forOutput-only SMUs and a measurement range for Voltage Monitor units (VMs) through system variables, asfollows, using the unit name:HRSMUx_COMP HPSMUx_COMP MPSMUx_COMPwhere x = 1, 2, 3, 4, 5, 6VMU1_RANGE_VALUE VMU2_RANGE_VALUE

IC-CAP assigns the following names to the units of the optional HP 41501A:

MPSMUn Medium Power Stimulus/Measurement Unit n (5, 6)HPSMU5 High Power Stimulus/Measurement Unit

A ground unit (GNDU) provides a means for connecting device terminals to a groundreference and can sink up to 1.6A. The ground unit is supported by IC-CAP but will notappear in the Hardware Editor Configuration dialog box. For information on how to use theground unit, refer to the section Adding a Ground Unit (measurement).

In both the user and internal sweep mode, voltage and current pulsed measurements aresupported. Only the SMUs can be specified as pulse units because the PGUs are notcurrently supported. For information on how to set up a pulsed measurement, refer to thePulse options in HP/Agilent 4155 (and HP/Agilent 4156) Option.

NoteThe HP/Agilent 4155 offers the internal secondary sweep capability known as VAR2. However, the internalSYNC sweep always depends on the primary sweep source VAR1. When a secondary SYNC sweep isdesired, use User Sweep.

NoteTo execute a user sweep measurement, IC-CAP sets the HP/Agilent 4155 to the Sampling mode with thenumber of samples equal to 1. The front panel screen activity is turned off at the start of themeasurement and is turned back on after the measurement is completed. Although the 4155 performs aninternal logarithmic sweep if the number of points per decade is 10, 25 or 50, IC-CAP will force themeasurement into the User Sweep for all specified logarithmic sweeps. If a Setup specification contains asingle Input with a sweep order of 1, IC-CAP will force the measurement into User Sweep.

The following table describes the 4155 options and their default values, where applicable.

HP/Agilent 4155 (and HP/Agilent 4156) Option

Option Description

Use UserSweep

Yes = use user mode sweep. No = use system mode, when required conditions are met.Default = No

Hold Time Time delay before starting an internal or user sweep to allow for DC settling. This optiondirectly controls the instrument firmware, and overrides similar delay/hold options set in otherinstrument drivers running on the same test system. Maximum is 655 seconds. Default = 0

Delay Time Time the instrument waits before taking a measurement at each step of an internal or usersweep. This option directly controls the instrument firmware, and overrides similar delay/holdoptions set in other instrument drivers running on the same test system. This value is not usedfor pulsed sweeps. Maximum is 65 seconds. Default = 0

Delay forTimeouts

For long-running measurements (that use a high number of averages, for example) use thisoption to avoid measurement timeouts. Default=0


Pulse Unit Enter the name of a pulsed unit when taking pulsed measurements.

Pulse Base Enter the value of the pulse base.

Pulse Width Enter the value of the pulse width.

Pulse Period Enter the value of the pulse period.

DisplayResolution

N=Normal, E=Extend.Changes the resolution of the measurement data, only available in 4155C/4156C.Default = N

InitCommand

Command field to set the instrument to a mode not supported by the option table. Thiscommand is sent at the end of instrument initialization for each measurement. Normal Cescape characters such as \n (new line) are available. Default = none

PowerCompliance †


† Supported for internal sweep mode only (USE USER SWEEP = NO) and DC onlymeasurement setups.This option applies to SMUs only. The allowable range of power compliance depends onthe sweep source (SMU type) and is not monitored by IC-CAP. Refer to instrument'sdocumentation for more details.IC-CAP requires rectangular datasets, thus when a power compliance is specified, theinstrument concludes the measurement at the power compliance limit, but IC-CAP fills thedatasets with the last point measured below power compliance. HP/Agilent 4156 Precision Semiconductor Parameter Analyzer

The HP/Agilent 4156 is equipped with the following units:

Four programmable high-resolution stimulus/measurement units (HRSMU)Two programmable voltage source units (VS)Two voltage monitor units (VM)

This instrument is designed for Kelvin connections and is capable of low- resistance andlow-current measurements.


HRSMUn High Resolution Stimulus/Measurement Unit n (1, 2, 3, 4)

VSUx Voltage Source Unit n (1, 2)

VMUx Voltage Monitor Unit n (1, 2)

The HP 41501A is an optional SMU and pulse generator expander box that can beattached to and controlled by the 4156. The HP 41501A can be equipped with thefollowing units:

High-power stimulus/measurement unit (HPSMU)


13

Medium power stimulus/measurement units (MPSMU)Pulse generator units (PGU-not supported by IC-CAP)

IC-CAP assigns the following names to the units of the optional HP 41501A:

MPSMUn Medium Power Stimulus/Measurement Unit n (5, 6)

HPSMU5 High Power Stimulus/Measurement Unit

A ground unit (GNDU) provides a means for connecting device terminals to a groundreference and can sink up to 1.6A. The ground unit is supported by IC-CAP but will notappear in the Hardware Editor Configuration dialog box. For information on how to use theground unit, refer to the section Adding a Ground Unit (measurement).

In both the user and internal sweep mode, voltage and current pulsed measurements aresupported. Only the SMUs can be specified as pulse units because PGUs are not currentlysupported. For information on how to set up a pulsed measurement, refer to the Pulseoptions in HP/Agilent 4155 (and HP/Agilent 4156) Option (measurement).

Notes

The HP/Agilent 4156 offers the internal secondary sweep capability known as VAR2. However, theinternal SYNC sweep always depends on the primary sweep source VAR1. When a secondary SYNCsweep is desired, use User Sweep.To execute a user sweep measurement, IC-CAP sets the HP/Agilent 4156 to the Sampling mode withthe number of samples equal to 1. The front panel screen activity is turned off at the start of themeasurement and is turned back on after the measurement is completed.Although the HP/Agilent 4156 performs an internal logarithmic sweep if the number of points perdecade is 10, 25 or 50, IC-CAP will force the measurement into the user sweep for all specifiedlogarithmic sweeps. If a Setup specification contains a single Input with a sweep order of 1, IC-CAPforces the measurement into user sweep.

Options for the HP 4156 are the same as for the HP 4155; refer to HP/Agilent 4155 (andHP/Agilent 4156) Option (measurement). Agilent E5260 Series Parametric Measurement Solutions

Agilent E5260 Series High Speed Measurement Solutions are built around the following:

E5260A 8-slot parametric measurement mainframeE5262A/3A 2-channel source/monitor units

Available Source/Monitor Units (SMUs):

E5290A High Power source/monitor unit (HPSMU)E5291A Medium Power source/monitor unit (MPSMU)

The E5260A 8-slot parametric measurement mainframe holds up to 8 single-slot modules,such as a medium power source/monitor unit (MPSMU), or up to 4 dual-slot modules, suchas a high power source/monitor unit (HPSMU).

The E5262A 2-channel source/monitor unit contains 2 medium power source/monitor units(SMUs).

The E5263A 2-channel source/monitor unit contains 1 high power and 1 medium powerSMU.

If you install 4 HPSMUs into the E5260A mainframe, you can output 1 Amp of current fromeach of these units simultaneously.

The E5260A/B mainframe's ground unit (GNDU) provides a means for connecting deviceterminals to a ground reference. The GNDU will sink 4 amps of current without having toworry about any resistive ground rise issues. This ground unit cannot be programmed ormonitored.

Unit names are dependent on the slot they occupy. A high power SMU occupies 2 slots inthe mainframe, a medium or a high resolution SMU occupies 1 slot; the value of slotnumber n is the higher of the 2 slots. IC-CAP assigns the following names to the units:



The E5260A 8-slot parametric measurement mainframe has a total maximum powerconsumption of 80W for all plug-in modules. The total maximum power consumption limitsfor the E5262A and E5263A are 8W and 24W respectively. If a measurement is performedand the power limitation is exceeded, the measurement will not be attempted and IC-CAPwill issue an error message.

HPSMU

The high power source monitor units will provide up to 50 milliamps of current at ±200volts and 1 amp of current at ±40 volts. Up to 4 HPSMUs can be used at one time in theE5260A mainframe. See manual for complete measurement and force rangesspecifications such as resolution and measurement accuracy.

MPSMU

The medium power source monitor units will provide up to 20 milliamps of current at±200 volts and 200 milliamps of current at ±20 volts. Up to 8 MPSMUs can be used at onetime in the E5260A. See manual for complete measurement and force rangesspecifications such as resolution and measurement accuracy.

Instrument Options

The following table describes the Agilent E5260A options and their default values, whereapplicable.

Agilent E5260A Options


14

Option Description

Use UserSweep

Yes = use user mode sweep. No = use internal sweep, when all required conditions are met.Default = No




PowerCompliance †


SMU FiltersON

Yes = filters ON, No = filters OFF. Applies to all SMUs in this E5260. Default = No

RangeManager Mode

The Range Manager command is used to avoid potential voltage spikes during current rangeswitching when using autorange. See Instrument Programming Guide † under RM commandfor details. Specify Range Manager mode: 1, 2, or 3 as explained below:

1 = deactivate Range Manager (default)2 = set Range Manager to mode 23 = set Range Manager to mode 3

RangeManagerSetting

Set the rate of the Range Manager command.Allowed values are between 11 and 100. This option is only active when Range Manager Modeis set to 2 or 3.

Enable <SMUname> RangeManager

Enables Range Manager at the setting values entered above for the named SMU. Default =No.

<SMU name>In/Out Range

Specify force (Input Sweep) and Output measurement ranges. Default is autorange (0 or 0/0)for both Input and Output measurement ranges. When an SMU is used in an IC-CAP inputdefinition to force voltage or current, a specific force range may be selected. The forceresolution † will depend on the selected range. When an SMU is used in an IC-CAP outputdefinition to monitor voltage or current, a specific measurement range may be selected. Themeasurement resolution will depend on the selected range. Both fixed (negative rangenumber) and limited auto (positive numbers) ranges are supported. Allowed ranges are SMUdependent and are forced by IC-CAP during initial measurement setup. See instrumentmanual † for allowed values for each SMU. When instrument supports 2 values for setting thesame range, IC-CAP only supports the smaller of the 2. For example, to select a 20 V range,the manual suggests using 12 or 200. Use the value 12, to select that range. Ranges must bein the format ForceRange/OutRange, e.g., 13/15 for a voltage SMU monitoring current meansForce Voltage Range=13 (40 V, 2mV resolution), Output Current Measurement Range=15 (10uA limited autorange).





Disable Self-Cal

Controls the status of the E5260A self-calibration routine during measurements. Yes = self-caldisabled. No= self-cal enabled. Default = No.

Output I/OPort (ERCCommand)

Send the user string with the ERC command

Output I/OPort (ERMCommand)

Send the user string with the ERM command

FormatParameter(FMTCommand)

Select the presicion of the measurement data.4 = 4 bytes binary. Select 4 to get better measurement speed.21 = ASCII data. Select 21 and longer Integ Time to get more accurate measurement data.Default = 4

Delay fortimeouts

Sets the delay before a measurement attempt times out.


† Supported for internal sweep mode only (USE USER SWEEP = NO) and DC onlymeasurement setups.The allowable range of power compliance depends on the sweep source (SMU type) and isnot monitored by IC-CAP. Refer to instrument's documentation for more details.IC-CAP requires rectangular datasets, thus when a power compliance is specified, theinstrument concludes the measurement at the power compliance limit, but IC-CAP fills thedatasets with the last point measured below power compliance.† Agilent E5260A, E5262A, E5263A Technical Overview-see Medium and High Power SMUstechnical specifications.† Agilent E5260A series Programming Guide-Chapter 4 "Command Reference"-Section"Command Parameters" Agilent E5270 Series Parametric Measurement Solutions

Agilent E5270 Series Parametric Measurement Solutions are built around the following:

E5270A 8-slot parametric measurement mainframe (obsolete)E5270B 8-slot parametric measurement mainframeE5272A/3A 2-channel source/monitor units (obsolete)

Available Source/Monitor Units (SMUs):

E5280A High Power source/monitor unit (HPSMU) for E5270A onlyE5280B High Power source/monitor unit (HPSMU) for E5270B onlyE5281A Medium Power source/monitor unit (MPSMU) for E5270A onlyE5281B Medium Power source/monitor unit (MPSMU) for E5270B onlyE5287A High Resolution source/monitor unit (HRSMU) for E5270B only

The E5270A 8-slot parametric measurement mainframe holds up to 8 single-slot modules,such as a medium power source/monitor unit (MPSMU), or up to 4 dual-slot modules, suchas a high power source/monitor unit (HPSMU).

The E5270B 8-slot parametric measurement mainframe holds up to 8 single-slot modules,such as a medium power source/monitor unit (MPSMU, HRSMU), or up to 4 dual-slotmodules, such as a high power source/monitor unit (HPSMU).

The E5272A 2-channel source/monitor unit contains 2 medium power source/monitor units(SMUs).

The E5273A 2-channel source/monitor unit contains 1 high power and 1 medium powerSMU.

If you install 4 HPSMUs into E5270A/B mainframes, you can output 1 Amp of current fromeach of these units simultaneously.

The E5270A/B mainframe's ground unit (GNDU) provides a means for connecting device


15

terminals to a ground reference. The GNDU will sink 4 amps of current without having toworry about any resistive ground rise issues. This ground unit cannot be programmed ormonitored.

Unit names are dependent on the slot they occupy. A high power SMU occupies 2 slots inthe mainframe, a medium or a high resolution occupies 1 slot; the value of slot number nis the higher of the 2 slots. IC-CAP assigns the following names to the units:



HRSMUn High Resolution Source/Monitor Unit in slot n (E5270B only)

The E5270A and E5270B 8-slot parametric measurement mainframes have a totalmaximum power consumption of 80W for all plug-in modules. The total maximum powerconsumption limits for the E5272A and E5273A are 8W and 24W respectively. If ameasurement is performed and the power limitation is exceeded, the measurement willnot be attempted and IC-CAP will issue an error message.

HPSMU

The high power source monitor units will provide up to 50 milliamps of current at ±200volts and 1 amp of current at ±40 volts. Up to 4 HPSMUs can be used at one time in theE5270A mainframe. Since SMUs characteristic may vary with version, see manual forcomplete measurement and force ranges specifications such as resolution andmeasurement accuracy.

MPSMU

The medium power source monitor units will provide up to 20 milliamps of current at±100 volts and 100 milliamps of current at ±20 volts (200 mA for the E5281A). Up to 8MPSMUs can be used at one time in the E5270A and E5270B mainframes. Since SMUscharacteristic may vary with version, see manual for complete measurement and forceranges specifications such as resolution and measurement accuracy.

HRSMU

The medium power/high resolution source monitor units provide up to 20 milliamps ofcurrent at ±100 volts and 100 milliamps of current at ±20 volts. Up to 8 HRSMUs can beused at one time in the E5270B mainframe. In the lowest current range, 10 pA, HRSMU'scurrent force resolution can be as low as 5 fA with a measurement resolution as low as 1fA.

Instrument Options

The following table describes the Agilent E5270A/B options and their default values, whereapplicable.

Agilent E5270A/B Options


16

Option Description

Use UserSweep





PowerCompliance †


SMU FiltersON

Yes = filters ON, No = filters OFF. Applies to all SMUs in this E5270. Default = No

RangeManager Mode

Specify Range Manager mode: 1, 2, or 3. 1 = deactivate Range Manager (default) 2 = setRange Manager to mode 2 3 = set Range Manager to mode 3 The Range Manager command isused to avoid potential voltage spikes during current range switching when using autorange.See Instrument Programming Guide † under RM command for details.

RangeManagerSetting

Set the rate of the Range Manager command. Allowed values are between 11 and 100. Thisoption is only active when Range Manager Mode is set to 2 or 3.

<SMU name>A/D converter

Sets A/D converter for higher resolution or higher speed. S = higher speed R = higherresolution (Default)

Enable <SMUname> RangeManager


<SMU name>In/Out Range

Specify force (Input Sweep) and Output measurement ranges. Default is autorange (0 or 0/0)for both Input and Output measurement ranges. When an SMU is used in an IC-CAP inputdefinition to force voltage or current, a specific force range may be selected. The forceresolution † will depend on the selected range. When an SMU is used in an IC-CAP outputdefinition to monitor voltage or current, a specific measurement range may be selected. Themeasurement resolution will depend on the selected range. Both fixed (negative rangenumber) and limited auto (positive numbers) ranges are supported. Allowed ranges are SMUdependent and are forced by IC-CAP during initial measurement setup. See instrumentmanual † for allowed values for each SMU. When instrument supports 2 values for setting thesame range, IC-CAP only supports the smaller of the 2. For example, to select a 20 V range,the manual suggests using 12 or 200. Use the value 12, to select that range. Ranges must bein the format ForceRange/OutRange, e.g., 13/15 for a voltage SMU monitoring current meansForce Voltage Range=13 (40 V, 2mV resolution), Output Current Measurement Range=15 (10uA limited autorange).





Disable Self-Cal

Controls the status of the E5270A self-calibration routine during measurements. Yes = self-caldisabled. No= self-cal enabled. Default = No.






select the presicion of the measurement data.4 = 4 bytes binary. Select 4 to get better measurement speed.21 = ASCII data. Select 21 and longer Integ Time to get more accurate measurement data.Default = 4

Delay fortimeouts



† Supported for internal sweep mode only (USE USER SWEEP = NO) and DC onlymeasurement setups.The allowable range of power compliance depends on the sweep source (SMU type) and isnot monitored by IC-CAP. Refer to instrument's documentation for more details.IC-CAP requires rectangular datasets, thus when a power compliance is specified, theinstrument concludes the measurement at the power compliance limit, but IC-CAP fills thedatasets with the last point measured below power compliance.† Agilent E5270A, E5272A, E5273A Technical Overview-see Medium and High Power SMUstechnical specifications.† Agilent E5270A series Programming Guide-Chapter 4 "Command Reference"-Section"Command Parameters" Agilent B1500A Semiconductor Device Analyzer

The Agilent B1500A Semiconductor Device Analyzer is a modular instrument with a ten-slot configuration that supports both IV and CV measurements.

The B1500A driver supports the following plug-in modules:

B1510A High Power Source Monitor Unit Module (HPSMU) for B1500B1511A Medium Power Source Monitor Unit Module (MPSMU) for B1500B1517A High Resolution Source Monitor Unit Module (HRSMU) for B1500B1520A Multi-Frequency Capacitance Measurement Unit Module (MFCMU) for B1500(combined DC-CV measurements not supported)N1301A SMU CMU Unify Unit (SCUU)

The B1500A driver does NOT support the following plug-in modules:

E5288A Auto Sense and Switch Unit for B1500

HPSMU

The high power source monitor units will provide up to 1 amp of current at ±200 volts. Upto 4 HPSMUs can be used at one time in the B1500A. Since SMUs characteristic may varywith version, see manual for complete measurement and force ranges specifications suchas resolution and measurement accuracy.

MPSMU

The medium power source monitor units will provide up to 100 milliamps of current at±100 volts. Up to 10 MPSMUs can be used at one time in the B1500A. Since SMUscharacteristic may vary with version, see manual for complete measurement and forceranges specifications such as resolution and measurement accuracy.

HRSMU


17

The medium power/high resolution source monitor units provide up to 100 milliamps ofcurrent at ±100 volts. Up to 10 HRSMUs can be used at one time in the B1500A. SinceSMUs characteristic may vary with version, see manual for complete measurement andforce ranges specifications such as resolution and measurement accuracy.

MFCMU

The multi frequency capacitance measurement units provide up to ±25 volts bias. If theSCUU is installed, up to ±100 volts range is available by automatically using SMUconnected to the SCUU as the bias source. Only one MFCMU can be installed and used inB1500A. Since CMU characteristic may vary with version, see manual for completemeasurement and force ranges specifications such as resolution and measurementaccuracy.

Instrument Options

The following table describes the Agilent B1500A options and their default values, whereapplicable.

Agilent B1500A Options

Option Description

Use UserSweep




Integ Time Instrument's integration time; can be set to S (short), M (medium), or L (long). Default = S.This setting is used for both SMU and CMU. For CMU, IC-CAP sets internally as follows:

S (short): Auto mode, N=2 (fixed),M (medium): Power line cycle (PLC) mode, N=1 (fixed),L (long): Power line cycle (PLC) mode, N=16 (defalut, user can set by Averaging Factoroption below)For details, see 'ACT' command description in Programming Guide.

PowerCompliance †

Specify power compliance in Watts with 1mW resolution. Specifying 0 (zero) disables powercompliance mode (default). This is only for SMU.

SMU FiltersON

Yes = filters ON, No = filters OFF. Applies to all SMUs in this B1500A. Default = No

CMU MeasFreq

Measurement Frequency. If the CV_FREQ system variable is defined and specified as 'P'(Parameter) mode input in the input/output setting of the setup, the value set in the input willbe used in measurements instead of the value set in this instrument option. The range is 1kHz to 5 MHz. Default = 1 MHz.Default frequencies of the instrument are 1 k, 2 k, 5 k, 10 k, 20 k, 50 k, 100 k, 200 k, 500 k,1 M, 1.2 M, 1.5 M, 2 M, 2.5 M, 2.7 M, 3 M, 3.2 M, 3.5 M, 3.7 M, 4 M, 4.2 M, 4.5 M, 5 MHz,which are stored in the instrument. If other frequencies are specified, the frequency valuesare added to the default frequency list when calibration is performed with IC-CAP.

CMU Osc Level Test signal level. Allowable voltage levels and resolutions are: Minimum = 10 mVrms;Maximum = 250 mVrms; Resolution = 1 mVrms. Default = 10 mVrms.

CMU Avr.Factor (IntegTime = L)

Averaging Factor which is valid only if Integ Time = L (Long), otherwise the value is ignored.See also Integ Time option description. Maximum = 100. Default = 16

Correction ON Yes = Set correction ON, No = Set correction OFF. Applies to CMU. Calibration (one or someof Open/Short/Load) need to be done before making measurement. Default = Yes

SamplingMode

No = No Sampling, Lin = Linear Sampling.Default = No

SamplingOutput Mode

Sim = Simultaneous, Seq = SequentialDefault = Sim

SequentialSampling Unit

Enter name of a sampling unit when taking sequential sampling measurements.

SamplingBase HoldTime

After the base hold time, the synchronous source SMUs force the bias value.

Sampling BiasHold Time

After the bias hold time, the measurement SMUs start measurement.

RangeManager Mode

Specify Range Manager mode: 1, 2, or 3 for SMU:

1 = deactivate Range Manager (default, autorange)2 = set Range Manager to mode 23 = set Range Manager to mode 3 (only applied to SMUs)The Range Manager command is used to avoid potential voltage spikes during currentrange switching when using autorange. See Instrument Programming Guide † under RMcommand for details. For CMUs, see RC command documentation.The Range Manger Options in this table are shared by both SMUs and CMUs. SMUs mayuse all three values of Range Manager Mode (1, 2 or 3). For CMUs, the following ruleapplies: when Range Manager Mode = 0 or 1 then autorange is active, if RangeManager Mode is >1 then the range is fixed (Range Manager Mode = 2 or 3 areequivalent for CMUs)

RangeManagerSetting

Set the rate of the Range Manager command. Allowed values are between 11 and 100. Thisoption is only active when Range Manager Mode is set to 2 or 3. This is only for SMU.

<SMU/CMUname> A/Dconverter

Sets A/D converter for higher resolution or higher speed.

S = higher speedR = higher resolution (Default)This is only for SMU, and CMU ignores this setting.

Enable<SMU/CMUname> RangeManager


<SMU/CMUname> In/OutRange

Specify force (Input Sweep) and Output measurement ranges for SMU. Default is autorange(0 or 0/0) for both Input and Output measurement ranges. For CMU, only Outputmeasurement range can be settable by user, so here just set the Output part ('/' is notnecessary). If user set in "xxx/yyy" format for CMU, the Input part is ignored and only theOutput part (yyy) is taken. When an SMU is used in an IC-CAP input definition to force voltageor current, a specific force range may be selected. The force resolution † will depend on theselected range. When an SMU is used in an IC-CAP output definition to monitor voltage orcurrent, a specific measurement range may be selected. The measurement resolution willdepend on the selected range. Both fixed (negative range number) and limited auto (positivenumbers) ranges are supported. Allowed ranges are SMU dependent and are forced by IC-CAPduring initial measurement setup. See instrument manual † for allowed values for each SMU.When instrument supports 2 values for setting the same range, IC-CAP only supports thesmaller of the 2. For example, to select a 20 V range, the manual suggests using 12 or 200.Use the value 12, to select that range. Ranges must be in the format ForceRange/OutRange,e.g., 13/15 for a voltage SMU monitoring current means Force Voltage Range=13 (40 V, 2mVresolution), Output Current Measurement Range=15 (10 uA limited autorange). For CMU, RCcommand is used for ranging. See Instrument Programming Guide † for details.

<SMU/CMUname>

Synchronous source channels force the base value.Only allowed in SMU.


18

SamplingBase Value



Pulse Period(common)*

Enter value of pulse period.

Pulse 2 Unit** Enter name of a pulsed unit when taking pulsed measurements.

Pulse 2Base**

Enter value of pulse base.

Disable Self-Cal

Controls the status of the B1500A self-calibration routine during measurements. Yes = self-caldisabled. No= self-cal enabled. Default = No.






Select the presicion of the measurement data.4 = 4 bytes binary. Select 4 to get better measurement speed.14 = 8 bytes binary. Select 14 and longer Integ Time to get more accurate measurementdata.21 = ASCII data. Select 21 for debugging, turn on the debug in hardware window, themeasurement result can be printed in status window.Default = 4

Delay fortimeouts



† Supported for internal sweep mode only (USE USER SWEEP = NO) and DC onlymeasurement setups.The allowable range of power compliance depends on the sweep source (SMU type) and isnot monitored by IC-CAP. Refer to instrument's documentation for more details.IC-CAP requires rectangular datasets, thus when a power compliance is specified, theinstrument concludes the measurement at the power compliance limit, but IC-CAP fills thedatasets with the last point measured below power compliance.† Agilent B1500A Technical Overview-see Medium and High Power SMUs technicalspecifications.† Agilent B1500A series Programming Guide-Chapter 4 "Command Reference"-Section"Command Parameters".* Common in channels.** Only for SMUs.

Capacitance-Voltage Measurement with B1500A

CMU of B1500A (MFCMU) has 2 port (High and Low), providing voltage bias andmeasuring capacitance/conductance/impedance/admittance, and so on. Additional bias tothird or fourth nodes from SMUs can be applied, but current/voltage cannot be measuredsimultaneously when measuring data by CMU. Also, in internal sweep mode, the primarysweep (bias sweep) must be controlled by CMU when using CMU. For user sweep mode,no such limitation of the sweep order is there. As an another aspect of the feature of theCMU, it can sweep frequencies. IC-CAP supports the swept frequency measurement,however if the frequency is the primary sweep in the setup, user sweep mode must beused. For internal sweep mode, frequency must be the second or higher order sweep. Tomake frequency input swept, use 'P' (Parameter mode) with parameter name "CV_FREQ".'F' mode cannot be used for this instrument. Calibration (one or some ofOpen/Short/Load) need to be done before making measurement. After performingcalibration, to make correction ON, set 'Correction ON'=Yes in the instrument optionstable.

Pulsed Measurement with B1500A

B1500A has the capability of (multi-channel) pulsed spot/sweep measurements. Both Ipulse and V pulse are available. Pulse timing among the units is controlled by theinstrument based on user-specified Hold Time, Pulse Width and Period. Pulse Period is acommon setting among the channels.

For all the SMUs (MP/HR/HP) of B1500A:

Pulse Delay = 0s Fixed (no change).Pulse Width set for Pulse1 is used for all the other pulse units.

You can specify Measure (timing) Point by setting The Measurement Delay Time. If notspecified (=Auto), it is determined automatically by a unit of which the pulse rises firstamong the units and the High Speed ADC integration time. If Measurement Delay Time isspecified, no check is done by the instrument about if the pulse is in Base or Top. Otherrestrictions in pulsed measurement include:

In internal sweep mode, if only one single channel is pulsed and the channel is notthe primary sweep, then only one Output can be measured.In internal sweep mode, if the primary sweep is pulsed, then multiple Outputs can bemeasured.When using CMU in pulse measurement, only one single pulse source is available.Internal bias source in MFCMU or SMU with SCUU in B1500 can be the pulsed biassource for pulsed CV measurement.

For more information on pulsed measurement with B1500A, refer to the Instrument UserGuide or the Programming Guide.


19

Sampling Measurement with B1500A

SMUs can be used, SMU can be constant source and make measurement at specified time.Output Sequence can be SEQUENTIAL and SIMULTANEOUS. Please check B1500 andEasyExpert doc for more details.

Inputs:Constant V/I Inputs: set the bias value.Linear Time Input: set the measurement time start, stop and interval.

Outputs:I/V Outputs: get the measurement value.Scope Output: get the measurement time stamp. Optional IC-CAP output, IC-CAP willread back the time stamp data if find any 'T' output in Measurement page.

Configuring the B1500A for IC-CAP Remote Control

Turn on the B1500.1.Login into Windows but do not start the EasyExpert software.2.Start the Agilent_Connection_Expert:3.Select Start > Programs > Agilent_IO_Library_Suites >Agilent_Connection_Expert.Select GPIB0 > Change_Properties, then uncheck the following checkboxes:4.

System_ControllerAuto-Discover_Instruments_Connected_To_The_Interface

Select OK and exit the dialog.Reboot the B1500A when prompted.5.Start the EasyExpert software:6.Select Start > Start_EasyExpert.Do not press the B1500A Start button, but leave the B1500A Start button on thescreen.

NoteFully starting the EasyExpert application would prevent IC-CAP from controlling the B1500A.

Connect the B1500A instrument to the IC-CAP computer via GPIB interface.7.From IC-CAP, rebuild the active instrument list:8.

Select Tools > Hardware Setup > Rebuild.After rebuild is completed, check that the B1500A is in the Active Instrument List.9.Select the instrument and configure its SMU names according to the names used in10.your measurement setups.

Agilent B1505A Power Device Analyzer/Curve Tracer

The Agilent B1505A Power Device Analyzer/Curve Tracer is a modular instrument with aten-slot configuration that supports both IV and CV measurements.

The B1505A driver supports the following plug-in modules:

B1510A High Power Source Monitor Unit Module (HPSMU) for B1505B1512A High Current Source Monitor Unit Module (HCSMU) for B1505B1513A High Voltage Source Monitor Unit Module (HVSMU) for B1505B1520A Multi-Frequency Capacitance Measurement Unit Module (MFCMU) for B1505

The B1505A driver supports the N1258A Module Selector.

The B1505A driver does NOT support the following plug-in modules:

B1511A MPSMUB1517A HRSMUB1525A HV-SPGUB1530A WGFMU

HPSMU

The high power source monitor units will provide up to 1 amp of current at ±200 volts. Upto 2 HPSMUs can be used at one time in the B1505A. Since SMUs characteristic may varywith version, see manual for complete measurement and force ranges specifications suchas resolution and measurement accuracy.

HCSMU

The high current source monitor units will provide up to 20 amp(Pulse),1 amp(DC) ofcurrent and at ±40 volts. Up to 2 HCSMUs can be used at one time in the B1505A. SinceSMUs characteristic may vary with version, see manual for complete measurement andforce ranges specifications such as resolution and measurement accuracy.

HVSMU

The high voltage source monitor units provide up to 4 milliamps of current at ±3000 voltsand 8 milliamps of current at ±1500 volts. Only 1 HVSMU can be used at one time in theB1505A. Since SMUs characteristic may vary with version, see manual for completemeasurement and force ranges specifications such as resolution and measurementaccuracy.

MFCMU

The multi frequency capacitance measurement units provide up to ±25 volts bias. Onlyone MFCMU can be installed and used in B1505A. Since CMU characteristic may vary withversion, see manual for complete measurement and force ranges specifications such asresolution and measurement accuracy.

Module Selector

The Agilent N1258A Module Selector is used to switch the measurementresources(HPSMU, HCSMU, and HVSMU) connected to DUT (device under test). Seemanual for more details.

Instrument Options

The following table describes the Agilent B1505A options and their default values, whereapplicable.

Agilent B1505A Options


20

Option Description

Use UserSweep




Integ Time Instrument's integration time; can be set to S (short), M (medium), or L (long). Default = S.This setting is used for both SMU and CMU. For CMU, IC-CAP sets internally as follows:

S (short): Auto mode, N=2 (fixed),M (medium): Power line cycle (PLC) mode, N=1 (fixed),L (long): Power line cycle (PLC) mode, N=16 (defalut, user can set by Averaging Factoroption below)For details, see 'ACT' command description in Programming Guide.

PowerCompliance †

Specify power compliance in Watts with 1mW resolution. Specifying 0 (zero) disables powercompliance mode (default). This is only for SMU.

SMU FiltersON

Yes = filters ON, No = filters OFF. Applies to all SMUs in this B1505A. Default = No

CMU MeasFreq

Measurement Frequency. If the CV_FREQ system variable is defined and specified as 'P'(Parameter) mode input in the input/output setting of the setup, the value set in the input willbe used in measurements instead of the value set in this instrument option. The range is 1kHz to 5 MHz. Default = 1 MHz.Default frequencies of the instrument are 1 k, 2 k, 5 k, 10 k, 20 k, 50 k, 100 k, 200 k, 500 k,1 M, 1.2 M, 1.5 M, 2 M, 2.5 M, 2.7 M, 3 M, 3.2 M, 3.5 M, 3.7 M, 4 M, 4.2 M, 4.5 M, 5 MHz,which are stored in the instrument. If other frequencies are specified, the frequency values areadded to the default frequency list when calibration is performed with IC-CAP.

CMU OscLevel

Test signal level. Allowable voltage levels and resolutions are: Minimum = 10 mVrms;Maximum = 250 mVrms; Resolution = 1 mVrms. Default = 10 mVrms.

CMU Avr.Factor (IntegTime = L)

Averaging Factor which is valid only if Integ Time = L (Long), otherwise the value is ignored.See also Integ Time option description. Maximum = 100. Default = 16

Correction ON Yes = Set correction ON, No = Set correction OFF. Applies to CMU. Calibration (one or some ofOpen/Short/Load) need to be done before making measurement. Default = Yes

SamplingMode

No = No Sampling, Lin = Linear Sampling.Default = No

SamplingOutput Mode

Sim = Simultaneous, Seq = SequentialDefault = Sim

SequentialSampling Unit

Enter name of a sampling unit when taking sequential sampling measurements.

SamplingBase HoldTime

After the base hold time, the synchronous source SMUs force the bias value.

Sampling BiasHold Time

After the bias hold time, the measurement SMUs start measurement.

RangeManagerMode

Specify Range Manager mode: 1, 2, or 3 for SMU:

1 = deactivate Range Manager (default, autorange)2 = set Range Manager to mode 23 = set Range Manager to mode 3 (only applied to SMUs)The Range Manager command is used to avoid potential voltage spikes during currentrange switching when using autorange. See Instrument Programming Guide † under RMcommand for details. For CMUs, see RC command documentation.The Range Manger Options in this table are shared by both SMUs and CMUs. SMUs mayuse all three values of Range Manager Mode (1, 2 or 3). For CMUs, the following ruleapplies: when Range Manager Mode = 0 or 1 then autorange is active, if Range ManagerMode is >1 then the range is fixed (Range Manager Mode = 2 or 3 are equivalent forCMUs)

RangeManagerSetting

Set the rate of the Range Manager command. Allowed values are between 11 and 100. Thisoption is only active when Range Manager Mode is set to 2 or 3. This is only for SMU.

<SMU/CMUname> A/Dconverter

Sets A/D converter for higher resolution or higher speed.

S = higher speed (Default for HCSMU and HVSMU)R = higher resolution (Default for HPSMU)Only higher speed supported in HCSMU and HVSMU.CMU ignores this setting.

Enable<SMU/CMUname> RangeManager

Enables Range Manager at the setting values entered above for the named SMU. Default = No.

<SMU/CMUname>In/Out Range

Specify force (Input Sweep) and Output measurement ranges for SMU. Default is autorange (0or 0/0) for both Input and Output measurement ranges. For CMU, only Output measurementrange can be settable by user, so here just set the Output part ('/' is not necessary). If userset in "xxx/yyy" format for CMU, the Input part is ignored and only the Output part (yyy) istaken. When an SMU is used in an IC-CAP input definition to force voltage or current, aspecific force range may be selected. The force resolution † will depend on the selected range.When an SMU is used in an IC-CAP output definition to monitor voltage or current, a specificmeasurement range may be selected. The measurement resolution will depend on the selectedrange. Both fixed (negative range number) and limited auto (positive numbers) ranges aresupported. Allowed ranges are SMU dependent and are forced by IC-CAP during initialmeasurement setup. See instrument manual † for allowed values for each SMU. Wheninstrument supports 2 values for setting the same range, IC-CAP only supports the smaller ofthe 2. For example, to select a 20 V range, the manual suggests using 12 or 200. Use thevalue 12, to select that range. Ranges must be in the format ForceRange/OutRange, e.g.,13/15 for a voltage SMU monitoring current means Force Voltage Range=13 (40 V, 2mVresolution), Output Current Measurement Range=15 (10 uA limited autorange). For CMU, RCcommand is used for ranging. See Instrument Programming Guide † for details.

Using ModuleSelector

Yes = enable module selector. No = disable module selector. Default = No.

ModuleSelectorInputs

specifies the modules connected to the input, Syntax is:hvsmu,hcsmu,hpsmu

hvsmu : HVSMU channel numberhcsmu : HCSMU channel numberhpsmu : HPSMU channel numberCharacters will be ignored in IC-CAP, that means 8,6,1 and HVSMU8,HCSMU6,HPSMU1are the same.If one unit is not installed on the instrument, use 0 or empty. 8,0,1 andHVSMU8,,HPSMU1 are the same.For details, see 'ERHPA' command description in Programming Guide.

ModuleSelectorOutput

sets the module selector input-output path to the HPSMU connect, HCSMU connect or HVSMUconnect. Syntax isoutput channel number

If 8,6,1 is the setting of the module selector inputs, specify 8 here means HVSMU8 isthe output of module selector.Characters will be ignored in IC-CAP, that means 8 or HVSMU8 are the same.For details, see 'ERHPP' command description in Programming Guide.

Enable SeriesResistor forHVSMU

Yes = When connects to HVSMU, also connects the series resistor. Only used when the outputis HVSMU.No = don’t connect the series resistor.Default = No


21

For details, see ' ERHPP' command description in Programming Guide.




Pulse Delay Enter value of pulse delay time.

Pulse Period(common)*

Enter value of pulse period.

Pulse 2 Unit** Enter name of a pulsed unit when taking pulsed measurements.

Pulse 2Base**

Enter value of pulse base.

Pulse 2Width**

Enter value of pulse width.

Pulse 2Delay**

Enter value of pulse delay time.

Disable Self-Cal

Controls the status of the B1505A self-calibration routine during measurements. Yes = self-caldisabled. No= self-cal enabled. Default = No.






Select the precision of the measurement data.4 = 4 bytes binary. Select 4 to get better measurement speed.14 = 8 bytes binary. Select 14 and longer Integ Time to get more accurate measurementdata.21 = ASCII data. Select 21 for debugging, turn on the debug in hardware window, themeasurement result can be printed in status window.Default = 4

Delay fortimeouts



† Supported for internal sweep mode only (USE USER SWEEP = NO) and DC onlymeasurement setups.The allowable range of power compliance depends on the sweep source (SMU type) and isnot monitored by IC-CAP. Refer to instrument's documentation for more details.IC-CAP requires rectangular datasets, thus when a power compliance is specified, theinstrument concludes the measurement at the power compliance limit, but IC-CAP fills thedatasets with the last point measured below power compliance.† Agilent B1505A User’s Guide-see High Power SMU, High Current SMU and High VoltageSMU technical specifications.† Agilent B1500A series Programming Guide-Chapter 4 "Command Reference"-Section"Command Parameters".* Common in channels.** Only for SMUs.

NoteIf you encounter an oscillation issue, try in the Instrument Options table to set “SMU Filters ON” = Yesand /or to change “In/Out Range” of the unit from 0 (Auto Range) to a specific value. Inserting a seriesresistor may also work for it.For range value, see "Command Parameters" in Agilent B1500A series Programming Guide.For more advice, see “4 Bytes Data Elements” in Agilent B1500A series Programming Guide.

NoteIn pulse mode measurement, make sure that the Pulse Duty (= pulse width / pulse period) is within alimitation which depends on the measurement range. (e.g. duty <= 1% for 20A range of HCSMU.) Formore information, refer to the instrument documentation.

Capacitance-Voltage Measurement with B1505A

CMU of B1505A (MFCMU) has 2 port (High and Low), providing voltage bias andmeasuring capacitance/conductance/impedance/admittance, and so on. Additional bias tothird or fourth nodes from SMUs can be applied, but current/voltage cannot be measuredsimultaneously when measuring data by CMU. Also, in internal sweep mode, the primarysweep (bias sweep) must be controlled by CMU when using CMU. For user sweep mode,no such limitation of the sweep order is there. As an another aspect of the feature of theCMU, it can sweep frequencies. IC-CAP supports the swept frequency measurement,however if the frequency is the primary sweep in the setup, user sweep mode must beused. For internal sweep mode, frequency must be the second or higher order sweep. Tomake frequency input swept, use 'P' (Parameter mode) with parameter name "CV_FREQ".'F' mode cannot be used for this instrument. Calibration (one or some ofOpen/Short/Load) need to be done before making measurement. After performingcalibration, to make correction ON, set 'Correction ON'=Yes in the instrument optionstable.

Pulsed Measurement with B1505A

B1505A has the capability of (multi-channel) pulsed spot/sweep measurements. Both Ipulse and V pulse are available. Pulse timing among the units is controlled by theinstrument based on user-specified Hold Time, Pulse Delay Time, Pulse Width and Period.Pulse Period is a common setting among the channels.

For HPSMU of B1505A:

Pulse Delay = 0s Fixed (no change).Pulse Width set for Pulse1 is used for all the other pulse units.


22

You can specify Measure (timing) Point by setting Measurement Delay Time. If notspecified (=Auto), it is determined automatically by a unit of which the pulse rises firstamong the units and the High Speed ADC integration time. If Measurement Delay Time isspecified, no check is done by the instrument about whether the pulse is in Base or Top.Other restrictions in pulsed measurement inlcude:

In internal sweep mode, if only one single channel is pulsed and the channel is notthe primary sweep, then only one Output can be measured.In internal sweep mode, if the primary sweep is pulsed, then multiple Outputs can bemeasured.When using CMU in pulse measurement, only one single pulse source is available.Internal bias source in MFCMU can be the pulsed bias source for pulsed CVmeasurement.

For more information on pulsed measurement with B1505A, refer to Instrument UserGuide or Programming Guide.

Sampling Measurement with B1505A

SMUs can be used, SMU can be constant source and make measurement at specified time.Output Sequence can be SEQUENTIAL and SIMULTANEOUS. Please check B1500/B1505and EasyExpert doc for more details.

Inputs:Constant V/I Inputs: set the bias value.Linear Time Input: set the measurement time start, stop and interval.

Outputs:I/V Outputs: get the measurement value.Scope Output: get the measurement time stamp. Optional IC-CAP output, IC-CAP willread back the time stamp data if find any 'T' output in Measurement page.

Configuring the B1505A for IC-CAP Remote Control

Turn on the B1505.1.Login into Windows but do not start the EasyExpert software.2.Start the Agilent_Connection_Expert:3.Select Start > Programs > Agilent_IO_Library_Suites >Agilent_Connection_Expert.Select GPIB0 > Change_Properties, then uncheck the following checkboxes:4.

System_ControllerAuto-Discover_Instruments_Connected_To_The_Interface

Select OK and exit the dialog.Reboot the B1505A when prompted.5.Start the EasyExpert software:6.Select Start > Start_EasyExpert.Do not press the EasyExpert Start button, but leave the EasyExpert Start button onthe screen.

NoteFully starting the EasyExpert application would prevent IC-CAP from controlling the B1505A.

Connect the B1505A instrument to the IC-CAP computer via GPIB interface.7.From IC-CAP, rebuild the active instrument list:8.

Select Tools > Hardware Setup > Rebuild.After rebuild is completed, check that the B1505A is in the Active Instrument List.9.Select the instrument and configure its SMU names according to the names used in10.your measurement setups.

Agilent B2900 Precision Source/Measure Unit

Agilent B2900 is a series of precision SMU (source/measure unit) and includes thefollowing:

B2912A: 2 SMUs models with better resolution (By default, in IC-CAP).B2911A: 1 SMU model with better resolution.B2902A: 2 SMUs models.B2901A: 1 SMU model.

Agilent B2900 Options

The following table describes the Agilent B2900 Options.


23

Option Description

Use User Sweep Yes = Use user mode sweep.No = Use internal sweep, when all required conditions are met.Default = No

Hold Time Time to allow for DC settling before starting internal or user sweep.This option directly controls the instrument firmware, and overrides similardelay/hold options set in other instrument drivers running on the same testsystem.Maximum value is 655 seconds.Default = 0

Delay Time Time the instrument waits before taking a measurement at each step of aninternal or user sweep.This option directly controls the instrument firmware, and overrides similardelay/hold optionsset in other instrument drivers running on the same test system.Maximum value is 65 seconds.Default = 100 msec

Integ Time Instrument's integration time; can be set to S (short), M (medium), N (normal),or L (long).Default = S

SMU Filters ON Yes = Filters ON, No = Filters OFF. Applies to all SMUs in this E2900.Default = No

Sampling Mode No = No Sampling, Lin = Linear SamplingDefault = No

<SMU name> In/OutRange

Specify force (Input Sweep) and Output measurement ranges.Default is autorange (0 or 0/0) for both Input and Output measurement ranges.

<SMU name> Auto SenseMode

Selects the operation mode of the automatic measurement ranging.N = NORMAL, R = RESOLUTION, S = SPEED.Default = N

<SMU name> 2/4 wires Remote sensing must be enabled to use the 4-wire connection (Kelvinconnection).Default = 2

<SMU name> Use Pulse Yes = Pulse ON, No = Pulse OFF.Default = No

<SMU name> Pulse Base Enter value of pulse base.

<SMU name> Pulse Width Enter value of pulse width.

<SMU name> Pulse Period Enter value of pulse period.

<SMU name> Source Wait Sets the gain/offset value used for calculating the source wait time for thespecified SMU.Default = 1/0

<SMU name> Sense Wait Sets the gain/offset value used for calculating the measurement wait time for thespecified SMU.Default = 1/0

Delay for timeouts Sets the delay before a measurement attempt times out.

Init Command Command field to set the instrument to a mode not supported by the optiontable.Command is sent at the end of instrument initialization for each measurement.Normal C escape characters such as \n (new line) are available.Default = none

For more information, refer to the B2900 User Guide and SCPI Reference.

Pulse Measurement in 2 SMUs models

One Pulse UnitThe simultaneous measurement is allowed, if one pulse SMU and one DC SMU is specified.

Two Pulse UnitsThe Simultaneous measurement is allowed, if pulse width/period setting are same in these2 SMUs.

Measurement with Multiple B2900

Internal SweepIf there is an internal sweep, the measure SMU and the internal source SMU must comefrom the same B2900. The SMUs in other B2900 can only set the bias, no measurementallowed.

User Mode SweepThe measurement SMU can be in different B2900, but no pulse is allowed.

Sampling Measurement

Inputs

Constant V/I Inputs: Set the bias value.Linear Time Input: Set the measurement time start, stop and interval.

Outputs

I/V Outputs: Get the measurement value.Scope Output: Fet the measurement time stamp. optional IC-CAP output, IC-CAP willread back the time stamp data, if there is any 'T' output in Measurement page.


24

Capacitance-Voltage metersFor all capacitance-voltage meters, issue the Calibrate command before starting ameasurement, otherwise calibration is carried out automatically at the start of themeasurement. This option directly controls the instrument firmware, and overrides similardelay/hold options set in other instruments' drivers running on the same test system.

Capacitance-voltage meters supported by IC-CAP are:

Defining the Reset State (measurement)HP 4194 Impedance Analyzer (measurement)HP 4271 1 MHz Digital Capacitance Meter (measurement)HP 4275 Multi-Frequency LCR Meter (measurement)HP 4280 1 MHz Capacitance Meter (measurement)HP and Agilent 4284 Precision LCR Meter (measurement)HP and Agilent 4285 Precision LCR Meter (measurement)Agilent E4980A Precision LCR Meter (measurement)Agilent 4294A Precision Impedance Analyzer (measurement)Agilent E4991A RF Impedance Material Analyzer (measurement)Agilent B1500A Semiconductor Device Analyzer (measurement)

Defining the Reset State

Using the prepare_CV_meter.mdl example model file, you can easily define the reset statefor the following instruments:

HP/Agilent 4284HP/Agilent 4285Agilent E4980

The IC-CAP drivers reset instruments to known states prior to configuring them for thecurrent measurement. For the 4284, 4285 and E4980, it sends the *RST command, whichresets the instruments to a known factory state. However, this default state (1V, 1KHzsignal) may cause damage to certain devices between the time the $RST is requested andthe time the requested signal level is set.

To avoid this problem, you can set the variable LCR_RST_MEM or LCR_RST_MEM_<unit>(e.g., LCR_RST_MEM_CM). The 4284, 4285, and E4980A instruments will use the value ofthis variable to set the instrument state. For example, if set to 1, the driver will recallinstrument state 1 instead of *RST.

Using the prepare_CV_meter.mdl example model file, you can programatically set anystate to be the *RST state with a zero signal level. It will also set the variable for you suchthat when a measurement is performed, this programatically set state is recalled insteadof sending *RST.

To prepare a memory location, open and Auto Executemodel_files/misc/prepare_CV_meter.mdl, then enter the 3 values below and click OK.

Name the unit associated with your instrument.1.Identify the memory location (0-9 recommended, but you can use any number from2.0-19 that your instrument supports.)Indicate if you want the unit number to apply to any 4284, 4284, or E4980A, or only3.to those with the specified unit name. This selection essentially chooses betweensetting LCR_RST_MEM or LCR_RST_MEM_<unit>.

HP 4194 Impedance Analyzer

The HP 4194 offers 2 measurement types: impedance analysis and gain-phasemeasurement. These occupy different test connectors on the test set. IC-CAP supports theimpedance analysis type, offering capacitance-voltage and conductance-versus-voltagemeasurements.

The frequency range is 100 Hz to 40 MHz; to extend this to 100 MHz use the impedanceprobe kit. An internal DC bias unit can deliver biases between -40V and +40V. An internaloscillator can deliver between 10 mV and 1V rms.

The HP 4194 driver is an example of a driver created using the Open MeasurementInterface. The driver source code can be found in the files user_meas.hxx anduser_meas.cxx in the directory $ICCAP_ROOT/src. For information, refer to Prober(measurement) and Matrix Drivers (measurement).

IC-CAP assigns the following name to the unit:

CM Capacitance Meter Unit

Notes

The short calibration of the HP 4194 driver is disabled by default because the CV measurementrarely needs this compensation. However, the SHORT_CAL4194 system variable may be defined andset to Yes to enable the short calibration.After a CV measurement is finished, you may notice that a DC bias light on the HP 4194 stays on.This indicates that a bias voltage is still being applied to the test setup. However, the IC-CAP driversets the DC sweep mode's bias voltage for the measurement so the DC bias is set to 0 V when thesweep starts and stops. There are 2 ways you can verify the bias voltage is set to zero. One way isto measure the test setup with a DMM. Another way is to enable IC-CAP's Screen Debug (Tools >Options > Screen Debug) and see that the following commands are being sent to the CV meter:START=0.0;STOP=0.0;NOP=2;MANUAL=0.0;OSC=0.01

SWM3;TRGM2

TRIG


HP 4194 Options


25

Option Description

Use UserSweep


Hold Time Time the instrument waits before starting an internal or user sweep. Default = 0

Delay Time Time delay, in seconds, the instrument waits before taking a measurement at each step of aninternal or user sweep. When biasing the device with an external DC source (e.g., an Agilent4142B or 4156C), the DC source's delay/hold options override this option. Default = 0

Meas Freq Oscillator frequency range 100 Hz to 40 MHz. The 41941A/B impedance probe kit extends thisto 100 MHz. If the CV_FREQ system variable is defined, it must be set equal to this frequency,otherwise an error is reported. Default = 1 MHz

Integ Time Instrument integration time: S (short), M (medium), or L (long). Default = S

Osc Level Test signal level. Allowable voltage levels and resolutions are: Minimum = 10 mV; Maximum =1V. Default = 10mV

Averages Number of averages. Maximum = 256. Default = 1

Delay forTimeouts


InitCommand

Command field to set the instrument to a mode not supported by the option table. Thiscommand is sent at the end of instrument initialization for each measurement. Normal C escapecharacters such as \n (new line) are available. Default = none

HP 4271 1 MHz Digital Capacitance Meter

IC-CAP supports only external bias sources when performing measurements using the HP4271. Both hardware and software calibrations are available. The instrument makesmeasurements in user sweep only. If the CV_FREQ system variable is defined, it must beset equal to 1 MHz before making a measurement with the HP 4271, otherwise an error isreported.

IC-CAP assigns the following name to this unit:



HP 4271 Options

Option Description


Delay Time Time, in seconds, the instrument waits before taking a measurement at each step of an internalor user sweep. When biasing the device with an external DC source (e.g., an Agilent 4142B or4156C), the DC source's delay/hold options override this option. Default = 0

InitCommand

This is a command field to set the instrument to a mode not supported by the option table. Thiscommand is sent at the end of instrument initialization for each measurement. Normal C escapecharacters such as \n (new line) are available. Default = none

HP 4275 Multi-Frequency LCR Meter

The HP 4275 includes an optional internal DC bias source. IC-CAP checks for this internalbias source when you issue the Rebuild command in the Hardware Setup window. For theinternal DC bias to be recognized, the DC BIAS selector switch must be set to Internal.Only hardware calibration is available and the instrument makes measurements in usersweep only.



The test signal level on the HP 4275 can only be set manually with the OSC LEVEL dial andMULTIPLIER switches. This signal level must be set by the user to a reasonable value suchas 10 mV to obtain accurate results, since a high signal level can modulate the DCoperating point. The MULTIPLIER is set to 1 when the instrument is powered up; adifferent setting must be selected manually.

When using the internal DC bias, the bias unit is also included in the CM unit. Therefore,the unit name of this CM unit should also be entered in the Unit fields of both the voltagebias Input and the capacitance Output specifications of the Setup.


HP 4275 Options

Option Description

Hold Time Time the instrument waits before starting an internal or user sweep. Range is 0 to 650 seconds in10 msec steps. Default = 0

DelayTime

Time the instrument waits before taking a measurement at each step of an internal or usersweep. Range is 3 msec to 650 sec. Resolution is in 1 msec steps for the 3 to 65 msec range; 10msec for the 65.01 to 99.99 msec range; and, 100 msec for the 100 msec to 650 sec range.When biasing the device with an external DC source (e.g., an Agilent 4142B or 4156C), the DCsource's delay/hold options override this option. Default = 3 msec

Meas Freq Measurement Frequencies. When the instrument is not equipped with option 004, it acceptsfrequency measurements at 10K, 20K, 40K, 100K, 200K, 400K, 1M, 2M, 4M, and 10M. Whenequipped with option 004, it accepts measurements at 10K, 30K, 50K, 100K, 300K, 500K, 1M, 3M,5M, and 10M. Enter valid frequencies only. If the CV_FREQ system variable is defined, it must beset equal to this frequency, otherwise an error is reported. Because the unit of CV_FREQ is Hz,divide it by 1K for this field. Default = 1 MHz

High Res Enables or disables high resolution mode. Yes = enabled; No = disabled. Default = No

InitCommand

Command field to set the instrument to a mode not supported by the option table. This commandis sent at the end of instrument initialization for each measurement. Normal C escape characterssuch as \n (new line) are available. Default = none

HP 4280 1 MHz Capacitance Meter

The HP 4280 measures the capacitance-voltage characteristics of semiconductor devices.The test signal of the instrument is a 1 MHz sine wave. The HP 4280 also contains a built-in DC bias source with an output capability of 0 to ±100V and a maximum settingresolution of 1 mV. Capacitance-voltage measurements can be taken using this internalvoltage source or an external bias unit. The HP 4280 includes an internal calibrationcapability. Measurements can be made in either internal or user sweep. If the CV_FREQsystem variable is defined, it must be set to 1 MHz before making a measurement withthe HP 4280, otherwise an error is reported.



When using the internal DC bias, this bias unit is also included in the CM unit. Therefore,the unit name of this CM unit should also be entered in the Unit fields of both the voltage


26

bias Input and the capacitance Output specifications of the Setup.


HP 4280 Options

Option Description

Use UserSweep


Hold Time Time the instrument waits before starting internal or user sweep. Range is 0 to 650 seconds in10 msec steps. Default = 3 msec

Delay Time Time delay before each measurement is taken when using internal sweep. Range is 3 msec to650 seconds. Resolution is in 1 msec steps for the 3 to 65 msec range, 10 msec for the 65.01 to99.99 msec range, and 100 msec for the 100 msec to 650 second range. When biasing thedevice with an external DC source (e.g., an Agilent 4142B or 4156C), the DC source's delay/holdoptions override this option. Default = 3 msec

Delay forTimeouts


MeasSpeed

Measuring speed: S (slow), M (medium), or F (fast). Default = S

Sig Level(10, 30)

Signal level: 10 or 30 mV rms. Default = 10

High Res Enables or disables high resolution mode. Yes = enabled. No = disabled. Default = No

Conn Mode Connection mode. When using the HP 4280 internal bias source, set to 10. When using anexternal bias source, connect the source to the EXT-SLOW connector on the HP 4280 rear paneland set the connection mode to 12. Default = 10

InitCommand


HP/Agilent 4284 Precision LCR Meter

The HP/Agilent 4284 is a general purpose LCR meter with a frequency range of 20 Hz to 1MHz. The instrument includes an internal calibration. Options 001 and 006 are supportedby IC-CAP. Option 001 includes a built-in internal bias source. Standard cable lengths are0 and 1 meter; option 006 supports 2 and 4 meter lengths as well. Measurements can bemade in user sweep mode only.




CautionPrior to configuring the HP/Agilent 4284 for the current measurement, the IC-CAP driver resets the 4284to a known state by sending the *RST command. The default reset state (1V, 1KHz signal) may causedamage to certain devices between the time the $RST is requested and the time the requested signal levelis set. To avoid this problem, you can define the reset state. See Defining the Reset State (measurement).


HP/Agilent 4284 Options

Option Description

Hold Time Time the instrument waits before starting an internal or user sweep. Range is 0 to 650 secondsin 10 msec steps. Default = 0

Delay Time Time the instrument waits before each sweep point is measured. The range is 0 to 60 seconds.When biasing the device with an external DC source (e.g., an Agilent 4142B or 4156C), the DCsource's delay/hold options override this option. Default = 0

Meas Freq Measuring frequency. Only a set of frequencies are available. The range is 20 Hz to 1 MHz. Ifthe CV_FREQ system variable is defined, it must be set equal to this frequency, otherwise anerror is reported. Default = 1 MHz

Integ Time Instrument integration time: S (short), M (medium), or L (long). Default = M

Osc Level Test signal level in volts or amps. Allowable voltage levels and resolutions are:Minimum = 5 mVMaximum = 20V with opt 001, 2V otherwiseBetween 5 mV and 200 mV: resolution = 1 mVBetween 200 mV and 2V: resolution = 10 mVBetween 2V and 20V: resolution = 100 mV (Opt. 001 only)Allowable current levels and resolutions are:Minimum level = 50 μA rmsMaximum level = 200 mA rms with opt 001, 20 mA otherwiseBetween 50 μA and 2 mA: resolution = 10 μABetween 2 mA and 20 mA: resolution = 100 μABetween 20 mA and 200 mA: resolution = 1 mA (Opt. 001 only)The Instrument Options folder accepts test signal levels outside these ranges. However, if ameasurement is attempted, an error message is issued and the measurement is not performed.Default = 10m

Osc Mode[V,I]

Specify V (voltage) or I (current). Automatic Level Control (ALC) is not supported. Default = V

Averaging[1-256]

The averaging rate of the instrument. Default = 1

Cable Length Cable length, in meters: 0, 1, 2, or 4. Default = 1M

Delay forTimeouts

For long-running measurements (that use a high number of averages, for example) use thisoption to avoid measurement timeouts.This option also means delay time for each data point because there is no internal sweep.Default=0

InitCommand


HP/Agilent 4285 Precision LCR Meter

The HP/Agilent 4285 is a general purpose LCR meter with a frequency range of 75 kHz to30 MHz. The instrument includes an internal calibration. Option 001, which adds a built-ininternal bias source, is supported by IC-CAP. Measurements can be made in user sweeponly.





27

CautionPrior to configuring the HP/Agilent 4285 for the current measurement, the IC-CAP driver resets the 4285to a known state by sending the *RST command. The default reset state (1V, 1KHz signal) may causedamage to certain devices between the time the $RST is requested and the time the requested signal levelis set. To avoid this problem, you can define the reset state. See Defining the Reset State (measurement).



Option Description


Delay Time Time the instrument waits before each sweep point is measured. Range is 0 to 60 seconds in 1msec steps. When biasing the device with an external DC source (e.g., an Agilent 4142B or4156C), the DC source's delay/hold options override this option. Default = 0

Meas Freq Measuring frequency. Range is 75 kHz to 30 MHz with 100 Hz resolution. If the CV_FREQsystem variable is defined, it must be set equal to this frequency, otherwise an error isreported. Default = 1 MHz


Osc Level Test signal level in volts or amps.The allowable voltage levels and resolutions are:Minimum level = 5 mV rmsMaximum level = 2 V rmsBetween 5 mV and 200 mV: resolution = 1 mVBetween 200 mV and 2V: resolution = 10 mVThe allowable current levels and resolutions are:Minimum level = 200 μA rmsMaximum level = 20 mA rmsBetween 200 μA and 2 mA: resolution = 20 μABetween 2 mA and 20 mA: resolution = 200 μAThe Instrument Options folder accepts test signal levels outside these ranges. However, if ameasurement is attempted, an error message is issued and the measurement is not performed.Default = 10m

Osc Mode[V,I]


Averaging[1-256]


CableLength

Cable length, in meters: 0, 1, or 2. Default = 1

Delay forTimeouts


InitCommand


Agilent E4980A Precision LCR Meter

The Agilent E4980A is a general-purpose LCR meter. The Agilent E4980A is used forevaluating LCR components, materials, and semiconductor devices over a wide range offrequencies (20 Hz to 20 MHz) and test signal levels (0.1 mVrms to 2 Vrms, 50 μA to 20mArms). With Option 001, the E4980A's test signal level range spans 0.1 mV to 20 Vrms,and 50 μA to 200 mArms. Also, the E4980A with Option 001 enables up to ±40 Vrms DCbias measurements (without Option 001, up to ±2 Vrms), DCR measurements, and DCsource measurements using the internal voltage source.




CautionPrior to configuring the Agilent E4980A for the current measurement, the IC-CAP driver resets the E4980to a known state by sending the *RST command. The default reset state (1V, 1KHz signal) may causedamage to certain devices between the time the $RST is requested and the time the requested signal levelis set. To avoid this problem, you can define the reset state. See Defining the Reset State (measurement).




28

Option Description

Hold Time Time the instrument waits before starting an internal or user sweep. Range is 0 to 650 secondsin 10 msec steps. Default = 0

Delay Time Time the instrument waits before each sweep point is measured. The range is 0 to 60 seconds.When biasing the device with an external DC source (e.g., an Agilent 4142B or 4156C), the DCsource's delay/hold options override this option. Default = 0

Meas Freq Measuring frequency. Only a set of frequencies are available. The range is 20 Hz to 1 MHz. Ifthe CV_FREQ system variable is defined, it must be set equal to this frequency, otherwise anerror is reported. Default = 1 MHz


Osc Level Test signal level in volts or amps.Allowable voltage levels and resolutions are:Minimum = 0 mVrmsMaximum = 20 Vrms with opt 001, 2Vrms otherwiseBetween 0 mVrms and 200 mVrms: resolution = 100 μVrmsBetween 200 mVrms and 500 mVrms: resolution = 200 μVrmsBetween 500 mVrms and 1Vrms: resolution = 500 μVrmsBetween 1 Vrms and 2 Vrms: resolution = 1 mVrmsBetween 2 Vrms and 5 Vrms: resolution = 2 mVrms (Opt. 001 only)Between 5 Vrms and 10 Vrms: resolution = 5 mVrms (Opt. 001 only)Between 10 Vrms and 20 Vrms †: resolution = 10 mVrms (Opt. 001 only)† When the test frequency is more than 1 MHz, the maximum oscillator voltage level that canbe set is 15 Vrms.Allowable current levels and resolutions are:Minimum level = 0 ArmsMaximum level = 100 mArms with opt 001, 20 mA otherwiseBetween 0 μArms and 2 mArms: resolution = 1 μArmsBetween 2 mArms and 5 mArms: resolution = 2 μArmsBetween 5 mArms and 10 mArms: resolution = 5 μArmsBetween 10 mArms and 20 mArms: resolution = 10 μArmsBetween 20 mArms and 50 mArms: resolution = 20 μArms (Opt. 001 only)Between 50 mArms and 100 mArms: resolution = 50 μArms (Opt. 001 only)The Instrument Options folder accepts test signal levels outside these ranges. However, if ameasurement is attempted, an error message is issued and the measurement is not performed.Default = 10m

Osc Mode[V,I]


Averaging[1-256]


Cable Length Cable length, in meters: 0, 1, 2, or 4. Default = 1M

Delay forTimeouts

For long-running measurements (that use a high number of averages, for example) use thisoption to avoid measurement timeouts.This option also means delay time for each data point because there is no internal sweep.Default=0

InitCommand


Agilent 4294A Precision Impedance Analyzer

The Agilent 4294A is a precision impedance analyzer designed to measure impedance(inductance, capacitance, and resistance) at frequencies between 40 Hz and 110 MHz. Theinstrument includes an internal calibration.




Frequency cannot be swept using IC-CAP.

The following table describes the Agilent 4294A options and their default values, whereapplicable.

Agilent 4294A Options

Option Description

Use UserSweep

Yes = use user sweep. No = use the instrument's internal sweep. Default = No.

Hold Time Time the instrument waits before starting an internal or user sweep. Default = 0.

Delay Time Time the instrument waits before each sweep point is measured. Range is 0 to 30 seconds.Resolution is 1 msec. Default = 0.

Meas Freq Measuring frequency. Only a set of frequencies are available. Range is 40 Hz to 110 MHz.Resolution is 1 mHz at 40 Hz and 1 kHz at 110 MHz. If the CV_FREQ system variable is defined,it must be set equal to this frequency, otherwise an error is reported. Default = 1 MHz.

Bandwidth Measurement bandwidth. 1 FAST (fastest measurement), 2, 3, 4, 5 PRECISE (highest accuracymeasurement). Default = 1.

Osc Level Test signal level. Allowable voltage levels and resolutions are: minimum = 5 mV, maximum = 1V. Default = 500 mV. Resolution = 1 mV.

Averages [1-256]

Point Averages, minimum 1, maximum = 256. Default = 1

Delay forTimeouts

For long-running measurements (that use a high number of averages, for example) use thisoption to avoid measurement timeouts. Default=0.

Meas Range Selects DC bias range. Three ranges: 1 mA, 10 mA, and 100 mA. Default = 1 mA.

InitCommand

Command field to set the instrument to a mode not supported by the option table. Thiscommand is sent at the end of instrument initialization for each measurement. Normal C escapecharacters such as \n (new line) are available. Default = no entry.

Agilent E4991A RF Impedance/Material Analyzer

The Agilent E4991A is an RF impedance/material analyzer designed to measureimpedance (inductance, capacitance, and resistance) at frequencies between 1 MHz and 3GHz. Measurements can be made in internal sweep mode only.




Frequency cannot be swept using IC-CAP.




29

Option Description

Use UserSweep

Yes = use user sweep, No = use the instrument's internal sweep, default = No.

Hold Time Time the instrument waits before starting an internal or user sweep, default = 0.

Delay Time Time the instrument waits before each sweep point is measured. Range is 0 to 20 seconds,default = 0.

Meas Freq Measuring frequency. Only a set of frequencies are available. Range is 1 MHz to 3 GHz with 1kHz resolution. If the CV_FREQ system variable is defined, it must be set equal to thisfrequency, otherwise an error is reported, default = 1 MHz.

Osc Level Test signal level in volts. Allowable voltage levels and resolutions are: minimum = 5 mV;maximum = 502 mV, default = 100 mV, resolution = 1 mV. The Instrument Options dialogaccepts test signal levels outside these ranges. However, if a measurement is attempted, anerror message is issued and the measurement is not performed.

Averages[1-100]

Point Averages, minimum = 1, maximum = 256, default = 1.

Delay forTimeouts

For long-running measurements (that use a high number of averages, for example), use thisoption to avoid measurement timeouts. Default=0.

Bias CurrentLimit

Bias current limit, minimum 2 μA, maximum 50 μA, resolution 10 μA, default 2 μA.

CalReferencePlane

Used to select the calibration reference plane, either Coaxial (C) or Fixture (F).

InitCommand

Command field to set the instrument to a mode not supported by the option table. Thiscommand is sent at the end of instrument initialization for each measurement. Normal C escapecharacters such as \n (new line) are available. Default = no entry.


30

Network AnalyzersA network analyzer is an integrated stimulus/response test system that measures themagnitude and phase characteristics of a 1-port or multi-port network by comparing theincident signal with the signal transmitted through the device or reflected from its inputs.A network analyzer provides a waveform with a specified attenuation and frequency asinputs to the network or device under test. It then measures the magnitude and the phaseinformation of both the reflected and transmitted waves.

The network analyzers supported by IC-CAP are:

S-parameters in Network Analyzers (measurement)Agilent E5071C ENA Series Network Analyzer (measurement)Agilent PNA Series Vector Network Analyzer (measurement)HP 3577 Network Analyzer (measurement)HP and Agilent 8510 Network Analyzer (measurement)HP and Agilent 8702 Network Analyzer (measurement)HP and Agilent 8719 Network Analyzer (measurement)HP and Agilent 8720 Network Analyzer (measurement)HP and Agilent 8722 Network Analyzer (measurement)HP and Agilent 8753 Network Analyzer (measurement)Wiltron360 Network Analyzer (measurement)Anritsu VectorStar Network Analyzer (measurement)

S-Parameters in Network Analyzers

A network analyzer contains an S-parameter test set that allows automatic selection of S

11, S21, S12, and S22 measurements. S-parameters are used to quantify the signals

involved in microwave design. S, for scattering, describes the act of an energy wave frontentering, exiting, or reflecting off the 2-port network being characterized. Physically, thewave is an electromagnetic flow of energy, a traveling complex voltage wave.Mathematically, the S-parameter is a voltage normalized by the impedance of theenvironment so that its expression relates all information about voltage, current, andimpedance at the same time.

The primary advantage of characterization with S-parameters is that they can bemeasured by terminating a network in its characteristic impedance instead of a short oropen. The following figure mathematically illustrates how S-parameters are defined.

Mathematical Description of S-parameters

Referring to the previous figure, when a network port is terminated so that there is noreflected energy, it is said to be terminated in its characteristic impedance Z0. If at port 2,a2 = 0 because b2 looked into a Z0 load and was not reflected, then

b1 = S11 • a1 + S12 • 0

or

This defines an input reflection coefficient with the output terminated by amatched load (Z0). Similarly,

defines an output reflection coefficient with the input terminated by Z0.

defines the forward transmission (insertion) gain with the output portterminated in Z0.

defines the reverse transmission (insertion) gain with the input port terminatedin Z0.

Graphic Description of S-parameters

The following figure is a graphic description of how S-parameters are defined.


31

NoteThe error terms saved to file during a network analyzer software calibration are not identified by errorcode.The order shown below represents the order in which they are saved and displayed in IC-CAP:0. EDF [directivity]1. EDR [directivity]2. EXF [isolation]3. EXR [isolation]4. ESF [source match]5. ERF [ref freq response]6. ESR [source match]7. ERR [ref freq response]8. ELF [load match]9. ETF [trans freq response]10. ELR [load match]11. ETR [trans freq response]

Agilent E5071C ENA Series Network Analyzer

IC-CAP supports the Agilent E5071C ENA Series RF network analyzer which supports 9Khzto 20GHz range of frequencies.


NWA Network Analyzer Unit

NoteIC-CAP loads the Instrument Options parameters, including Source Power, Sweep Time, and so on, duringan ENA measurement. Since this involves setting values critical to the calibration, an error or warning maybe issued.

The ENA Series network analyzers are recognized when you issue the Rebuild, Measure, orCalibrate command.

This driver only supports Frequency mode with sweep types of Linear, List, Log, andConstant.

Linear sweep mode allows you to specify the start/stop frequencies, number ofpoints, and step size.List sweep mode allows you to sweep up to 202 individual frequencies.Log sweep mode allows you to specify start/stop frequencies, number of decades,and points per decade. The points are log spaced and you can specify a total of 202points.Constant mode allows you to measure 1 individual frequency.

The table that follows describes the E5071C ENA options and their default values, whereapplicable. For more information on options, refer to the E5071C ENA Series NetworkAnalyzer Help file located in the analyzer.

A self-test function is not provided for this instrument.

Calibration

The IC-CAP Calibrate command loads Setup information into the ENA prior to calibrating.When running a measurement afterwards, the calibration set must match IC-CAP's Setupand it must be valid.

Only hardware calibration is supported. The calibration must be either manually executedor executed using dedicated calibration software and saved in a directory in the ENA. Thecalibration and state file must have extension .sta. To measure calibrated data, set theinstrument option Cal Type to H (Hardware) and specify a file name with a .sta extensionin the Instrument Option field Cal/State File Name.

On the ENA mainframe, the default directory for saving and reading calibration and statefiles is D:\State. You can save the calibration file in a different directory and still recall itfrom IC-CAP by setting the system variable ENA_CAL_FILE_PATH to the new directory(use full path such us _D:\my_dir_).

When running a measurement recalling a calibration set, the frequency sweep and theinstrument options should be consistent with the calibration set. Warnings will be issued inthe IC-CAP Status window when relevant ENA measurement settings (such as IFBandwidth or Port Power) differ from the calibration settings.

NoteThe .sta file type should be a save state file that includes the instrument state and the calibration data. Sowhen saving the .sta file inside the instrument for further use, make sure to use the State & Cal save typein the Save/Recall menu.

The ENA has the capability to interpolate between points. Therefore, you can specify a different frequencyrange and number of points during a measurement as long as the measured frequency range is within thecalibrated frequency range. However, be aware that a loss in accuracy occurs due to interpolation.

Agilent E5071C ENA Options


32

Option Description

Use UserSweep[Yes/No]

Yes = use user sweep. No = use instrument's internal sweep.Default = No

Hold Time(sec)

Time, in seconds, the instrument waits before each sweep to allow for DC settling.Default = 0

Delay Time(sec)

Time the instrument waits before setting each frequency in user sweep mode.Default = 0

Sweep Time(sec)

Time the instrument takes for each sweep. 0 = AutoDefault = 0

Sweep Type[SA]

S = Stepped mode. A = Analog (ramp) modeDefault = S

Port PowerCoupled[Yes/No]

Yes = Coupled mode. No = Non-Coupled mode.Default = No.When ports are coupled, the Port1 Src Power value is used for both Port 1 and 2. Port2 SrcPower is ignored.

Port1 SrcPower (dBm)

Defines the source Power for Port 1 and 2 when ports are coupled or the source power for Port1 when ports are uncoupled. The power range depends on the ENA model and options.

Port2 SrcPower (dBm)

Defines the source power for Port 2 when ports are uncoupled. This option field is ignoredwhen ports are coupled. The power range depends on the ENA model and options.

Power Slope(db/GHz)

Can be any value between -2 and +2 dB/GHzDefault = 0

IF Bandwidth(Hz)

Range 10 Hz to 500 kHz Nominal settings are: 10, 15, 20, 30, 40, 50, 70, 100, 150, 200, 300,400, 500, 700, 1k, 1.5k, 2k, 3k, 4k, 5k, 7k, 10k, 15k, 20k, 30k, 40k, 50k, 70k, 100k, 150k,200k, 300k, 400k, 500kHzDefault = 1000 HzNote: If an invalid value is specified, the ENA will not round it to the nearest available value. Itwill round up to the next higher value.

Avg Factor[1-1024]

Number of averages per measurement. [1-1024]Default = 1

Cal Type[HN]

H = Hardware calibration. N = No calibrationDefault = H

Cal/State FileName [ .staonly]

Name of .sta file (with stored calibration and instrument state) to be used.Default = none

Use ENACalibrationSettings[Yes/No]

This setting can be set to Yes only if a calibration file is available and Calibration Type is set to H(Hardware).Default = NoWhen set to Yes, IC-CAP loads the calibration and runs the measurement without furtherinitializing the instrument (i.e. without downloading the current Instrument Table settings).Although IC-CAP uses the calibration settings for measurements, it still sets the sweep settings(e.g. Start, Stop, Sweep Type, etc.). Therefore, make sure the requested sweep setting isconsistent with the calibration settings as IC-CAP attempts to run the measurement withoutperforming any frequency range checking. Also note that when this option is set to Yes, thedriver responds as if MEASURE_FAST =Yes (i.e., calibration is loaded only when themeasurement is first run or after errors or warnings occur).

Delay fortimeouts(sec)

For long-running measurements (that use a high number of averages, for example) use thisoption to avoid measurement timeouts.Default = 0

InitCommand

Command field to set the instrument to a mode not supported by the option table. Commandis sent at the end of instrument initialization for each measurement. Normal C escapecharacters such as \n (new line) are available.Default = none

Technical Notes

You can perform averaging by increasing the number of averages or decreasing theIF filter bandwidth. Both methods result in more samples taken at each frequencypoint. Decreasing the IF filter bandwidth not only increases the number of samplesbut also the time at each frequency point resulting in a longer sweep time. Increasingthe number of averages, increases the number of sweeps. Although the driversupports both modes, using IF bandwidth for averaging is generally more efficient.Coupled ports have the same source power connected to Port 1 and Port 2 forforward and reverse S-parameter measurements.If you have significant insertion loss due to cables or bias networks, use power slope.Using the appropriate power slope can compensate for insertion loss as the frequencyincreases. However, if the network's return loss is too high, increasing the powerslope will not compensate because the power is reflected back.Step sweep mode is more accurate than analog (ramp) mode, but analog mode istypically faster than step sweep mode. In step sweep mode, RF phase locking isperformed at each frequency, which ensures that the frequency value is veryaccurate. This results in a longer transition time from 1 frequency point to the nextand a longer total sweep time. In analog mode, the RF frequency is swept across thefrequency range and its frequency accuracy depends on the linearity of the VCO(Voltage Controlled Oscillator).Sweep time is the total time to sweep from Start to Stop frequency. Several factorscontribute to sweep time. For example at each point in step mode, sweep time is thesummation of transient time due to phase locking, settling time, and measurementtime, which depends on the IF Bandwidth filter. Although you can specify a sweeptime, you should use auto mode (Sweep Time field = 0). This allows the ENA todetermine the fastest sweep time based on the other settings. To view the actualsweep time, select Sweep Setup/Sweep Time on the ENA application's main window.For additional details on sweep time, see the E5071C ENA's online help.

Agilent PNA Series Vector Network Analyzer

IC-CAP supports the Agilent PNA Series vector network analyzers grouped as the AgilentPNA. The following table lists each analyzer and its frequency range:

Supported PNA Series Vector Network Analyzers

The following table lists the supported PNA Series Vector Network Analyzers.


33

Instrument Name Low Frequency High Frequency Notes

Agilent N5241A 10 MHz 13.5 GHz PNA-X Series



Agilent N5245A 10 MHz 50 GHz PNA-X Series

Agilent N5247A 10 MHz 67 GHz &110GHz

PNA-X Series

Agilent N5221A 10 MHz 13.5 GHz PNA Series



Agilent N5225A 10 MHz 50 GHz PNA Series

Agilent N5227A 10 MHz 67 GHz PNA Series

Agilent E8362C 10 MHz 20 GHz PNA Series




Agilent N5250C 10 MHz 110 GHz 110 GHz PNA System

Agilent N5230C 300 KHz 50 GHz PNA-L series, frequency range depends on option

Agilent E8356A 300 kHz 3 GHz Discontinued



Agilent E8361A 10 MHz 67 GHz Discontinued


Agilent E8362B 10 MHz 20 GHz Discontinued








Agilent N5250A 10 MHz 110 GHz Discontinued

Notes

IC-CAP does not supports X-parameters measurements with the PNA-X.

Support for Multiport and Pulsed S-parameters Measurements

IC-CAP supports the Multiport and Pulsed S-parameters measurements with any of thePNA's that have the capabilities, as described in the Supported PNA Series Vector NetworkAnalyzers section.

The Examples on pulsed S-Parameter measurements are available at the followinglocations:

/$ICCAP_ROOT/examples/demo_features/3_MEAS_ORGANIZE_n_VERIFY_DATA/0_MASTER_FILES/30_DEEMBEDDING/more/9_S_Y_Z_conversions_2port_and_Nport.mdl/$ICCAP_ROOT/examples/demo_features/3_MEAS_ORGANIZE_n_VERIFY_DATA/0_MASTER_FILES/20_NWA_CAL_VERIFICATION/4Port/CAL_VERIFY_ENA_4port_MASTERFILE_demodata.mdl/$ICCAP_ROOT/examples/demo_features/3_MEAS_ORGANIZE_n_VERIFY_DATA/0_MASTER_FILES/20_NWA_CAL_VERIFICATION/4Port/CAL_VERIFY_PNA_4port_MASTERFILE_demodata.mdl/$ICCAP_ROOT/examples/demo_features/DEPOTS.mdl and then SetupPEL_DEPOTS/DEEMB

You can also find the examples using the Search Examples wizard in demo_featuresavailable at /$ICCAP_ROOT/examples /demo_features/0____FIND_EXAMPLES_IN_DEMO_FEATURES filepath.



NoteIC-CAP loads the Instrument Options parameters, including Source Power, Attenuation, and so on, duringa PNA measurement. Since this involves setting values critical to the calibration, an error or warning maybe issued.

The PNA Series network analyzers are recognized when you issue the Rebuild, Measure, orCalibrate command.

This driver only supports Frequency mode with sweep types of Linear, List, Log, andConstant.

Linear sweep mode allows you to specify the start/stop frequencies, number ofpoints, and step size.List sweep mode allows you to sweep up to 202 individual frequencies.Log sweep mode allows you to specify start/stop frequencies, number of decades andpoints per decade. The points are log spaced and you can specify a total of 202points.Constant mode allows you to measure 1 individual frequency.

The table that follows describes the PNA options and their default values, whereapplicable. For more information on options, refer to the PNA Series Network AnalyzerHelp file located in the analyzer.


NoteSupport for Millimeter wave SystemsWhen IC-CAP initializes the PNA instrument and checks for its availability on the GPIB bus, it also queriesfor the frequency range supported by the PNA (Min and Max frequency values). For systems designed tohandle frequencies higher than 110GHz, it is possible that the frequency range returned by the instrumentis not the supported system range. In this scenario, you can override the Min and Max frequenciesreturned by the PNA by defining the PNA_MIN_FREQ and PNA_MAX_FREQ system variables as described inMeasurement Options (extractionandprog) section. The PNA_MIN_FREQ and PNA_MAX_FREQ systemvariables must be defined at the IC-CAP/Main level.

Calibration

The IC-CAP Calibrate command loads Setup information into the PNA prior to calibrating.When running a measurement afterwards, the calibration set must match IC-CAP's Setupand it must be valid.

Only hardware calibration is supported. The calibration must be either manually executedor executed using dedicated calibration software and saved in a directory in the PNA. Thecalibration file must have extension .cst.


34

NoteThe .cst file type includes the instrument state and a pointer to the internal calset. The .cst file does notsave the calibration coefficients (the internal calset). Do not delete the internal calset referenced by the.cst file otherwise the IC-CAP measurement will issue an error.If you wish to save the calibration coefficients, save the active calset using a .cal file extension. If theinternal calset is accidentally deleted, you can reinstate it by loading the .cal file from the front panel. Dothis BEFORE running an IC-CAP measurement that uses the .cst file.

To measure calibrated data, set the instrument option Cal Type to H (Hardware) andspecify a file name with a .cst extension in the Instrument Option field Cal/State FileName.

On the PNA mainframe, the default directory for saving and reading calibration and statefile is C:\Program Files\Agilent\Network Analyzer\Documents. You can save the calibrationfile in a different directory and still recall it from IC-CAP by setting the System VariablePNA_CAL_FILE_PATH to the new directory. Use a full path, such as C:\my_dir\.

When running a measurement recalling a calibration set, the frequency sweep and theinstrument options should be consistent with the calibration set. Warnings will be issued inthe IC-CAP Status Window when relevant PNA measurement settings (such as IFBandwidth or Port Power) differ from the calibration settings.

NoteThe PNA has the capability to interpolate between points. Therefore, you can specify a different frequencyrange and number of points during a measurement as long as the measured frequency range is within thecalibrated frequency range. However, be aware that a loss in accuracy occurs due to interpolation.

Agilent PNA Options

Option Description

Use UserSweep

Yes = use user sweep. No = use instrument's internal sweep.Default = No

Hold Time Time, in seconds, the instrument waits before each sweep to allow for DC settling.Default = 0

Delay Time Time the instrument waits before setting each frequency in user sweep mode.Default = 0

Sweep Time Time the instrument takes for each sweep. 0 = Auto Default = 0

SweepType[SA]

S = Stepped mode. A = Analog (ramp) modeDefault = S

Port PowerCoupled

Yes = Coupled mode. No = Non-Coupled mode.Default = No.When ports are coupled, the Port Src Power value is used for both Port 1 and 2. Port 2 SrcPower is ignored. Attenuators are also coupled so that Port Src Atten is used for both ports andPort 2 Src Atten is ignored.

Port SrcPower

Defines the source Power for Port 1 and 2 when ports are coupled or the source power for Port1 when ports are uncoupled. The power range depends on the attenuator settings and the PNAmodel and options.

Port 2 SrcPower

Defines the source power for Port 2 when ports are uncoupled. This option field is ignoredwhen ports are coupled. The power range depends on the attenuator settings and the PNAmodel and options.

Port AttenAuto

Yes = Auto mode. No = Non-Auto mode.Default = No.When attenuators are in auto-mode, the PNA will set the most efficient values for theattenuators to obtain the requested output power at the port. In auto-mode, the full powerrange is directly available at the output port. In auto-mode, the instrument options Port SrcAtten and Port 2 Src Atten are ignored.

Port SrcAtten

Possible Values: 0, 10, 20, 30, 40, 50, 60, 70 dBDefault = 0The available range depends on the PNA model. For example, the E8364A attenuator range is0-60 dB. This option is ignored when attenuators are in auto-mode.

Port 2 SrcAtten

Possible Values: 0, 10, 20, 30, 40, 50, 60, 70 dBDefault = 0The available range depends on the PNA model. For example, the E8364A attenuator range is0-60 dB. This option is ignored when attenuators are in auto-mode.

Power Slope Can be any value between -2 and +2 dB/GHzDefault = 0

Dwell Time Sets the dwell time, in seconds, between each sweep point. Only available in Stepped sweeptype.Default = 0 (Auto - PNA will minimize dwell time)

IF Bandwidth Possible Values: 1, 2, 3, 5, 7, 10, 15, 20, 30, 50, 70, 100, 150, 200, 300, 500, 700, 1k, 1.5k,2k, 3k, 5k, 7k, 10k, 15k, 20k, 30k, 35k, 40kDefault = 1000 HzNote: If a invalid value is specified, the PNA will not round it to the nearest available value. Itwill round up to the next higher value.

Avg Factor Number of averages per measurement. [1-1024]Default = 1

Cal Type[HN] H = Hardware calibration. N = No calibrationDefault = H

Cal/State FileName

Name of .cst file (cal file and instrument state) to be used.Default = none

Use PNACalibrationSettings[Yes/No]

This setting can be set to Yes only if a calibration file is available and Calibration Type is set to H(Hardware).Default = NoWhen set to Yes, IC-CAP loads the calibration and runs the measurement without furtherinitializing the instrument (i.e., without downloading the current Instrument Table settings).Although IC-CAP uses the calibration settings for measurements, it still sets the sweep settings(e.g., Start, Stop, Sweep Type, e.t.c.). Therefore, make sure the requested sweep setting isconsistent with the calibration settings as IC-CAP attempts to run the measurement withoutperforming any frequency range checking. Also note that when this option is set to Yes, thedriver responds as if MEASURE_FAST =Yes (i.e., calibration is loaded only when themeasurement is first run or after errors or warnings occur).

Delay fortimeouts

For long-running measurements (that use a high number of averages, for example) use thisoption to avoid measurement timeouts.Default = 0

InitCommand

Command field to set the instrument to a mode not supported by the option table. Command issent at the end of instrument initialization for each measurement. Normal C escape characterssuch as \n (new line) are available.Default = none

Technical Notes

You can perform averaging by increasing the number of averages or decreasing theIF filter bandwidth. Both methods result in more samples taken at each frequencypoint. Decreasing the IF filter bandwidth not only increases the number of samplesbut also the time at each frequency point resulting in a longer sweep time. Increasingthe number of averages, increases the number of sweeps. Although the driversupports both modes, using IF bandwidth for averaging is generally more efficient.Coupled ports have the same source power connected to Port 1 and Port 2 forforward and reverse S-parameter measurements. In addition, the attenuator settingsare coupled.When port attenuators are set to auto mode, the PNA automatically chooses theattenuator value that provides the requested power level at the output port. AccurateS-parameter calibration requires that the attenuator settings do not change during


35

measurements or calibration, therefore auto mode is not recommended.

If you have significant insertion loss due to cables or bias networks, use power slope.Using the appropriate power slope can compensate for insertion loss as the frequencyincreases. However, if the network's return loss is too high, increasing the powerslope will not compensate because the power is reflected back.Step sweep mode is more accurate than analog (ramp) mode, but analog mode istypically faster than step sweep mode. In step sweep mode, RF phase locking isperformed at each frequency, which ensures that the frequency value is veryaccurate. This results in a longer transition time from 1 frequency point to the nextand a longer total sweep time. In analog mode, the RF frequency is swept across thefrequency range and its frequency accuracy depends on the linearity of the VCO(Voltage Controlled Oscillator).Sweep time is the total time to sweep from Start to Stop frequency. Several factorscontribute to sweep time. For example at each point in step mode, sweep time is thesummation of transient time due to phase locking, settling time, dwell time, andmeasurement time, which depends on the IF Bandwidth filter. Although you canspecify a sweep time, you should use auto mode (Sweep Time field = 0). This allowsthe PNA to determine the fastest sweep time based on the other settings. To view theactual sweep time, select Sweep/Sweep Time on the PNA application's main window.For additional details on sweep time, see the PNA's online help.Dwell time is the time spent at each frequency point before sampling starts. For mostapplications, you should set dwell time to auto mode. In auto mode, the PNAincreases the dwell time as the sweep time increases to comply the total sweep time.If long delays are present in the circuit and additional settling time is needed, set thedwell time to an appropriate value.

Dwell time is not active in analog mode-only in step mode. If the sweep time inanalog mode is increased significantly (because of a setting), the PNA caninternally switch to step mode and set an optimum value for the dwell time.

HP 3577 Network Analyzer

The HP 3577 has a frequency range of 5 Hz to 200 MHz (100 kHz to 200 MHz with HP35677A/B S-Parameter Test Set). The RF source is an integral part of this instrument; DCbias levels must be supplied from external sources.



Because this instrument does not offer full 2-port calibration, IC-CAP provides a popular12-term correction for this instrument that is widely used for 2-port measurements.Manual operation is required to measure standards interactively. Separate calibration datacan be obtained for each Setup; the data is saved and retrieved when Setups are writtento or read from files.

Though IC-CAP supports the HP 3577A and B models, the Discrete Sweep capability of HP3577B is not available with IC-CAP. Therefore, the log and list frequency sweeps must beperformed as a User Sweep.

For most 2-port AC measurements, the network analyzer units must be biased with acurrent or voltage source to supply DC power to the DUT. A DC analyzer can be used forthis. Therefore, a typical S-parameter measurement Setup specification would use the unitname of the network analyzer unit (NWA) in the Unit field of the Output and the unitnames of the DC analyzer units in the Unit fields of the biasing Inputs.

NoteThe HP 35677A/B S-Parameter Test Set has a maximum DC bias range of ±30 V and ±20m A with somedegradation of RF specifications; ±200 mA damage level.

The measurement methods, listed in the following table, are selected by setting the UseUser Sweep and Use Fast CW flags in the HP 3577 Instrument Options folder.

HP 3577 Measurement Modes

Mode Description

Slow CWSweepMode

Use User Sweep = Yes. Use Fast CW = No The instrument sets each frequency then measures all4 S-Parameters. Although somewhat slow, this method has the advantage of gathering all of theparameters for a frequency at approximately the same time.

Fast CWSweepMode

Use User Sweep = Yes. Use Fast CW = Yes This mode is faster than Slow CW Sweep because itperforms just 2 user sweeps. The instrument first measures the forward parameters (S 11 ~~

and S 21), then changes the test set direction and measures the reverse parameters (S 12 and S

22).

Single FreqCW Mode

Use User Sweep = Yes. Use Fast CW = No The instrument performs a spot frequencymeasurement. Except for the number of frequencies, this mode is the same as the Slow CWSweep Mode.

InternalSweepMode

Use User Sweep = No Fastest available sweep type. Sweep must be linear. Values for start, stop,and number of points are stored in the instrument. The number of points in the linear sweepmust match 1 of the HP 3577's allowed number of points choices. When IC-CAP is unable to fit aninternal sweep, it attempts to use the Fast CW Mode.

The following table describes the HP 3577 options and their default values, whereapplicable. For more information on options, refer to the HP 3577 Operating andProgramming Manual.

HP 3577 Options


36

Option Description

Use UserSweep

Yes = use user sweep. No = use instrument's internal sweep. Default = No

Hold Time Time, in seconds, the instrument waits before each sweep to allow for DC settling. Default = 0

Delay Time Time the instrument waits before setting each frequency in user sweep mode. Default = 100msec

Input A Attn Sets Input A attenuation. Choices are 0 or 20 dB. Default = 20 dB

Input B Attn Sets Input B attenuation. Choices are 0 or 20 dB. Default = 20 dB

Input R Attn Sets Input R attenuation: 0 or 20 dB. Default = 20 dB

Source Power Source signal level. Range is -45 to 15 dBm. Default = -10 dBm

Sweep Time[.05 - 16]

Instrument internal sweep time, in seconds. Default = 100 msec.

IF Bandwidth Instrument receiver resolution, in Hz. Default = 1000 Hz

Use Fast CW Enables Fast CW mode. Default = Yes

Avg Factor [1-256]

Number of averages per measurement. Default = 1

Cal Type[SN] S = Software calibration. N = No calibration. Default = S

Soft CalSequence

Software calibration requires measurement of (L)oad, (O)pen, (S)hort, (T)hru, and optionally(I)solation in a certain order. This string defines the sequence of these standardmeasurements by these letters (L, O, S, T, I). Default = LOST

Delay forTimeouts



The system variables used by the 12-term software calibration are listed in the followingtable. They primarily affect S 11 and S 22 corrections at high frequencies. Load and Short

standards are assumed ideal in the calibration frequency range. These variables can bedefined at Setup or higher levels.

System variables

Variable Description

CAL_OPEN_C0CAL_OPEN_C1CAL_OPEN_C2

Define a capacitance of an Open standard in Farads. This value applied to port 1 andport 2. A second-order polynomial is assumed for its frequency response. Copen =C0 + C1 • F + C2 • F2. Default = 0 (for C0, C1, and C2)

TWOPORT_Z0 Defines impedance of port 1 and port 2, in Ohms. This and the open capacitancevalue are used to calculate open gamma correction data. Also used by TwoPortfunction. Default = 50 Ohms

Note: CAL_OPEN_C is replaced by CAL_OPEN_C0; CAL_Z0 is replaced by TWOPORT_Z0.Use the new variables when possible; the old variables are effective for the softwarecalibration when the new variables are undefined. HP/Agilent 8510 Network Analyzer

The HP/Agilent 8510 is identical to the HP 8753 except:

The 8510A has a frequency range of 45 MHz to 26.5 GHz.The 8510B options can source frequencies up to 100 GHz.The RF source is a separate external instrument.The 8510A does not support frequency list mode-it cannot run internal log and listsweeps.

IC-CAP assumes an A model if the instrument is manually added to the Instrument List (inthe Hardware Setup window) by selecting it and clicking the Add button. For IC-CAP torecognize a newer model, use the Rebuild command or perform a dummy measurement:use a linear sweep with the Use Linear List option set to No. Note that the 8510C istreated as the B model.



NoteIC-CAP loads the Instrument Options parameters, including Source Power, Attenuation, and so on, duringan 8510 measurement. Because this involves setting values critical to the calibration, the followingwarning may be issued: Calibration may be invalid.

If the IC-CAP Calibrate command is used to load Setup information to the 8510 prior to calibrating, thecalibration set must match IC-CAP's Setup and be valid.

For information on the processes listed below, refer to the section, HP 3577 NetworkAnalyzer (measurement).

Use of DC bias sourcesAvailable measurement modesSystem variables used in software calibration

NoteUse the 12-term software calibration carefully at very high frequencies where accuracy of the Loadtermination generally degrades.

The following table describes the 8510 options and their default values, where applicable.For more information on options, refer to the HP 8510 Operating and ProgrammingManual.



37

Option Description

Use UserSweep

Yes = use user sweep. No = use instrument's internal sweep. Default = Yes

Hold Time Time, in seconds, the instrument waits before each sweep to allow for DC settling. Default =0


Port 1 Attn Sets Port 1 attenuation. This option is ignored by the 8510XF. Range is 0 to 90 dB. Default =20 dB

Port 2 Attn Sets Port 2 attenuation. This option is ignored by the 8510XF. Range is 0 to 90 dB. Default =20 dB

Source Power Range is -90 to 30 dbm. Default = -10 dBm

Power Slope Range is 0 to 1.5 dbm/GHz. Default = 0

Fast Sweep(Ramp)

Enables ramp sweep. Default = No

Sweep Time[.05 - 100]

Instrument sweep time. Default = 100 msec


Trim Sweep Adjusts frequency at each band edge. Default = 0

Avg Factor [1-4096]


Cal Type[SHN] S = Software calibration. H = Hardware calibration. N = No calibration. Default = H

Cal Set No. [1-8]

Specifies an instrument calibration set. Default = 1

Soft CalSequence


Delay forTimeouts


Use Linear List Yes = load linear sweeps into the 8510's frequency list instead of 1 of the fixed point counts.(Not available on 8510A.) Default = Yes

Init Command Command field to set the instrument to a mode not supported by the option table. Commandis sent at the end of instrument initialization for each measurement. Normal C escapecharacters such as \n (new line) are available. Default = none

When performing 2 sequential CW measurements that use different CW cal subsets, the8510 may report the error RF UNLOCKED. A system variable is available in IC-CAP, in theMeasurement Options group, to ignore this error:

IGNORE_8510_RF_UNLOCK

When defined as Yes, IC-CAP ignores a temporary and benign RF UNLOCKEDerror from the 8510.

Making Measurements with Uncoupled Ports

To calibrate using the 8510XF driver:

Set input sweeps and instrument options. To set port 1 power, set Source Power. To1.set port 1 power slope, set Power Slope. Set averaging. Ignore the Port 1 and 2Attenuators fields as the 8510XF does not have attenuators.In the Init Command field, type the following command string to set port 2 power2.and slope:PPCOUPLEOFF;POWP2 <power>SLPP2ON <slope>

Example:

PPCOUPLEOFF;POWP2 -20;SLPP2ON 0.05;

sets P2=-20 dB and Power slope 2 to 0.05 dB/GHzClick Calibrate. This downloads the sweep settings, the instrument option settings,3.and sets the 8510XF with uncoupled ports.Perform RF Calibration and save the results in one of the Calsets.4.

When making a measurement using the 8510XF driver, the driver recalls the calibrationdata and the setting used during calibration. If you want to use the same power level andslope, you do not need to make any changes. If you want to change the port 2 powersetting, use the Init Command field as in step 2 (you do not need PPCOUPLEOFF since theports are already off when calibration is recalled). Be aware that the 8510XF will issue awarning if you set a different port power for the measurements. HP/Agilent 8702 Network Analyzer

The HP/Agilent 8702 network analyzer has a frequency range of 300 kHz to 3 GHz (IC-CAP does not support the lightwave analyzer features). Use Option 006 and turn on thefrequency doubler from the front panel if 6 GHz is desired. The RF source is an integralpart of this instrument. For other features, refer to the section, HP/Agilent 8753 NetworkAnalyzer (measurement) because the HP/Agilent 8702 is almost identical to the HP/Agilent8753 in the E/E mode. IC-CAP supports both HP/Agilent 85046A and HP/Agilent 85047AS-Parameter Test Sets for the HP/Agilent 8702.



For most 2-port AC measurements, the network analyzer units must be biased with acurrent or voltage source to supply DC power to the DUT. A DC analyzer can be used tosupply this current or voltage source. Therefore, a typical S-parameter measurementSetup specification would use the unit name of the network analyzer unit (NWA) in theUnit field of the Output and the unit names of the DC analyzer units in the Unit fields ofthe biasing Inputs.

For information on the topics listed below, refer to the section, HP/Agilent 8753 NetworkAnalyzer (measurement).

Measurement modesOptions

NoteThe 8702 occupies 2 GPIB addresses, the instrument itself and the display. The display address isderived from the instrument address by complementing the least significant bit. Hence, if theinstrument is at an even address, the display occupies the next higher address; if the instrument isat an odd address, the display occupies the next lower address.

HP/Agilent 8719 Network Analyzer

The HP/Agilent 8719 is identical to the HP/Agilent 8720 except the 8719 has a frequencyrange of 50 MHz to 13.5 GHz. For information, refer to the next section, HP/Agilent 8720Network Analyzer (measurement).


38

HP/Agilent 8720 Network Analyzer

The HP/Agilent 8720 network analyzer has a frequency range of 50 MHz to 20 GHz. TheRF source and S-parameter test set are an integral part of this instrument. IC-CAPsupports the HP/Agilent 8720 A, B, C, and D models. (The 8720 D is the only model thatsupports uncoupled port power.)



NoteThe 8720 occupies 2 GPIB addresses, the instrument itself and the display. The display address is derivedfrom the instrument address by complementing the least significant bit. Hence, if the instrument is at aneven address, the display occupies the next higher address; if the instrument is at an odd address, thedisplay occupies the next lower address.


Measurement modes for the 8720 are the same as for the 8753; refer to SupportedMeasurement Modes (measurement) for this information.

For system variables used in the software calibration, refer to System variables(measurement) in the HP 3577 section.

The following table describes the 8720 options and their default values, where applicable.


Option Description

Use UserSweep

Yes = use user sweep. No = use instrument's internal sweep Default = No

Hold Time Time, in seconds, that the instrument waits before each sweep to allow for DC settling.Default = 0

Delay Time Time the instrument waits before setting each frequency. Default = 100 msec

Port 1 SourcePower

Range is -65 to 10 dBm. Default = -10.00 dBm

Port 1 PowerRange

Specifies which instrument power range to use. Range is 1 to 12 for models A, B, and C;range is 0 to 11 for model D. (The Hardware calibration is turned off by the instrument whencalibrated Power Range and requested Power Range don't match.) Default = 1

Port 1 AutoPower Range †

Enables auto power ranging on port 1. Default = Yes

Coupled PortPower †

Enables/disables coupled test port power. When disabled, Port 2 options are ignored. Default= Yes

Port 2 SourcePower †

Range is -65 to 10 dBm. Default = -10.00

Port 2 PowerRange †

Specifies which instrument power range to use. Range is 1 to 12 for models A, B, and C;range is 0 to 11 for model D. (The Hardware calibration is turned off by the instrument whencalibrated Power Range and requested Power Range don't match.) Default = 1


Enables auto power ranging on port 2. Default = Yes

Sweep Time Instrument sweep time. A zero sweep time turns on the Auto Sweep Time, which ensures theminimum sweep time. Default = 100 msec

IF Bandwidth(Avg)

Instrument's receiver IF bandwidth. Default = 1000 Hz


Avg Factor [1-999]


Cal Type[SHN] S = Software calibration. H = Hardware calibration. N = No calibration. Default = H

Cal Set No. Models A, B, and C: 1 through 5 specifies which instrument calibration sets to use; 6specifies the active instrument state. Model D: 1 through 32 specifies which instrumentcalibration sets to use; 33 specifies the active instrument state. Default = 1

Soft CalSequence


Delay forTimeouts


Use Linear List Yes = load linear sweeps into the HP/Agilent 8720's frequency list instead of one of the fixedpoint counts. This mode should be faster than using the instruments linear frequency sweep.Default = Yes


† These options apply only when using built-in test set (Model D). HP/Agilent 8722 Network Analyzer

The HP/Agilent 8722 is identical to the HP/Agilent 8720 except for its frequency range-theHP/Agilent 8722 has a frequency range of 50 MHz to 40 GHz. HP/Agilent 8753 Network Analyzer

The HP/Agilent 8753 network analyzer has a frequency range of 300 kHz to 3 GHz (6 GHzwith Option 006). The instrument contains an RF source for frequency sweeps, but DCbias must be supplied from external sources to acquire biased RF data.



IC-CAP supports the HP/Agilent 8753 A, B, C, D, E, and D opt 011 models (D models mustbe firmware revision 6.14 or higher). The standard D and E models have a built-in testset; the A, B, C, and D opt 011 models are used in conjunction with an external test set.IC-CAP supports both the HP/Agilent 85046A and HP/Agilent 85047A S-Parameter TestSets.


39

Notes

IC-CAP cannot differentiate between model D and D opt 011. When using the 8753D with anexternal test set, alias it as an 8753C in the instraliases file in the $ICCAP_ROOT/iccap/lib directory.The 8753 occupies 2 GPIB addresses, the instrument itself and the display. The display address isderived from the instrument address by complementing the least significant bit. Hence, if theinstrument is at an even address, the display occupies the next higher address; if the instrument isat an odd address, the display occupies the next lower address.

The model is recognized when you issue the Rebuild, Measure, or Calibrate command. Ifyou manually add the instrument to the active instrument list (by clicking the Add button),IC-CAP assumes the instrument is an A model until one of the previously describedcommands is issued.

NoteSome early models of the 8753C (ROM 4.00 and 4.01) have GPIB problems that prevent IC-CAP fromfinding this instrument during Rebuild. Add the instrument manually after PRESET when IC-CAP ignoresthis instrument. The model is recognized when Measure or Calibrate is performed.



Hardware calibration is only supported when using Internal Sweep mode or Single FreqCW mode. For measurement modes that do not support internal instrument calibration,software calibration is provided. When software calibration is set in the instrumentoptions, use the Calibrate command to initiate the calibration. IC-CAP will load thefrequency values and options into the instrument and then direct you to connect thevarious calibration standards required to perform the calibration.

The Calibrate command can also be used to download the desired instrument state whenrequesting a hardware calibration You must then calibrate the instrument manually (referto the instrument manual) and store the results in one of the instrument's state registers.With this method there is no need to manually input the instrument state to match the IC-CAP settings.

The measurement modes listed in Supported Measurement Modes are selected by settinga combination of the following fields (details follow) in the 8753 Instrument Optionsfolder:

Use User SweepUse Fast CWUse Linear ListCal Type field

IC-CAP contains routines that compare its sweep values with those stored in the 8753. Incase of discrepancies, IC-CAP prompts you to specify whether the sweeps should bemodified to match the instrument. This may not be practical when variables are includedin the sweep specifications.

Error checking ensures a valid measurement mode. When discrepancies are found, thefollowing changes are made:

For CON frequency, Use User Sweep is set to Yes and Use Fast CW is set to No.For internally calibrated sweeps, Use User Sweep is set to No.When frequency is not the main sweep, Use Fast CW is set to No, Cal Type is set toN, and Use User Sweep is set to Yes.

Refer to HP 3577 Network Analyzer (measurement) for system variables used in thesoftware calibration.

Options for the HP/Agilent 8753 describes the 8753 options and their default values,where applicable. Differences in options when using the 8753 with an external test setversus a built-in test set are noted.

For optimum performance of the HP/Agilent 85047 test set, 6 GHz mode requires SourcePower to be +20dBm. The Instrument Options folder should show 20 as the Source Powerlevel when the Freq Range is 6 GHz. Setting Source Power to less than 20 can cause No IFFound errors in the 8753. Further information on the power requirements for 6 GHzoperation can be found in the instrument's operation manual.

When the test set switches between 3 and 6 GHz operation, the 8753 automaticallychanges Source Power level.

3 to 6 GHz Switching: 20 dBm.6 to 3 GHz Switching: 0 dBm.

When Hardware calibration is used, a specified calibration set recalls the originalcalibration power level. When Software or no calibration is used, the Source Power will beforced to one of the default levels if the test set has to switch modes. When a SourcePower level other than the above forced values is required, perform one of the following:

Make a dummy measurement first to switch the test set to the desired frequencymodeManually switch the test set to the desired frequency modeUse the Calibrate command to download the desired instrument state

NoteThe 8753 may not reflect the power level specified in the Instrument Options folder if the analyzer isin HOLD mode. When the 8753 is in HOLD mode and receives a remote command to switch thefrequency mode of the test set, it postpones switching modes until an actual measurement sweep istriggered. When the Measure or Calibrate command is issued, IC-CAP initializes the state beforetriggering a measurement. Thus IC-CAP will download the power level specified in the InstrumentOptions folder and the analyzer will force it to its default value when the measurement is triggered.

For descriptions of the variables used in software calibration, refer to System variables(measurement) in the HP 3577 Network Analyzer section.

Supported Measurement Modes


40

Mode Description UseUserSweep

UseFastCW

UseLinearList

CalType

Slow CWSweep

IC-CAP performs a spot measurement of 2-port data by setting theinstrument to each frequency point individually and measuring all 4S-parameters. Although slow, this method has the advantage ofgathering all of the parameters for a frequency at approximatelythe same time. Only uncalibrated data can be obtained from thistype of measurement since each frequency point is measured in CWmode. Typically used when frequency is not the primary sweep(Sweep Order ≠ 1).

Yes No Ignored S, N

Fast CWSweep

Similar to Slow CW Sweep, this mode is faster because it firstmeasures the forward parameters (S 11 ~~ and S 21) with a single

sweep, then the reverse parameters (S 12 ~~ and S 22). This is

accomplished by using the dual channel feature of the instrument.As with Slow CW Sweep, instrument calibration is not possible andonly uncalibrated data can be obtained.

Yes Yes Ignored S, N

SingleFreq CW

This is the only user sweep mode capable of acquiring 2-port datausing hardware calibration. A CW mode calibration can beperformed and saved in one of the state registers to be recalledwhen a measurement is executed.

Yes No Ignored H,S,N

InternalSweep

Fastest available sweep type. Sweeps can be linear, log, or list. †

Since this is an internal sweep, hardware calibration is possible. IC-CAP expects that the calibration over the appropriate frequencieshas been completed before the measurement is performed.

No No Yes orNo

H,S,N

Notes^† ^ For linear sweeps, the number of points requested must fit one of the 8753's predefined number ofpoints. If the desired number of points is not one of the legal set values, IC-CAP checks to see if it still canmake a valid measurement by increasing the number of points on the instrument such that data at thedesired frequencies can be acquired. For example, a 300 to 500 kHz sweep in 6 steps internally requiresIC-CAP to set the instrument to 11 points because 11 is a legal value. When IC- CAP is unable to fit aninternal sweep, it attempts to use the Fast CW mode. If CW mode is not desired, set Use Linear List = Yes.For log and list sweeps, set Use Linear List = Yes. This uses the instrument's frequency list capability.Because the 8753 is limited to thirty sub-sweeps, it can store no more than sixty frequencies.Instrument options must match those for which the 8753 was calibrated.

Options for the HP/Agilent 8753

Option Description

Use UserSweep

Yes = use user sweep. No = use instrument's internal sweep Default = No

Hold Time Time, in seconds, the instrument waits before each sweep to allow for DC settling. Default =0

Delay Time Time the instrument waits before setting each frequency. Default = 100 msec

Port 1 Atten † Sets Port 1 attenuation. Range is 0 to 70 dB. Default = 20 dB

Port 2 Atten † Sets Port 2 attenuation. Range is 0 to 70 dB. Default = 20 dB

Source Power† Range is -10 to 25 dbm. Default = -10

Power Slope Models A, B, C, and D opt 11: Range is 0 to 2 dbm/GHz. Default = 0 Models D and E: Rangeis -2 to +2dBm/GHz. Default = 0 dBm/GHz


Sets Port 1 source power level. Range is -85 to +10 dBm. Default = -10 dBm

Port 1 PowerRange[0-7] †

Sets Port 1 source power range. The valid range is 0 to 7. Default = 0


Enables auto power ranging on port 1. Default = No

Coupled PortPower †

Enables/disables coupled test port power. When disabled, Port 2 options are ignored. Default= Yes


Sets Port 2 source power level. Range is -85 to +10 dBm. Default = -10 dBm

Port 2 PowerRange[0-7] †

Sets Port 2 source power range. The valid range is 0 to 7. Default = 0


Enables auto power ranging on port 2. Default = No

Sweep Time Instrument sweep time. Zero sweep time turns on the Auto Sweep Time, which ensures theminimum sweep time. Default = 100 msec

IF Bandwidth(Avg)

Instrument receiver IF bandwidth setting in the Averaging menu. Default = 1000 Hz


Avg Factor [1-999]


Cal Type [SHN] S = Software calibration. H = Hardware calibration. N = No calibration. Default = H

Cal Set No. Models A, B, C, and D opt 11: 1 through 5 specifies which instrument calibration sets to use;6 specifies the active instrument state. Models D and E: 1 through 32 specifies whichinstrument calibration sets to use; 33 specifies the active instrument state. Default = 1

Soft CalSequence


Delay forTimeouts


Use Linear List Yes = load linear sweeps into the 8753 frequency list instead of one of the fixed point counts.This mode should be faster than using the instrument's linear frequency sweep. Default = Yes

Freq Range[36N]†

This option sets Frequency Range to 3 GHz, 6 GHz, or No change. Default = N

Init Command This command field sets the instrument to a mode that is not supported by the option table.This command is sent at the end of instrument initialization for each measurement. Normal Cescape characters such as \n (new line) are available. Default = none

Notes† These options apply only when using external test set (Models A, B, C, and D opt 11).† These options apply only when using built-in test set (Models D and E).

Wiltron360 Network Analyzer

The Wiltron360 network analyzer has a frequency range of 10 MHz to 60 GHz dependingon the RF source. If the frequency sweep requested exceeds the limits of the source, IC-CAP issues an error message, Parameter Out of Range. Check Inputs. The RF source is anintegral component of the system for frequency sweeps, but DC bias must be suppliedfrom external sources to acquire biased RF data.

The Wiltron360 can be added to the active instrument list by issuing the Rebuild commandfrom the Hardware Setup window. If the Wiltron360 is manually added to the activeinstrument list using the Add button, IC-CAP verifies that the instrument is available onthe bus when either the Measure or Calibrate command is first issued.




41

IC-CAP supports only hardware calibration for this instrument. After a broadbandcalibration, the 360 can perform a spot measurement of swept calibrated data-softwarecalibration is not required. This capability also allows a CON frequency input defined in anIC-CAP setup to be used with a broadband calibration. For either measurement method,the requested frequency points must be a subset of the frequency sweep currently set upon the instrument. If the requested frequency point is not part of the instrument sweep,IC-CAP will issue an error message.

The measurement modes listed in Measurement Modes are selected by setting thefollowing fields in the Wiltron360 Instrument Options folder:

Use User SweepCW Mode SetupCal Type

IC-CAP supports recalling calibration sets from the Wiltron360 internal disk drive. The CalFile Name option is provided to recall the desired calibration state from disk. If nocalibration file name is supplied, the current active instrument state is used.

IC-CAP loads the Instrument Options parameters during a measurement. Because thisinvolves setting stimulus values sensitive to the calibration, instrument options mustmatch those for which the Wiltron360 was calibrated; otherwise, the Wiltron360 will issuea Calibration may be invalid message if any of the downloaded stimulus values aredifferent from the current calibration. If this message is displayed, check the InstrumentOptions folder to verify which value is different and modify as appropriate. Use theCalibrate command from the Setup menu to download the options information to theWiltron360 prior to calibrating. This ensures that the calibration will match IC-CAP's Setupand be valid.

Measurement Modes

Mode Description UseUserSweep

CWModeSetup

CalType

CWSweep

Used when frequency is not the primary sweep (Sweep Order = 1). IC-CAP performs a spot measurement of 2-port data by setting theinstrument to each frequency point individually and measuring all S-parameters. A broadband hardware calibration can be performed. Thecalibration does not have to match the IC-CAP sweep exactly; however,the desired swept frequency points must be a subset of the calibratedfrequencies.

Yes No H orN

SingleFreq CW

Used when a CW mode hardware calibration is performed. Yes Yes H

InternalSweep

Linear, log, or list sweeps. Hardware calibration over requestedfrequencies is completed before an IC-CAP measurement is performed.Unlike CW Sweep, the calibration frequencies must match the setup.

No No H orN

The following table describes the Wiltron360 options and their default values, whereapplicable.

Wiltron360 options

Option Description

Use UserSweep

Yes = Use user sweep No = use instrument's internal sweep. Default = No

Hold Time Time, in seconds, the instrument waits before each sweep to allow for DC settling. Default = 0


Port 1 SrcAtten

Sets Port 1 source attenuation. Range is 0 to 70 dB, in 10 dB increments. Default = 0 dB

Port 2 SrcAtten

Sets Port 2 source attenuation. Range is 0 to 70 dB, in 10 dB increments. Default = 0 dB

Port 2 TestAtten

Sets test port attenuation (port 2). Range is 0 to 40 dB, in 10 dB increments. Default = 0 dB

Source Power Range is dependent on test set used. Default = 0 dBm

IF Bandwidth[NRM]

Sets instrument's receiver IF Bandwidth. N = Normal R = Reduced M = Minimum. Default = N

Avg Factor[1-4095]

Sets number of averages per measurement Default = 1

Use CW ModeSetup

Indicates to IC-CAP that NWA has been set up in single point (CW) measurement mode.Default = No

Cal Type[HN] H = Hardware calibration N = No calibration Default = H

Cal File Name Specifies instrument calibration file to recall. If hardware calibration is requested and thisoption is empty, IC-CAP will use the current active instrument state. Default = Null

Soft CalSequence


Delay forTimeouts


InitCommand

Command field to set to a mode not supported by the option table. This command is sent atthe end of instrument initialization for each measurement. Normal C escape characters such as\n (new line) are available. Default = none

SystemVariables

None. Software calibration is not provided for the Wiltron360.

Anritsu VectorStar Network Analyzer

The IC-CAP Anritsu VectorStar Network Analyzer driver is supported by Anritsu.Please contact Anritsu at [email protected] for questions about driverdocumentation and support.

mailto:[email protected]

mailto:[email protected]


42

OscilloscopesThe oscilloscopes supported by IC-CAP are:

HP 54120T Series Digitizing Oscilloscopes (measurement)HP 54510 Digitizing Oscilloscope (measurement)Agilent Infiniium Oscilloscope (measurement)HP 54750 Series Digitizing Oscilloscopes (measurement)Differential TDR and TDT Capability (measurement)

HP 54120T Series Digitizing Oscilloscopes

The HP 54120 Series of digitizing oscilloscopes measure time-domain responses, includingTDR (time-domain reflectometry).

HP 54121T measures signals from DC through 20 GHz.HP 54122T (does not have a step generator and cannot perform TDR measurements)provides programmable input attenuation. Bandwidth is reduced to 12.4 GHz due tothe input attenuators.HP 54123T operates up to 34 GHz; it operates up to 20 GHz on channel 1 (thechannel on which the step generator is available).


CHn Channel Unit n (1, 2, 3, and 4)

A Setup configured for measurements using an HP 54120 Series is in the model file54120.demo.mdl. These files also include examples using an HP 54120 Series oscilloscopewith an HP 8130 pulse generator and provides hints for obtaining good alignment betweenmeasured and simulated waveforms when a pulse generator is used.

The following instrument capabilities are supported by IC-CAP:

Time-domain measurements nested within DC bias settings provided by DC SMUs.4-channel concurrent data acquisition.Offset, range, and probe attenuation adjustment for each channel. HP 54122Tincludes options to set internal attenuation for each channel (refer to HP 54120Series Options).Averaging of between 1 and 2048 waveform acquisitions on each channel.Automatic Pulse Parameter Measurements, such as risetime and peak-to-peakvoltage. These are requested in Outputs of Mode T. For help on the available choices,click the middle mouse button over the Pulse Param field and see the Status window.Square-wave generation (except HP 54122T) on CH1, the left-most connector on thetest set. Frequency can be adjusted from 15.3 Hz to 500 kHz. To activate the stepgenerator, the Setup should include an Input with Mode V and Type TDR. In theabsence of a type TDR Input, the step generator is not activated.

The instruments do not support some of the fields present in a TDR Input. For example, itis not possible for the instrument to offer other than a 50-ohm source impedance. The onefield that is of consequence to oscilloscope measurements is Period. IC-CAP directs theinstrument to use the closest value supported.

The other TDR Input fields are ignored during measurement, and the following hardware-imposed values of the instrument's step function apply:

Initial value of 0VPulsed value of 200 mV into 50 ohms; 400 mV into an open-circuitDelay of approximately 17 nsecRisetime of approximately 40 psecPulse width equal to about 50 percent of the specified periodSource impedance of 50 ohmsTime-Domain Reflectometry (except HP 54122T). When a type TDR Input is presentin the Setup, the reflected signal is available on the unit designated CH1.

To make a time-domain measurement, a Setup must have these inputs and outputs:

An Input with Mode T and Type LIN. Here, the values of Start, Stop, and Number ofPoints govern the time axis of the measurement. Start and Stop values define thetime viewing window, and are relative to the trigger event used by the oscilloscope.Optionally, an Input of Mode V and Type TDR or PULSE. The Period field in this Inputcontrols the rate of the oscilloscope's internal square-wave generator. If Period is setto 0, or if this Input is absent from the Setup, the oscilloscope's internal square-wavegenerator is not activated for the measurement. In this case, a trigger signal must beprovided on the oscilloscope's trigger input.

If the Input's Unit field is set to ground, IC-CAP ignores the Input during themeasurement. In this manner, measurements can be performed using a pulsegenerator controlled by its front panel. If the Input's Unit field is set to the pulseunit of a supported pulse generator (for example, PULSE1 for an HP 8130generator), then IC-CAP will control the pulse generator to provide stimulus tothe DUT and oscilloscope.

Refer to the HP 8130 Pulse Generator documentation provided with IC-CAP.

To capture a waveform from any of the instrument's 4 channels requires an Output ofMode V. The Output Editor permits you to specify from which channel a waveform isdesired. Define an Output for each channel of interest.To obtain automatically extracted pulse parameters at any of the 4 channels requiresan Output of Mode T.

The following pulse parameters can be requested: DUTYCYCLE, FALLTIME,FREQ, OVERSHOOT, PERIOD, PRESHOOT, RISETIME, VPP, VRMS, +WIDTH, and-WIDTH. Consult the instrument's Front Panel Operation Reference fordefinitions of these parameters or information on the process by which theinstrument computes them.

By defining multiple Outputs for a scope channel, it is possible to obtain both thefull time-domain waveform and any number of automatically extracted pulseparameters for that channel, all in the same measurement. This can be donewith any or all of the 4 channels within the same measurement.

The following table describes the HP 54120 series options and their default values, whereapplicable.

HP 54120 Series Options


43

Option Description

Hold Time Time, in seconds, prior to performing time-domain measurement. Can be used to permitadditional DC stabilization when a time-domain sweep is nested within DC steps provided by a DCbias unit. Default = 0

Averages Number of averages. Maximum = 2048. Default = 1

CH1 Offset†

DC offset value of Channel 1 in volts. Does not directly affect waveforms returned from theoscilloscope. However, an improper setting can cause the instrument to fail when measuringpulse parameters, such as RISETIME. Set to a value close to the middle of the expected range ofthe output voltage waveform to maximize the instrument's ability to achieve high resolutionwithout experiencing clipping. Valid range is ± 500mV • (CH1 Probe Attn) • (CH1 Internal Attn).Default = 200.0mV

CH1 ProbeAttn †

Set to 10 if the channel 1 probe provides a divide by 10 functionality (20dB). Specifying theattenuation of the probe permits the oscilloscope to generate data in which the probe attenuationis corrected out. Values between 1 and 1000 are accepted. Default = 1.0

CH1InternalAttn †

HP 54122T only. This option causes IC-CAP to control attenuators inside the 54122 test set. Theattenuators have limited power-handling ability †. Measured voltages will take the attenuationsetting into account. Values 1, 3, 10, and 30 are valid. Default = 1.0

CH1 Range†

Set in excess of the maximum anticipated signal swing for this channel. Does not affectwaveforms returned from the oscilloscope. However, an improper setting can cause theinstrument to fail when measuring pulse parameters, such as RISETIME. Specify a Range valuebetween

CH1 Range† (cont'd)

8mV • (CH1 Probe Attn) • (CH1 Internal Attn) and 640mV • (CH1 Probe Attn) • (CH1 InternalAttn). Default = 640.0 mV

Notes^†^Option table entries are also provided for Offset, Probe Attn, Internal Attn, and Range on channelsCH2, CH3, and CH4.† Changing the Probe Attn options for CH1-CH4 and the trigger input does not attenuate the input signals.It only changes the results reported by the instrument. To deliver signals exceeding 2V DC or 16 dBm ACpeak, use an external attenuator.By using the internal attenuators of the HP 54122T (via the Internal Attn options), larger voltages can beaccepted. Limitations on attenuator voltage and power handling are described in the Internal Attendocumentation in the Channels Menu chapter of the HP 54122T Front Panel Reference.The external trigger is ignored if a TDR type Input is defined in the Setup. In the presence of a TDR typeInput, the scope is triggered by its internal TDR step generator.

The TRG options listed in the following table apply when driving the trigger input of theoscilloscope with an external signal. This is typically done with the trigger output from asignal generator.

Trigger Options for the HP 54120T Series

Option Description

TRG ProbeAttn

Set to 10 if the trigger probe is fitted with a 10X (20dB) divider. Values between 1 and 1000 areaccepted. Default = 1.0

TRG Slope Specify triggering on a rising (+) or falling (-) edge. Default = +

TRG Level Voltage threshold at which triggering occurs. Valid range is ±1V • (TRG Probe Attn). Default =100.0mV

NormalizeTDR

If Yes, TDR waveform data from CH1 is subject to the HP 54120 series reflection normalizationprocess. This can substantially improve waveform integrity when cabling and test fixtures haveimpedance mismatches. Prior to using this option perform calibration of the network reflectionpath via the front panel Network page. Default = No (This option is not supported by the HP54122.)

Delay forTimeouts


InitCommand


HP 54510 Digitizing Oscilloscope

The HP 54510 is a 1 giga-sample/second, 2-channel digitizing oscilloscope. The HP 54510driver is an example of a driver created using the Open Measurement Interface. Thedriver's source code can be found in the files user_meas3.hxx and user_meas3.cxx in thedirectory $ICCAP_ROOT/src. For information, refer to Prober (measurement) and MatrixDrivers (measurement).


CHn Channel Unit n (1 and 2)


Time-domain measurements nested within DC bias settings provided by DC SMUs.2-channel concurrent data acquisition.Offset, range, and probe attenuation adjustment for each channel. (Refer to HP54510 Options.)Averaging 1 to 2048 waveform acquisitions on each channel.Automatic Pulse Parameter Measurements, such as risetime and peak-to-peakvoltage. These are requested in Outputs of Mode T. For help on the available choices,click the middle mouse button over the Pulse Param field and see the Status window.

To make a time-domain measurement, a Setup must contain these Inputs and Outputs:

An Input with Mode T and Type LIN. Here, the values of Start, Stop, and Number ofPoints govern the time axis of the measurement. Start and Stop values define thetime viewing window, and are relative to the trigger event used by the oscilloscope.The HP 54510 driver uses the repetitive sampling mode and therefore alwaysmeasures 501 points. The timebase range is set to 500 × step size of the inputsweep. The timebase requires a value in the sequence 1-2-5, that is, 1 nsec, 2 nsec,5 nsec, 10 nsec, ..., 1 sec, 2 sec, or 5 sec. If the Input step size does not correspondto a valid timebase, the driver aborts the measurement and recommends new stopand step values for the input sweep.Optionally, an Input of Mode V and Type PULSE. A trigger signal must be provided onthe oscilloscope's trigger input. If the Input Unit field is set to ground, IC-CAP ignoresthe Input during the measurement. In this manner, you may perform measurementsusing a pulse generator controlled by its front panel. If the Input Unit field is set tothe pulse unit of a supported pulse generator (for example, PULSE1 for an HP 8130generator), then IC-CAP will control the pulse generator to provide stimulus to theDUT and oscilloscope. For more information, refer to HP 8130 Pulse Generator(measurement). Also refer to the documentation for the HP 54120 in the54120.demo.mdl file.To capture a waveform from either of the instrument's 2 channels requires an Outputof Mode V. The Output Editor permits you to specify from which channel a waveformis desired. Define one such Output for each channel of interest.To obtain automatically extracted pulse parameters at either of the 2 channels, theSetup must include an Output of Mode T.

The following pulse parameters can be requested: DUTYCYCLE, FALLTIME,FREQ, OVERSHOOT, PERIOD, PRESHOOT, RISETIME, VPP, VRMS, +WIDTH, and


44

-WIDTH. Consult the instrument's Front Panel Operation Reference fordefinitions of these parameters, or information on the process by which theinstrument computes them.

By defining multiple Outputs for a scope channel, both the full time-domainwaveform and any number of automatically extracted pulse parameters for thatchannel can be obtained, all in the same measurement. This can be done witheither or both of the channels in the same measurement.

The following table describes the HP 54510 options and default values, where applicable.

HP 54510 Options

Option Description

Hold Time Time, in seconds, prior to performing time-domain measurement. Can be used to permitadditional DC stabilization when a time-domain sweep is nested within DC steps provided by a DCbias unit. Default = 0.0

Averages Number of averages. Maximum = 2048. The HP 54510 rounds the number of averages to thenearest power of 2. If the value is exactly halfway between, it takes the higher value. Default = 1

CH1 Offset†

DC offset value of Channel 1, in volts. This does not directly affect waveforms returned from theoscilloscope. However, an improper setting can cause the instrument to fail when measuring pulseparameters, such as RISETIME. Set this to a value close to the middle of the expected range ofthe output voltage waveform; this will maximize the instrument's ability to achieve highresolution without experiencing clipping. Valid range is ±250V • (CH1 Probe Attn). Default = 0.0

CH1 ProbeAttn †, †

Set to 10 if the Channel 1 probe provides a divide by 10 functionality (20 dB) and 50 ohm inputimpedance is selected. Specifying the attenuation of the probe permits the oscilloscope togenerate data in which the probe attenuation is corrected out. Values between 0.9 and 1000 areaccepted. Default = 1.0

CH1Range †

Set in excess of the maximum anticipated signal swing for this channel. This option does notaffect waveforms returned from the oscilloscope. However, an improper setting can cause theinstrument to fail when measuring pulse parameters, such as RISETIME. Default = 2.0

Notes† Option table entries are also provided for Offset, Probe Attn, and Range for CH2.† Changing Probe Attn options for CH1, CH2 and the External Trigger input does not attenuate the inputsignals. It only changes the results reported by the instrument. To deliver signals exceeding 5V rms (50ohm) or 250V (1 Mohm), an external attenuator should be used.

Refer to the following table for oscilloscope trigger options. The TRG/TRIG options apply tothe trigger input. This is typically done with the trigger output from a signal generator.When using the EXT TRIG channel, be sure the TRG Source option is set to "E" (ExternalTrigger).

NoteInstrument settings not included in the Instrument Options folder, such as input impedance, can be setmanually before executing Measure.

Oscilloscope Trigger Options for the HP 54510

Option Description

EXT TRIGAttn

Attenuation of the EXT TRIG channel. Set to 10 if the trigger probe is fitted with a 10X (20dB)divider and the EXT TRIG channel is set to 50 ohms. Values between 0.9 and 1000 areaccepted. Default = 1.0

TRG Source Specify the trigger source channel: 1 (CH1), 2 (CH2) or E (External Trigger). Default = E

TRG Slope Specify + (rising edge) or - (falling edge). Default = +

TRG Level Voltage threshold at which triggering occurs. Valid range is ±2V • (TRG Probe Attn) for the EXTTRIG channel and ±1.5 • (full scale from center of screen) for channels CH1 and CH2. Default =0.0

Delay forTimeouts

For long-running measurements (that use a high number of averages, for example), use thisoption to avoid measurement timeouts. Default = 0.0

InitCommand

Use to set the instrument to a mode not supported by the option table. This command is sent atthe end of instrument initialization for each measurement. Normal C escape characters such as\n (new line) are available. Default = none

Agilent Infiniium Oscilloscope

The Agilent Infiniium scopes are available as 2 or 4-channel digitizing oscilloscopes. IC-CAP supports the following Infiniium scopes:

54810A, 54815A, 54820A, 54825A 500 MHz bandwidth, 1 GSa/s sample rate and32K of memory width.54835A, 1 GHz bandwidth, 4 GSa/s sample rate, 62K memory width.54845A, 1.5 GHz bandwidth, 8 GSa/s sample rate, 64K memory width.

The IC-CAP driver supports acquisition only from Channels 1 and 2.


CHn Channel Unit n (1 and 2)


Time-domain measurements nested within DC bias settings provided by DC SMUs.2-channel concurrent data acquisition (Channel 1 and 2 only).Offset, range, and probe attenuation adjustment for each channel. (Refer to InfiniiumOptions.)Averaging 1 to 2048 waveform acquisitions on each channel.Automatic Pulse Parameter Measurements, such as risetime and peak-to-peakvoltage. These are requested in Outputs of Mode T. For help on the available choices,click the middle mouse button over the Pulse Param field and see the Status window.

To make a time-domain measurement, a Setup must contain these Inputs and Outputs:

An Input with Mode T and Type LIN. Here, the values of Start, Stop, and Number ofPoints govern the time axis of the measurement. Start and Stop values define thetime viewing window, and are relative to the trigger event used by the oscilloscope.The Infiniium acquisition range is given by the number of acquisition points multipliedby the sampling period (1/Acquisition Rate). Acquisition points and frequency are setin the instrument option table. If the time viewing window set by the Start and Stopvalues is wider than the acquisition range, the driver aborts the measurement. TheAcquisition rate must be in the 1, 2.5, 5, 10 sequence, that is, 1MSa/s, 2.5 MSa/s, 5MSa/s, etc. The maximum acquisition rate depends on the scope model. Theacquisition mode may be Real or Equivalent Time. Real time mode usually is used forsingle events, such as transients, while equivalent time may be used for periodicsignals.Optionally, an Input of Mode V and Type PULSE. A trigger signal must be provided onthe oscilloscope's trigger input. If the Input Unit field is set to ground, IC-CAP ignoresthe Input during the measurement. In this manner, you may perform measurementsusing a pulse generator controlled by its front panel. If the Input Unit field is set to


45

the pulse unit of a supported pulse generator (for example, PULSE1 for an HP 8130generator), then IC-CAP will control the pulse generator to provide stimulus to theDUT and oscilloscope. For more information, refer to HP 8130 Pulse Generator(measurement).To capture a waveform from either of the instrument's 2 channels requires an Outputof Mode V. The Output Editor permits you to specify from which channel a waveformis desired. Define one such Output for each channel of interest. When acquisitionrange and points differ from sweep time interval and points, the waveform is actuallyinterpolated by the actual measured data. It is a good practice to use an acquisitionrange that is slightly greater than the time window, but not too much greater.To obtain automatically extracted pulse parameters at either of the 2 channels, theSetup must include an Output of Mode T.

The following pulse parameters can be requested: DUTYCYCLE, FALLTIME, FREQ,OVERSHOOT, PERIOD, PRESHOOT, RISETIME, VPP, VRMS, +WIDTH, and -WIDTH. Consultthe instrument's Front Panel Operation Reference for definitions of these parameters, orinformation on the process by which the instrument computes them.

By defining multiple Outputs for a scope channel, both the full time-domain waveform andany number of automatically extracted pulse parameters for that channel can be obtained,all in the same measurement. This can be done with either or both of the channels in thesame measurement.

As shown in the following instrument options table, the trigger source may be set toChannel 1 or 2, or to EXT or AUX. The trigger sweep may be Auto, Triggered or Single.Trigger level and slope are also specified.

The following table describes the Infiniium options and default values, where applicable.

Infiniium Options

Option Description

Hold Time Time, in seconds, prior to performing time-domain measurement. Can be used to permitadditional DC stabilization when a time-domain sweep is nested within DC steps provided by aDC bias unit. Default = 0.0

Sample Rate The internal sample frequency. It must be in the 1, 2.5, 5, 10 sequence. In real-time mode themaximum sample rate is 1 GSa for the 54810A/15A, 2GSa for the 54820A/25A, 4 GSa for the54835A and 8 GSa for the 54845A (2 channel mode). Default = 1 GSa.

AcquisitionMode

Can be real time (R) or equivalent time (E). Real time is used for single events such astransients while equivalent time may be used to increase the "equivalent" sampling rate whenthe waveform is periodical. Default = R.

AcquisitionCount

Turns averaging on or off, and (when on) sets the number of averages. Allowed range is 1through 4096. Use 1 to turn averaging off. Use 2 through 4096 to turn on averaging and setthe count. Default = 1.

AcquisitionPoints

Number of acquired points at the sample rate. The acquisition range is defined as theacquisition period 1/(Sample rate) multiplied by the number of points. The number of points islimited by the memory depth: 32,768 points for the 54810A/15A/20A/25A and 65,536 pointsfor the 54835A/45A.

CH1 Scale †

[V/div.]DC vertical sensitivity in Volts per division. When probe attenuation is 1 maximum sensitivity is5 V/div. Minimum sensitivity is 1 mV/div for 54810A/15A/20A/25A and 2 mV/div for 54835Aand 54845A. Default = 500 mV/div.

CH1 Offset †

[V]DC available offset. It depends on the scale. Maximum offset is ±250V when CHn Scale = 5V/div.

CH1 Input † Channel input impedance: DC 50 ohm (DC50),1 Mohm (DC), AC. LFR1 and LFR2 are alsopossible when using the Agilent 1153A differential probe. Default is DC.

CH1 ProbeAttn † †

Set to 10 if the Channel 1 probe provides a divide-by-10 functionality (20 dB) and 50 ohm inputimpedance is selected. Specifying the attenuation of the probe permits the oscilloscope togenerate data in which the probe attenuation is corrected out. Values between 0.9 and 1000are accepted. Default = 10.0.

Trigger Input Set trigger input source (1, 2, AUX or EXT). Default is 1 for Channel 1.

TriggerSweep

Set trigger sweep Modes to Auto (A), Triggered (T), or Single (S). Default is Auto (A).

TriggerSlope

The only supported trigger mode is Edge. Trigger slope may be positive () or negative (-).Default is positive ().

Trigger Level[V]

Sets voltage level at which trigger occurs. Level range depends on sweep mode and scope type.Default is 500 mV.

Delay forTimeouts

For long-running measurements, such as collecting a high number of averages, use this optionto avoid measurement timeouts. Default = 0.0

InitCommand

Sets the instrument to a mode not supported by the option table. This command is sent at theend of instrument initialization for each measurement. Normal C escape characters such as \n(new line) are available. Default = none

Notes† ^ Option table entries are also provided for Scale, Offset, Input and Probe Attn, for CH2. ^† Changing Probe Attnoptions for CH1, CH2 and the External Trigger input does not attenuate the input signals. It only changesthe results reported by the instrument. To deliver signals exceeding 5V rms (50 ohm) or 250V (1 Mohm),an external attenuator should be used.

HP 54750 Series Digitizing Oscilloscopes

The IC-CAP driver for the HP 54750 supports the following plug-in modules:

HP 54753A. This module is a 2-channel vertical plug-in with a TDR step generatorbuilt into channel one. The bandwidth of the TDR/vertical channel is 18 GHz. Thebandwidth of channel 2 is 20 GHz.HP/Agilent 54754A. This module has 2 independent vertical channels and 2independent step generators. The bandwidth of both channels is 18 GHz.HP 54752A and HP 54752B. The 54752A has two 50 GHz bandwidth channels and54752B provides a single cost-effective channel.HP 54751A. This module has two 20 GHz bandwidth channels.

Since the instrument is configurable, the insertion of the instrument in the activeinstrument table must be done using rebuild active list. Plug-in modules must be placedstarting from slot 1 without discontinuities. IC-CAP assigns the following names to theunits:

TDRn for TDR channelsCHn for normal scope acquisition channels

Example files: A Setup configured for measurements using the HP 54750, is in the modelfile /examples/model_files/misc/hp54750.mdl.


Time domain acquisition for each channel (TDR or CH).Offset, Scale, and Probe Attenuation adjustment for each channel.Averaging of between 1 and 4096 waveform acquisition.Automatic Pulsed/Waveform parameter measurements for each TDR or CH type


46

channel.Trigger Probe Attenuation, Slope, Level, Mode as well as the trigger slot (2 or 4 incase 2 plug-ins are present) can be set in the instrument table.Start time, Stop time, and number of points are set in the Input Time sweep.Step generator on TDR channels. Frequency rate can be adjusted between 50 and250 kHz. To activate the step generator the setup should include an Input with modeV and Type TDR. Note that only one TDR step generator can be active per setup(differential TDR is not supported). The period on the TDR input is used to calculatefrequency for TDR/TDT measurements.TDR normalized measurements are supported for each of the TDR channels. Toacquire a normalized TDR response, perform either software or hardware calibration,then set Normalize to Y in the TDR channel and measure. To perform software TDRcalibration, first set the normalization option to TDR, then run Calibrate and followthe steps.TDT normalized measurements are supported for each plug-in. Plug-in channel 1must be the TDR source, and channel 2 must be the TDT sink. To acquire anormalized TDT measurement on channel 2, perform either software or hardwareTDT calibration, then set Normalize to Y on channel 2 (sink) and measure. Toperform software TDT, set the normalization option to TDT, run Calibrate, and followthe steps.

You must be sure to insert the oscilloscope into IC-CAP's instrument table. Connect theinstrument, switch it on, and perform Rebuild in the Hardware Setup. The HP 54750should be now present in the Instrument List. Select HP 54750 in the list and selectConfigure. The units should reflect the hardware configuration and the plug-in type in theUnit Table. Here are examples of what should appear in the table:

If module HP/Agilent 54754 occupies slots 1 and 2, TDR1 and TDR2 units shouldappear in the Unit Table.If module HP 54753 occupies slots 1 and 2, TDR1 and CH2 units should appear in theUnit Table.

To make time-domain measurements (acquisition only), a setup must contain theseInputs and Outputs:

An Input with Mode T, and Type LIN. Minimum start time is 20 nsec; max start timeis 10 sec. The minimum time range is 100 psec while the maximum range is 10 sec.During acquisition (no internal TDR) the number of acquired points can be set to anynumber between 16 and 4096.A trigger signal must be provided at the trigger input (slot 2 or 4) to acquire anywaveform. Trigger Mode can be selected in the instrument option. Use trigger modeFREErun or TRIGgered for periodic waveforms. Use option TRIGgered when usingexternal trigger for example for acquiring transients or when using external TDR stepgenerator.To capture a waveform, an Output of Mode V is required. Define an output for eachchannel of interest.To obtain automatically-extracted pulse parameters, the setup must include anoutput of mode T that specifies the unit and the requested parameter. Examples ofparameter values are VPP and VRMS.

To make TDR or TDT measurements, a setup must contain these Inputs and Outputs:

An Input with Mode T and Type LIN. Minimum start time is 20 nsec, max start time is10 sec. The minimum time range is 100 psec while the maximum range is 10 sec.When the instrument's Normalize option is turned OFF, the number of acquired pointscan be set to any number between 16 and 4096. When the Normalize option is ON,the number of points can be set to any number between 16 and 4096 that is amultiple of 2, such as 512 or 1024.An Input with Mode V and Sweep Type TDR. The unit is set to the TDR sourcechannel. Only the value of the Period is used during measurement for setting thefrequency of the internal step generator. Use a value between 50 Hz (20 msec) and250 kHz (4 usec). The other fields, such as Delay and Width, are used only by thesimulator. If an external TDR step generator is used, then Unit must be set to GND,and all parameters (including Period) are used only by the simulator.To capture the output waveforms, insert 1 or 2 Outputs of mode V referring to theTDR source channel for TDR measurements, or to the sink channel for TDTmeasurements.(Optional). To measure waveform parameters such as VPP and RISETIME, insert 1 ormore Outputs of mode T.

TDR or TDT measurements can be done with or without normalization. Normalizationestablishes a reference plane different from the oscilloscope output. The reflection andohm measurements are based on the actual measured step height. Also, from thisinformation, the scope builds a filter, which can be applied to any reflected signal. Therisetime of the filtered step can be selected. The filtered step removes any losses ordiscontinuities from the reference plane generated by the plug-in.

To measure without normalization, simply set the Normalize flag to N in the instrumentoptions for any channel involved in the TDR or TDT measurement.

To make normalized TDR measurements, either hardware or software normalization mustbe performed prior to measurement. To perform software calibration, set theNormalization mode to TDR in the Instrument Option Table. Then run Calibrate. Thisroutine will load current sweeps (start, stop and period) in the instruments and then willask the operator to insert the calibration standards (short and load) at the referenceplane.

Once the instrument has been successfully calibrated, set the Normalize flag of the TDRsource channel to 'Y' before running a measurement to acquire normalized data. Set thenormalized response Unit to VOLT (default), REF or OHM in the Instrument Options Table.When setting response scale to VOLT, IC-CAP will acquire the actual normalized response.When the response scale is OHM, IC-CAP will acquire the normalized-to-50 ohm response.This is particularly useful when evaluating characteristic impedance of different line series.Setting the scale to REF will acquire the reflection due to a change of impedance. Thenormalized rise time can also be set in the instrument option table. The minimum settablerise time actually depends on the number of points. Generally speaking, increasing thenumber of points allows a smaller rise time and therefore improves the space resolution(minimum distance between 2 discontinuities to distinguish them in the space/timedomain).

To make normalized TDT measurements, either hardware or software normalization mustbe performed prior to measurement. To perform software calibration, set Normalizationmode to TDT in the Instrument Option Table. Then connect source and sink together(without DUT) and run Calibrate.

Once the instrument has been successfully calibrated, set the Normalize flag of the TDT


47

sink channel to 'Y' before running a measurement to acquire normalized data. NormalizedResponse unit can be set to VOLT (default) or GAIN. The normalize risetime can also bevaried with the same limitation described above. Differential TDR/TDT Capability

New addition to TDR driver: Differential TDR/TDT capabilities.

Two new entries have been added to the Agilent 54750 Instrument table:

Differential Mode

Set the instrument in differential mode.

Channel 1 and 2 are the TDR channels.

The differential stimulus on channel 1 and 2 can be Differential (DIFF) or Common(COMM).

Default is no differential stimulus (NONE).

Once the instrument has been calibrated in differential Response mode, the responsereading can be set to Differential (DIFF) Mode or Common (COMM).

Note that this field is active only when the Normalize Flag of the response channels is setto yes.

Default is DIFF.

To make TDR differential measurements, place the Agilent 54754A plug-in in the first 2instrument slots (channel 1 and 2).

In the IC-CAP measurement page insert 1 input of type TDR (Unit TDR1 or CH1).

Insert 1 input of Mode T (Type LIN) and set the time interval and the number of points.

Insert 2 outputs of Mode V monitoring channel 1 and 2.

In the 54750 Instrument Option Table, set the Differential Mode to DIFF or COMM.

To measure raw data simply set the Normalize flags of CH1 and CH2 to N and run themeasurements.

To measure normalized data, perform the TDR normalization before running themeasurements. Follow the instructions in the 54754 manual to calibrate in differential TDRmode.

Once the instrument has been successfully calibrated, set the Normalize Mode to TDR, setDifferential Response Mode to DIFF or COMM. To measure the normalized response simplyset the Normalize flag of channel 1 and 2 to yes.

Summary differential TDR

Differential Mode Differential Response Mode Response Mode CH1 CH2

Raw DIFF/COMM Not Relevant Nor relevant N N

Norm DIFF/COMM DIFF/COMM TDR Y Y

To make TDT differential measurements place 1 Agilent 54754A plug-in in the first 2instrument slots (channel 1 and 2) and second 54754 plug-in in the third and fourth slots.When measuring differential TDT, the driver assumes that Channel 1 and 2 supply thedifferential stimulus (input).

In the IC-CAP measurement setup page insert 1 input of type TDR (Unit TDR1 or CH1).

Insert 1 input of Mode T (Type LIN) and set the time interval and the number of points.

Insert 4 outputs of Mode V monitoring channel 1 to 4. In the 54750 Instrument OptionTable, set the Differential Mode to DIFF or COMM.

To measure raw data simply set the Normalize Flags of CH1,CH2,CH3 and CH4 to N andrun the measurements.

To measure normalized data, the user needs to perform the TDT normalization beforerunning the measurements.

Follow the instructions in the 54754 manual on how to calibrate in differential TDT mode.

Once the instrument has been successfully calibrated, set the Normalize Mode to TDT, setDifferential Response Mode to DIFF or COMM.

To measure the normalized response simply set the normalized flag of channels 3 and 4 toyes.

Summary differential TDT:

Differential Mode Differential Response Mode Response Mode CH1 CH2 CH3 CH4

Raw DIFF/COMM Not Relevant Not relevant N N N N

Norm DIFF/COMM DIFF/COMM TDT N N Y Y

The following table describes the HP 54750 options and default values, where applicable.

HP 54750 Options Table


48

Option Description

Hold Time Time, in seconds, prior to performing time-domain measurements. Default = 0

Averages Number of averages per sample. Min = 1 (off), Max = 4096. Default = 16

NormalizationMode

Two modes supported for calibration and measurements: TDR or TDT. Default = TDR

NormalizedResponse Unit

Sets the type of unit for the acquired normalized response. Possible choices are VOLT, REFor OHM for TDR type measurements and VOLT or GAIN for TDT measurements. Default =VOLT

NormalizedResponseRisetime

Set the risetime for the normalized response. Minimum risetime depends on number ofpoints. In case specified rise time is greater than the minimum allowed for that number ofpoints, IC-CAP will set the minimum possible value. Default = 40 psec

CHn ProbeAttenuation

Probe Input impedance is always 50 ohm. Specifying the attenuation of the probe permitsthe oscilloscope to generate data in which the probe attenuation is corrected out. Forexample, set it to 10 if the channel 1 probe provides a divide by 10 functionality. Valuesbetween 0.9 and 1000 are accepted. Default = 1.0

CHn Offset DC offset value of Channel 1, in volts. This does not directly affect waveforms returned fromthe oscilloscope. However, an improper setting can cause the instrument to fail whenmeasuring pulse parameters, such as RISETIME. Set this to a value close to the middle ofthe expected range of the output voltage waveform; this will maximize the instrument'sability to achieve high resolution without experiencing clipping. Valid range is ±250V.Default = 200.0 mV

CHn Scale Default = 100.0 mV/div.

CHn Normalize Normalization Flag. When set to 'Y', IC-CAP acquires the normalized response with unit asspecified in The Normalized Response Unit. Default = N

TRG ProbeAttenuation

Default = 1.0

TRG Slope Specifies triggering on a rising (+) or falling (-) edge. Default = +

TRG Level Voltage threshold at which triggering occurs. Range depends on attenuation. Default = 0.0mV

TRG Slot Choose the input trigger channel. For example, when 54754 plug-in is present on slot 1 and2, trigger will be on slot 2. When another TDR plug-in is present on slot 3 and 4, slot 4 isanother possible choice for trigger. Default = 2

TRG Mode Used when acquiring a waveform not in TDR mode (internal trigger is used in that case).Possible choices are freerun (FREE) usually used for periodic waveform or triggered (TRIG)for transients. Default = FREE

Delay fortimeout

Increase this delay when acquiring a large number of points or averages. This gives moretime for the instrument to digitize the waveform and save it into memory. Default = 3

Init Command Use to set the instrument to a mode not supported by the option table. This command issent at the end of instrument initialization for each measurement. Normal C escapecharacters such as \n (new line) are available. Default = None


49

Pulse GeneratorsThis section describes the HP 8130 and the HP 8131 pulse generators.

HP 8130 Pulse Generator (measurement)HP 8131 Pulse Generator (measurement)

HP 8130 Pulse Generator

The HP 8130 is a programmable pulse generator controllable by IC-CAP. It providesexcellent features for time-domain characterization using pulse stimuli. The followingpulse characteristics are programmable:

Period, Width, and DelayRisetime and FalltimeInitial and Pulsed Voltage Levels

IC-CAP assigns the following name to the channel 1 output unit:

PULSE1

The HP 8130 offers a fixed source impedance of 50 ohms. Pulse period can be varied from3 nsec to 99.9 msec. Rise and falltimes can be varied from 670 psec to 100 μsec. Theoutput voltage range is from -5.2 to +5.2V, but the maximum voltage swing must be lessthan or equal to 5.2V. A complementary output signal is available (refer to the followingtable).

A Setup configured for measurements using the HP 8130, along with HP 54120 Seriesdigitizing oscilloscopes is in the model file 54120.demo.mdl. These files also includeexamples using an HP 54120 Series oscilloscope with no pulse generator, or with amanually controlled pulse generator.


HP 8130 Options

Option Description

Width atTop

Flag provided to aid simulator compatibility. The HP 8130 defines pulse width to include the topsection of the pulse plus one-half of the rising and falling edges. SPICE defines pulse width toinclude the top of the pulse only. For compatibility with SPICE, set this option to Yes (the 8130pulse will become wider). Default = No

EnableComp Out

If Yes, complementary data can be obtained by cabling to the complementary output connectoron the HP 8130. Default = No

Pulse DelayOffset

The HP 8130 has a delay between its trigger output and signal output (SPICE has nothing likethis). The value is added to the TDR or PULSE sweep Delay value. Positive values will shift thewaveform to the right; negative values will shift the waveform to the left. This option permitsone to align the simulated and measured waveforms. The option may need adjustment if theperiod is changed. Default = 0

InitCommand


HP 8131 Pulse Generator

The HP 8131 is a programmable pulse generator controllable by IC-CAP. It providesexcellent features for time-domain characterization using pulse stimuli. The followingpulse characteristics are programmable:

Period, Width, and DelayInitial and Pulsed Voltage Levels

IC-CAP assigns the following name to the channel 1 output unit:

PULSE1

The HP 8131 offers a fixed source impedance of 50 ohms. Pulse period can be varied from2 nsec to 99.9 msec. Rise and fall times are fixed <200 psec; if >200 psec, IC-CAP willissue a warning PULSE1 Rise/Fall time fixed at less than 200ps.

The output voltage range is from -5.0 to +5.0V; the maximum voltage swing must be lessthan or equal to 5.0V. The offset voltage is from -4.95V to +4.95V. A complementaryoutput signal is available (refer to the following table).

A Setup configured for measurements using the HP 8131, along with HP 54120 Seriesdigitizing oscilloscopes is in the model file 54120.demo.mdl. These files also includeexamples using a HP 54120 Series oscilloscope without a pulse generator, or with amanually controlled pulse generator.


Options for the HP 8131

Option Description

Width atTop

Flag provided to aid compatibility with SPICE and other simulators. The HP 8131 defines pulsewidth to include the top section of the pulse plus one-half of the rising and falling edges. SPICEdefines pulse width to include only the top of the pulse; as a result, SPICE pulses are wider. ForSPICE compatibility, set this option to Yes. Default = No

EnableComp Out

If Yes, complementary data can be obtained by cabling to the complementary output connectoron the HP 8131. Default = No

Pulse DelayOffset

The HP 8131 has a delay between its trigger output and signal output (SPICE does not). Thevalue is added to the TDR or PULSE sweep Delay value. Positive values will shift the waveform tothe right; negative values will shift the waveform to the left. This option permits alignment ofsimulated and measured waveforms. The option may need adjustment if the period is changed.Default = 0

InitCommand



50

Dynamic Signal AnalyzersIC-CAP supports the HP/Agilent 35670A dynamic signal analyzer.

HP/Agilent 35670A Dynamic Signal Analyzer

The HP/Agilent 35670A portable 2- or 4-channel dynamic signal analyzer evaluates signalsand devices under 102.4 kHz real-time rate at 800 lines of resolution. It providesspectrum, network, and time- and amplitude-domain measurements from virtually DC toslightly over 100 kHz.


CHn Channel Unit (1 and2)

SRC Source Unit

The following table describes the HP/Agilent 35670A options and their default values,where applicable.

Table: HP/Agilent 35670A Options

Option Description

Hold Time Time delay, in seconds, before each primary sweep begins.

Delay Time Time delay, in seconds, before each sweep point is measured.

Averages Defines the averaging of the instrument. Maximum is 9,999,999.

SourceMode

Source waveforms: (R) random noise, (B) burst random, (P) periodic chirp, or (S) fixed sine.

DC Offset Specifies a DC offset for the source output.

Source Freq Sets the frequency of the sine source.

WindowType

Type of windowing function: (H) Hanning, (U) uniform, (F) flat or (E) exponential.

CHn Units Vertical unit for the specified display's Y axis: (V) volts, (V2) square volts, (V/RTHZ) square rootpower spectral density, or (V2/HZ) power spectral density.

InitCommand

Extra command to initialize the instrument to a certain mode.


51

DriversAdding an Instrument Driver (measurement)Prober Drivers in IC-CAP (measurement)Matrix Drivers in IC-CAP (measurement)Driver Examples (measurement)Handling Signals and Exceptions in Prober and Matrix Drivers (measurement)


52

Adding an Instrument DriverMany instruments can measure a device or a circuit. While IC-CAP supports major HP orAgilent instruments, other instruments manufactured by HP or Agilent, or other vendorscould be used for characterization work within IC-CAP. The Open Measurement Interface(OMI) is a part of IC-CAPs open system philosophy that allows the addition of newinstrument drivers.

NoteAs creating new drivers requires C++, you must obtain C++ software that will compile with both youroperating system and with IC-CAP. To determine appropriate software media options and obtain the mostup-to-date part numbers, consult an appropriate pricing and configuration guide, or contact your salesrepresentative.

This section provides information on OMI and the basic form of an OMI driver. Alternativesto creating a new driver are also addressed.

Using the Open Measurement Interface

The Open Measurement Interface enables you to add drivers for other instruments. User-added drivers can be full-featured, fully integrated, and indistinguishable from the Agilent-provided drivers. Like the Agilent-provided drivers, they are written using C++. OMI wasdesigned to ensure that C Language programmers do not experience language barrierswhen creating new drivers.

Much of the work necessary to lay out the required code is performed by a tool kitcomprised of Driver Generation Scripts described in Adding a Driver. These scripts alsowrite all necessary code for the Instrument Options editors for a new driver, and allnecessary code for the driver to be included in the Instruments Library shown in theHardware Setup window.

The user is responsible for filling in the bodies of a set of functions that IC-CAP callsduring measurements. A set of reusable software constructs is provided for accomplishingcommon programming tasks; refer to Programming with C++.

With the first version of the Open Measurement Interface (IC-CAP version 4.00), onlyGPIB based instrument I/O is formally supported.

OMI Guidelines

To use the Open Measurement Interface, the following qualifications are recommended.

One year of C programming experience or recent completion of a good course in C.Familiarity with the use of struct data types in C (or record data types in PASCAL) isessential, because C++ classes build upon the struct concept.Experience writing code to control an instrument.Familiarity with the particular instrument's features and operation.A willingness to learn the details of the requests IC-CAP places on drivers, and theorder in which they occur during principal operations: Measure, Calibrate, andRebuild (instrument list).A copy of the C++ language system provided by your computer vendor, includingmanuals and a license.

Driver Development Concepts

The basic form of user-added drivers involves 1 file with declarations of data types andfunctions, and 1 file with implementations of functions. Because Driver Generation Scriptsare provided, very few modifications to the declarations file are necessary; work is largelyconfined to the function implementations file. The separation of declarations andimplementations is common practice, and has been used with User C Functions. Thesource directory $ICCAP_ROOT/src is used for OMI compilation, just as it is for User CFunctions.

The default source files for new drivers already contain example drivers:

HP 4194: user_meas.hxx and user_meas.cxxHP 4140: user_meas2.hxx and user_meas2.cxxHP 54510: user_meas3.hxx and user_meas3.cxx

Unless you choose to add files to optimize your compilation process, the$ICCAP_ROOT/src/Makefile permits the make(1) command to create an up-to-date IC-CAP executable file with your latest modifications. This Makefile accounts for the distinctcompilation needs of the C++ and C source files by invoking the appropriate compiler. Bydefault, make(1) understands a .cxx suffix to mean C++ compilation, and .c to mean Ccompilation; the Open Measurement Interface follows this convention.

The process for building the shared libraries libicuserc.<ext> and libicusercxx.<ext> isdemonstrated in the following figure. It is not necessary to know the details; the make(1)command can perform the entire process (provided the $ICCAP_ROOT/src/Makefile iscorrect).

The user driver files, user_meas.hxx and user_meas.cxx, to which your driver is added bydefault, already contain an example driver. This keeps the facility simple but could slowyour compilation. If you choose to add your code to other files, adjust the Makefile.Otherwise, do not modify the Makefile.

Flow Diagram for the User Build Process


53

NoteThe pbench.o file is supplied since it is required to build the shared library. However, the source is notprovided, so you cannot modify it. Additional information is available online in example drivers, headerfiles, and comments inside the code generated by the driver generation scripts.

Example Drivers Three example drivers HP 4194, HP 4140, and HP 54510 can be seenin the Instrument Library in the Hardware Setup window.

Source files for the HP 4194 are user_meas.hxx and user_meas.cxxSource files for the HP 4140 are user_meas2.hxx and user_meas2.cxxSource files for the HP 54510 are user_meas3.hxx and user_meas3.cxx

The information provided by these example drivers should serve as valuable referencematerial for adding a new driver.

Header Files Files that are normally modified and re-compiled, user_meas.hxx anduser_meas.cxx, use include (or header) files. The most important header files are unit.hxx, user_unit.hxx, instr.hxx, and user_instr.hxx. These files declare all of the virtualfunctions for each driver, and provide information to write (or avoid writing) eachfunction.

Generated Code and Comments The driver generation scripts generate both code andcomments. Generally, the comments state what each required function must return, whenit is invoked, and its purpose. Code examples are often provided that you can use as thebasis for the code you must provide. To access this information, run the scripts. Forinformation, refer to Driver Generation Scripts.

Binary Byte Order

For information on transferring binary data between an instrument and IC-CAP, see theREADME.byteorder file in the source directory $ICCAP_ROOT/src. It contains importantinformation with respect to the order of bytes in a multi-byte number.

Adding a Driver

The basic steps for adding a driver include:

Run the Driver Generation Scripts.1.Fill in functions that control your instrument.2.Inform IC-CAP of the new instrument type.3.Build the IC-CAP executable file.4.Debug the new driver.5.

Details for adding a driver are provided in the following paragraphs.

Driver Generation Scripts

The driver generation scripts provide a framework of functions in which a users drivercode is placed.

mk_unit

The mk_unit script generates code for units in an instrument. For example, in HP 4141,there are 8 units, including 4 DC SMUs, 2 VS and 2 VM units. The HP 4194 example hasjust 1 unit, which is typical for a CV driver.

A transcript of the mk_unit session used for the HP 4194 driver is:

$ mk_unit

Enter a name for the unit class for which you want code:

cvu_4194

Enter a name for the instrument class that will use this unit

class: hp4194

Enter the full name of the .hxx file that will declare hp4194

default: user_meas.hxx]:

Enter a name of twelve characters or less; the emitted code

will be appended to .cxx and .hxx files with this basename

[default: user_meas]:

Done. C++ code was added to user_meas.hxx and user_meas.cxx.

You should re-run mk_unit if more unit types are needed.

Otherwise, you probably need to run mk_instr now.

You must supply the name of 2 C++ classes. A class is a name for a user-defined C++type and is like a struct in C. The mk_unit script uses the chosen class names throughoutthe generated code. In this example, a unit class name (cvu_4194) is chosen to denoteCV Unit in a 4194, and an instrument class name (hp4194) to reflect the name of theinstrument. Try to select class names in the same style. Class names should bemeaningful and specific, since this helps to avoid name collisions during compilation. Forexample, a suffix relating to the instrument or company. C and C++ compilers generallyaccept very long names. The use of long descriptive names helps prevent compilation orlinking problems due to name collisions.


54

If the instrument has more than 1 kind of unit to drive, for example, HP 4141, runmk_unit repeatedly. If it has several identical units, do not re-run mk_unit. Identical unitscan be taken into account after running mk_instr.

mk_instr

The mk_instr script generates code for instrument-wide functionality in a driver, such ascalibration, self-test, and getting the instrument recognized during Rebuild (instrumentlist).

A transcript of the mk_instr session used for the HP 4194 driver is:

$ mk_instr

Enter the name of the instrument class for which you want

code: hp4194




Done. C++ code was added to user_meas.hxx and user_meas.cxx.

Now you can go take a look at user_meas.cxx, and start doing

the real work.

NOTE: in user_meas.cxx you may eventually need to add

#include statements to ensure that user_meas.cxx sees the

class declarations of any unit classes used by hp4194.

Disregard this if the necessary unit declarations appear at

the beginning of user_meas.hxx.(The mk_unit script should

generally have put them there.)

You WILL need to declare some units in the class declaration

of hp4194 in user_meas.hxx (see comments therein).

After running this script, you generally need to run

mk_instr_ui next.

This script requires the class name hp4194 to be repeated again, exactly as it is entered inmk_unit. (In your own driver, use another class name besides hp4194, but repeat thesame instrument class name when each script requires.)

The script mentions the need to declare some units, which is accomplished by manualedits to the user_meas.hxx file; for example, cvu_4194* cv_unit; accomplishes that forthe HP 4194 driver in the user_meas.hxx file. If HP 4194 had 2 identical CV unitsavailable, this declaration is:

cvu_4194* cv_unit_1;

cvu_4194* cv_unit_2;

mk_instr_ui

The mk_instr_ui script generates code that fully implements the Instrument Options(measurement) tables appearing in Setups that use the instrument driver. Within thesetables, an IC-CAP operator can specify Delay Time, Integration Time, and otherinstrument-specific options. Since this script completely writes out the necessary C++code for this user-interface functionality, it asks more queries than the previous scripts.

A transcript of the mk_instr_ui session used for the HP 4194 driver is:

$ mk_instr_ui

NOTE: valid types for editor fields are these:

{ real \| int \| char \| boolean \| string }

Enter the name of the instrument class for which you want UI

code: hp4194




Enter the label for an editor field (or enter a null string

if no more fields are desired): Use User Sweep

Enter a type for editor field 'Use User Sweep' [h for help] :

boolean

Enter an initial value for this field [ 0 or 1 ] : 0


if no more fields are desired): Hold Time

Enter a type for editor field "Hold Time" [h for help] : real

Enter the minimum legal value for this field: 0

Enter the maximum legal value for this field: HUGE

Enter a granularity value (for rounding this field; 0 for no

rounding): 0

Enter an initial value for this field: 0


if no more fields are desired): Delay Time

Enter a type for editor field "Delay Time" [h for help] :

real


Enter the maximum legal value for this field: 3600


rounding): 0



if no more fields are desired): Meas Freq

Enter a type for editor field "Meas Freq" [h for help] :

Sorry, "" is not a valid type.

The valid types are: { real \| int \| char \| boolean \| string }

Enter a type for editor field "Meas Freq" [h for help] : real


Enter the maximum legal value for this field: 100e6


rounding): 1

Enter an initial value for this field: 1e6


if no more fields are desired): Integ Time

Enter a type for editor field "Integ Time" [h for help] :

char

This field will force the user to enter one character, from

within a set of valid characters you will specify now.

Example set of valid characters: TFYN

Enter the set of character values that this field can take

on: SML

Enter whether this field should force user input to

uppercase [y/n]: y

Enter an initial value for this field: S


if no more fields are desired): Osc Level [.01-1Vrms]

Enter a type for editor field "Osc Level [.01-1Vrms]"

[h for help] : real

Enter the minimum legal value for this field: .01



rounding): 0 Enter an initial value for this field: .01



55

if no more fields are desired): Averages [1-256]

Enter a type for editor field "Averages [1-256]" [h for help]

: int





if no more fields are desired): Delay for Timeouts

Enter a type for editor field "Delay for Timeouts"

[h for help] : real


Enter the maximum legal value for this field: HUGE


rounding): 0



if no more fields are desired):

Done. All necessary C++ UI code was added to user_meas.hxx

and user_meas.cxx.

From the nature of the queries in this script, this process defines an editor table for theinstrument. The table offers some advanced features, such as constraining the type andthe range of values that an operator can enter in each field.

Running the Scripts on Windows

To run the mk_instr, mk_unit, and mk_instr_ui scripts on Windows:

First edit the $ICCAP_ROOT/bin/icrun.bat file. You must set ICCAP_ROOT variable by1.modifying the following statement:SET ICCAP_ROOT=<Path to IC-CAP>

for example, if you have installed IC-CAP at C:\Agilent\ICCAP_2011, edit the file toread:SET ICCAP_ROOT=C:\Agilent\ICCAP_2011

Once ICCAP_ROOT variable is set, change <name> in the below statement to mk_instr,2.mk_unit, or mk_instr_ui to run those scripts.icrun <name>

Running the Scripts on UNIX

This section contains information on running the scripts on UNIX, queries asked by thescript, and the form of user responses.

The scripts are invoked as UNIX commands. Execute cd $ICCAP_ROOT/src unless youjust want to experiment with the scripts in another directory like /tmp. The cdcommand ensures that the code moves where the Makefile expects.Generate backup copies of user_meas.hxx and user_meas.cxx files before using thescripts.The scripts run in following order:

mk_unitmk_instrmk_instr_uiRunning the scripts out of order may cause compilation errors when thecompiler encounters types, classes, or variables before they are properlydeclared.

All the scripts prompt with a series of queries. The effect of the scripts is to writeC++ code onto the end of the user_meas.hxx header and the user_meas.cxximplementation files.

Plan your response by reviewing the transcripts and comments shownpreviously for each script to avoid re-running the script. Multiple passes bythe scripts could put declarations into user_meas.hxx more than oncecausing error messages from the scripts or a compile-time message suchas error 1113: class <some_class_name> defined twice. To re-run thescript, restore user_meas.hxx and user_meas.cxx to the same state asthey appear on the IC-CAP product media.

When providing real number values to mk_instr_ui, supply values that a C compileraccepts. The engineering notation accepted by IC-CAP's PEL interpreter, such as15meg, or 2k, is not accepted by the compiler. Examples of acceptable real numbersare:

1.010.5e6HUGE (a constant from /usr/include/math.h)

granularity as used in real fields, refers to a flexible rounding feature. For example, ifyour instrument has an option Osc Level (for which the instrument has only 10 mVresolution) enter granularity as 10e-3. The Instrument Options editor then protectsthe IC-CAP operator from entering values the instrument cannott support.

The scripts require you to fill the functions in user_meas.cxx. They also require a fewminor adjustments in user_meas.hxx. These adjustments are:

The instrument class should declare any units owned by the instrument. This isdiscussed under mk_instr.You may encounter compilation errors when unit and instrument functions attemptaccess to each others data members, since this violates normal C++ access rules.For example, in user_meas.cxx, in hp4194::init_instr(), a function of the hp4194class accesses a data member of the cvu_4194 class with this statement: cv_unit ->oscillator_on = 1;A typical compiler error message could be:

error 1299: init_instr() cannot access cvu_4194::oscillator_on:

private member

One workaround is to let the unit and instrument class declare each other as friends.For example, the declarations

friend class cvu_4194;

and

friend class hp4194;

in user_meas.hxx, permit the hp4194 functions to access the cvu_4194 datamembers, and vice-versa.


56

Filling in Necessary Functions

After running the scripts, you must write the body portions of the functions added touser_meas.cxx. This section provides hints on how you can accomplish this.

For help filling in a function body, look at the declarations and functions generated by thescripts. These provide comments explaining the purpose, return values, and invocationtime of each function.

Next, look at the declarations and functions of the HP 4194 example driver. This sectioncontains examples of code accomplishing required tasks. The following manual sectionsmay also be helpful.

Programming with C++Order in Which User-Supplied Functions are Called. Provides useful information aboutthe sequence in which functions are invoked. Decisions must often be made aboutwhich function should perform particular instrument manipulations; these decisionscan be aided by seeing when each function runs.What Makes up an IC-CAP Driver. Explains the functionality expected in areas suchas Calibration and Hardware Setup Operations. The functions whose bodies you needto write are grouped in that section by functional category.

You may want to proceed in stages. For example, start with Hardware Setup Operations todemonstrate that Rebuild (instrument list) can find the instrument and display the driverand instrument in the Hardware window. Then implement the functions that supportMeasure. Address those functions that support Calibrate, if desired. During the time yourdriver is partially implemented, compiler warnings serve as a rough indication of functionsnot yet implemented.

The GPIB analyzer (Tools menu), and especially its macro features (described elsewherein this section), are helpful when developing the appropriate sequence of commands touse with the instrument.

Making a New Instrument Type Known to IC-CAP

Running the mk_instr script makes a new instrument type known to IC-CAP. The codeinvolves an add_user_driver() function call, placed in user_meas.cxx by the mk_instrscript.

Creating a New Shared Library

After any series of edits to the source files, you must generate 1 or 2 new shared librariesto pick up the modified files. The shared library names are libicuserc .<ext> andlibicusercxx.<ext> where ext is a platform-specific extension. Use the extension .so forSolaris. The library libicuserc.<ext> holds C code and is used to add user C functions. Thelibrary libicusercxx.<ext> holds C++ code and is used to add instruments. The defaultlocation of these files on SUN Solaris 2.X is $ICCAP_ROOT /lib/sun2x. When you issue themake command, you will create a local version of the same file that includes yourmodifications. By setting an environment variable, you can direct IC-CAP to use your newshared library instead of the default library.

To generate the new shared library:

Create a work directory for the source files (for example, mkdir my_source, and1.change it to (cd my_source).Copy the set of source files from $ICCAP_ROOT/src to the new work directory (cp2.$ICCAP_ROOT/src/* .).Use the touch command on the *.o files so that all *.c and *.o files appear to have3.been created at the same time (touch *.o). (This step is important for the makeprocedure.)

NoteIf the drive you're copying to is NFS mounted, clock skews can result if the NFS drive's system has aslightly different system time than the local system. If you think this might apply to you, first,execute touch * then execute touch *.o. The first touch synchronizes all files to your local system'stime; the following touch causes the make system to believe that all of the .o files were generatedlater than the source files, so it will not attempt to rebuild any unnecessary files.

Copy your source code to the working directory. Modify the function4.add_users_c_funcs() in userc.c to add your C functions to IC-CAP's list of functions,and/or modify the function add_users_drivers() in user_meas.cxx to add your driversto IC-CAP's library of instrument drivers. Modify the Makefile to add your source codemodules to the list of objects.Issue the make command and debug any compiler errors.5.Set the environment variable ICCAP_OPEN_DIR to point to the directory containing6.the libicuserc .<ext> or libicusercxx.<ext> file where ext is a platform-specific fileextension (ext is .so on Solaris).

Alternately, if you want to use the new files site wide, you can replace theoriginal files (after copying to another name to preserve them) under$ICCAP_ROOT/lib/<platform>.

Start IC-CAP as usual.7.

Troubleshooting Compiler Errors

The definitive authority on compiler errors is your compiler documentation. This sectionoffers assistance with some of the common messages you may encounter when compilingOMI drivers.

The message CC: "user_meas.cxx", line 899: warning: outptr not used (117)

usually indicates that you have not yet filled in a function, with the result that the functionis not using all of its arguments. In some cases the function may not use all of itsarguments, so the message may not be important.

Resolution of the message error 1299: some_unit_func cannot accesssome_instr_class_name::some_member: private member

is discussed in Running the Scripts on Windows.

The message CC: "user_meas.hxx", line 9: error: class x defined twice (1113)

indicates that the Driver Generation Scripts were probably run twice.

For help, refer to Running the Scripts on Windows.

Debugging


57

This section provides information about debugging driver code, after iccap.new has beencompiled, including the xdb debugger and GPIB analyzer (Tools menu).

Using the xdb Debugger

The default Makefile arranges for debug information to be available after linking theexecutable file. This is done with the -g flag among the CFLAGS in the Makefile.

The debugger commands described in the following table should be tried in the orderpresented.

Debugger commands for the xdb debugger

Command Action

cd$ICCAP_ROOT/src

Changes directories. Debugging works best when the current directory contains thesource files and the binary.

xdb iccap.new Starts the debugger.

z 8 isr; z 16 isr; z18 isr

Tells the debugger not to interfere with 3 signals managed by other parts of iccap.new.

r bjt_npn.mdl Runs iccap.new, specifying bjt_npn.mdl as a command line argument. If the debuggerstops with a message such as bad access to child process, ignore it and enter c tocontinue.

BREAK or CTRL-C Suspends iccap.new, to give further debugger commands. xdb does not executecommands unless iccap.new is suspended.

v user_meas.cxx Enables you to view and edit a source file. This command is helpful for settingbreakpoints.

td Toggle display. Toggles the display mode between assembly code and C/C++. Use this ifthe preceding command displays assembly code on the screen, or if no code is displayed.

/hp4194::find Searches forward (as in vi) to view the source for the function hp4194::find_instr().

v nnn Enables you to view line nnn at the center of the screen where nnn is the line numberyou want to view.

b Sets a breakpoint at the line currently centered on the screen. Sometimes the debuggerchooses another nearby line, especially if the currently centered line is blank, or is only adeclaration statement. When iccap.new resumes running, the debugger stops iccap.newwhenever this line of code is about to be executed. You may set several breakpoints.

b nnn Similar to the last command. Sets a breakpoint at line nnn where nnn is the line numberyou specify. Sometimes the debugger chooses another nearby line, especially if youchose a blank line, or a line with only a declaration statement.

S Big step. Steps through 1 line of source code without stopping inside any procedure callsencountered.

Little Like S, but this stops inside any debuggable procedure that is encountered whileexecuting the line of code.

c Continues execution of iccap.new.

Execute a menufunction

To reach breakpoints in the driver code, use Measure, Calibrate, or Rebuild asappropriate. For help in making this choice, refer to Order in Which User-SuppliedFunctions are Called. Be sure that the function will actually be called if you want thebreakpoint reached.

p address When the debugger hits a breakpoint in a procedure, this command prints the value of anargument passed to the procedure, or a local variable in the procedure. In this example,the argument/variable is named address.

p address=23 To assign a new value to an integer variable named address, employ this special form ofthe p (print) command.

p *this\K Prints the member data of the C++ object in whose member function the currentbreakpoint is located.

GPIB Analyzer (Tools menu) and IC-CAP Diagnostics

In addition to xdb, debugging capabilities are built into IC-CAP.

The GPIB analyzer (Tools menu) in the Hardware Setup window includes the followingfeatures.

The I-O Screen Debug On menu selection can monitor all activity on the GPIB bus.Observe the GPIB commands and responses associated with your driver, as well asother IC-CAP drivers.The analyzer can be used for interactive I/O activities, to force an instrument state,poll the instrument, or test the effect of a command.Analyzer operations can be collected into a file for macros for rapidly prototyping theGPIB commands to be used in a driver. For more information about macro files of thissort, refer to GPIB Analyzer (measurement).

The generation of IC-CAP diagnostic messages can be activated by menu functions underTools in the IC-CAP Main window.

Alternatives to Creating New Drivers

If you don't need an instrument driver to be as fully integrated as HP/Agilent-provideddrivers, it may be worthwhile to consider controlling the instrument by means less formalthan creating a driver using the Open Measurement Interface.

NoteThere is an important shortcoming with these suggestions. An IC-CAP measurement currently provides nomechanism for Program Transforms or Macros to be invoked at critical times in the interior of themeasurement (for example, at the instant when DC bias levels have just been established by SMUs, and itis time for a main sweep instrument to stimulate the DUT and collect data). Use of the Open MeasurementInterface overcomes such limitations.

Use the PRINT statement in an IC-CAP Macro to direct commands to an instrument,when a suitable device file has been established using the mknod command.Use the functions listed with USERC_write and USERC_read in a Program Transformor Macro to provide limited instrument control. For descriptions of the User Cfunctions in general, refer to Main ICCAP Functions (extractionandprog). For detailsand examples of the input/output functions, refer to User C Functions(extractionandprog).Rather than using the Measure menu selection directly, construct Macros in thefollowing style to enclose the measurement between operations controlling otherhardware:

! Steps 1, 2, and 3 are assumed to be implemented by PRINT.

! 1) Force next desired set point on temperature chamber.

! 2) Enable waveform generator.

iccap_func("/opamp/time_domain/positive_slew","Measure")

! 3) Disable waveform generator.

! One way to control the values desired for temperature and

! frequency is to access IC-CAP system variables.

What Makes up an IC-CAP Driver

In addition to measurement capabilities, each IC-CAP driver possesses other capabilities,


58

such as the user interface functionality provided in Instrument Options folder and theability to participate in Input, Output and Setup Checking prior to measuring. Each ofthese essential areas is discussed in this section. In each area, information is providedabout the specific functions necessary to complete that part of a driver.

In the tables throughout this section, the prefix unit:: means the class name(s) youprovided for units when you ran the mk_unit script. The prefix instr:: should beconsidered to mean the class name you provided for the instrument when you ran themk_instr script. The column Importance indicates whether you typically need to write anycode for the function. Because of the inheritance features of C++, you must often rely oninherited default functions. Functions important to write, typical return values, and otherinformation can be determined from the comments for the function in$ICCAP_ROOT/src/user_meas.cxx.

Instrument Options

The Instrument Options folder provides a method for selecting certain instrumentconditions for a measurement. Certain instrument conditions are separated into differentgroups of instrument options (rather than appearing in Input sweep editors) because theyare highly instrument specific, and play no role in simulation. The options displayed in theInstrument Options folder typically vary with each setup that participates in themeasurement involving a particular instrument.

The Driver Generation Scripts, described in Procedure for Adding a Driver, can write all theC++ code that is necessary to establish appropriate instrument options tables for a newdriver. The driver generation script named mk_instr_ui prompts for the desired contentsof the instrument options tables, after which it proceeds to generate the necessarydeclarations and implementations in C++. The generated code will contain data structuresin which options are stored, as well as the user interface linkages that display the optionsfor editing.

Input, Output and Setup Checking

When you initiate Measure or Calibrate for a Setup, IC-CAP first verifies the validity of themeasurement Setup. This permits many operator errors to be detected and reportedbefore IC-CAP undertakes instrument I/O.

IC-CAP performs the following 3 kinds of checks:

Checks Input (Sweep) specifications; for example, does a Start or Stop value exceedthe instrument's range?Checks Output specifications; for example, can the instrument measure the type ofdata desired, such as capacitance?Checks overall Setup structure; for example, is there more than 1 time or frequencysweep being requested?

The following table describes the functions related to input (sweep) checking.

IC-CAP Input (Sweep) Modes shows a summary of the supported Input (Sweep) modes inIC-CAP. The column Character Used in Driver Functions shows the character passed whenan Input Mode is passed to a function, such as unit::can_source.

Functions for Output Checking describes the functions related to output checking.

IC-CAP Output Modes shows a summary of the supported Output modes in IC-CAP. Thecolumn Character Used in Driver Functions shows the character passed when an OutputMode is passed to a function, such as unit::can_measure.

Functions for Input Checking

Function Name Purpose Importance

instr::use_second_sweep tells if unit has 2 internal sweeps default usually OK

unit::can_source tells if unit can source a given Mode important

unit::can_source_vs_time tells if unit can source time-domainsignals

important for pulse generators

unit::check_bias_swp reserved for future use default is OK

unit::check_sweep lets unit check/preview Input data set important

unit::check_sync checks sync sweep spec. important if implementing sync sweeps IC-CAP Input (Sweep) Modes

Character Used in Driver Functions Meaning

V Voltage

I Current

F Frequency

T Time

P Parameter

U User (refer to User-Defined Input and Output Modes) Functions for Output Checking


unit::can_measure tells if unit can measure a given Mode important

unit::can_measure_vs_time tells if unit can measure time-domainsignals

important for oscilloscopes

unit::check_out lets unit check/preview Output data set important if measures multiple datasets

IC-CAP Output Modes

Character Used in Driver Functions Meaning

V Voltage

I Current

C Capacitance

G Conductance

T Time Domain Pulse Parameter (like RISETIME)

S, H, Z, Y, K, A Two-Port

U User (refer to User-Defined Input and Output Modes)

Setup checking is performed primarily by logic embedded in IC-CAP. A limited amount ofthe checking is accomplished with user-supplied functions. The following table describesthe user functions related to overall Setup checking.

Functions for Overall Setup Checking


59


instr::find_instr checks GPIB for instrument necessary

instr::find_units locates optional units default usually OK

instr::set_found remembers instrument was found; could set internal flagsconcerning presence of optional hardware modules

default usually OK

instr::use_second_sweep tells if unit has 2 internal sweeps default usually OK

unit::bias_compatible checks if this unit can tolerate signal or bias from anotherunit

could potentiallysave fuses.

unit::can_do_second_sweep tells if another sweep and unit can be an internal sweepsecondary to the sweep for this unit

default is OK

User-Defined Input and Output Modes

Mode U is a reserved user-defined mode that allows some flexibility for safely checkingany new signal modes to be sourced or measured. This feature is for situations where it isnot practical or safe to use existing Input or Output modes (such as voltage orcapacitance).

The following considerations apply:

Units associated with existing drivers are likely to reject U. For example, a HP 4141VM unit will not force or measure U. In such a case the measurement is disallowed.(It does not make sense for the IC-CAP HP 4141 driver to try to force or measure U-Mode data, since it does not know what U-Mode means.)The unit functions associated with the new driver can enforce any desired policy for aU-mode Input or Output, as well as the other Input and Output Modes.With a U-Mode Input or Output in an IC-CAP Setup, do not expect the Simulate menufunction to work on that Setup.

Calibration

Calibration functions are associated with the instrument, not its units. To performcalibration procedures initiated from the IC-CAP program, implement the functions shownin the following table.

Functions for Calibration


cal_possible tells if the other 2 functions do anything These functions are necessary if IC-CAP is tocalibrate the instrument.

do_cal downloads Setup, leads operator throughcalibration procedure

recall_n_chk_calib activates calibration during Measure;checks sweep

Several of the functions required for Measure are also used during Calibrate. Refer toOrder in Which User-Supplied Functions are Called in this section for a list of functionscalled during Calibrate.

Storage is provided in the instr_options class for limited calibration data for a particularinstrument in a particular Setup. The instr_options class is declared in instr.hxx.

The data members in instr_options for holding calibration results are:

String cal_data; //declare your own data if String is not an appropriate typecalib_status last_cal_status; //calib_status is an enumeration with these possible values:

CAL_OKCAL_ERRORCAL_ABORTEDCAL_NEVER_DONE

Set calib_status during do_cal() and test it during recall_n_chk_calib(). Recall thatcal_possible() and do_cal() are invoked (in that order) during Calibrate, whilerecall_n_chk_calib() is later called during Measure, with the purpose of enabling thedesired calibration set.

Derived from the class instr_options (declared in instr.hxx) is user_instr_options, declaredin user_instr.hxx. For the new driver, a further derived class will have been declared inuser_meas.hxx by the mk_instr_ui script. The section Class Hierarchy for User-Contributed Drivers clarifies the relationships of these classes.

The class in user_meas.hxx that is derived from user_instr_options is an appropriate placeto declare additional calibration data the workstation should retain, because a distinctobject (or data structure) of this type exists in every situation where distinct instrumentcalibration data might be needed. In other words, an instrument has a distinctuser_instr_options object in every Setup where the instrument is used. For the example ofthe HP 4194 driver, such data (if any) would be declared in the class named hp4194_tablein user_meas.hxx. You might declare several double numbers, to keep a record of sweeplimits that were in effect at the time of Calibrate, so that they can be verified duringMeasurement. (With many instruments, calibration is not valid unless measurementsemploy the same sweep limits that were in effect during calibration.)

NoteTo simplify an initial pass at implementing calibration, do not declare additional data structures forremembering sweep parameters, and do not perform much verification during recall_n_chk_calib().

If you choose to declare additional calibration-related data in the class derived fromuser_instr_options, it is possible for this data to be archived and re-loaded with IC-CAPModel(.mdl), DUT(.dut), and Setup(.set) files. Note that the archiving of user-definedcalibration data is an advanced feature that most implementations can probably avoidconsidering.

To archive user-defined calibration data, your class derived from user_instr_options mustredeclare and implement 2 virtual functions. These functions are read_from_file andwrite_from_file, declared for the class instr_options, in the file instr.hxx. When called,these functions receive an open stdio FILE*, which provides read or write access to the IC-CAP archive file at the appropriate time during a Read From File or Write to File menufunction.

Measurement: Initialization, Control and Data Acquisition

The functions in this area perform the real work of the instrument driver; this areaaccounts for the largest number of functions present in each driver.

Initialization functions are listed in the following table.

Initialization Functions


60


instr::init_instr downloads information from the InstrumentOption Table

necessary

instr::reset_instr_info clears flags in driver, refer to instr.hxx default usually OK

instr::reset_outptrs nulls out output data set pointers, refer toinstr.hxx

default usually OK

instr::zero_supplies puts instrument to safe state, turns offsources

necessary

unit::enable_output enables any output unit needing explicitenabling (refer to user_meas.cxx)

necessary with some instruments,such as the HP 4141

unit::init_unit reserved for future use default is OK

unit::reset_inassign reserved for use by 4142 and 4145 default is OK

unit::reset_outassign reserved for use by 4142 and 4145 default is OK

unit::set_2_internal_sweeps downloads specifications for 2 nestedinternal sweeps

default usually OK

unit::set_internal_sweep downloads specifications for internal sweep necessary

unit::set_sync downloads specifications for sync sweeps default usually OK

unit::zero_supply puts unit to safe state, suppresses bias andso on

necessary, if the unit can sourcebias or other signal

Control and data acquisition functions are shown in the following table.

Because many of the functions in this category must perform non-trivial work, such asinstrument communication and error reporting, refer to Programming with C++, wheresuch operations are explained. The examples for the cvu_4194 member functions and thehp4194 member functions in user_meas.cxx are also helpful.

A few of the functions in this area are provided for the support of a particular instrument,for example, the HP 4145. The intermediate classes user_unit and user_instr do notredeclare some of these low-usage functions, though their declarations are inherited fromthe unit and instr classes, so they could be used in a new driver if needed. For example,instr::use_second_sweep() is re-declared and used only by the HP 4145 driver.

Control and Data Acquisition Functions


instr::copy_outds does delayed data set stuffing (refer to instr.hxx) default usuallyOK

instr::fill_outds similar to copy_outds (refer to instr.hxx) default usuallyOK

instr::get_outptr gives pointer to Output data set default usuallyOK

instr::keep_mdata keeps 1 data point default is OK

instr::out_count tells number of output pointers in instr class default usuallyOK

unit::can_do_second_sweep tells if 2 internal sweeps OK (refer to unit.hxx) default is OK

unit::define_channel reserved for 4145 default is OK

unit::enable_sync reserved for future use default is OK

unit::fill_outds any data this unit has kept internally, or in data structures ofthe instrument or unit driver, that belong in outptr, should besaved there now (refer to user_meas.cxx)

default usuallyOK

unit::get_data gets data from the instrument (Refer to user_meas.cxx) necessary

unit::get_int_bias reserved for future use default is OK

unit::get_scalar_data reserved for 54120 series default is OK

unit::list_chan_num reserved for 4142 default is OK

unit::list_output_name reserved for 4145 default is OK

unit::meas_err used by some drivers to make error messages default is OK

unit::set_bias forces a bias value necessary foruser sweep

unit::set_data_out reserved for 4145 default is OK

unit::set_scalar reserved for 54120 series default is OK

unit::source_const_unit reserved for 4145 default is OK

unit::source_unit reserved for 4145 default is OK

unit::trigger directs a unit to perform the measurement specified earlier viaset_internal_sweep

necessary

unit::turn_chan_OFF reserved for 4145 default is OK

unit::wait_data_ready allows the instrument to finish measurement before trying toget data after trigger. In the cvu_4194 code, this wait isaccomplished in the trigger function.

default usuallyOK

unit::wait_delay_time implements Delay Time prior to a spot measurement. Refer tocvu_4194 case in user_meas.cxx.

necessary

unit::wait_hold_time implements Hold Time prior to a User Main Sweep. Refer tocvu_4194 case in user_meas.cxx.

necessary

Hardware Setup Operations

The Hardware Setup functions, listed in the following table, are used in the followingoperations:

Maintaining lists of instruments and unitsAdding and deleting instrumentsMaintaining the unit table, including the addition of entries due to newly addedinstrumentsThe Rebuild (instrument list) function

Self-testing the instrumentsPolling the instruments

Hardware Setup Functions


61


instr::addl_addr_label reserved for 8510 and 8753 default is OK

instr::build_units creates unit objects. Refer to hp4194::build_units inuser_meas.cxx.

necessary

instr::find_instr checks GPIB for instrument necessary

instr::find_units locates optional units default usually OK

instr::get_addl_addr reserved for 8510 and 8753 default is OK

instr::get_ID gets instrument ID string. Refer to user_meas.cxx. necessary

instr::get_unit_by_name finds a unit in the instrument default is OK

instr::instr * initializes data members of instrument object necessary

~instr::instr * cleans up members of instrument object necessary

instr::read_units reserved for 4142 default is OK

instr::rebuild_units reserved for 4142 default is OK

instr::set_found remembers that instr found; may test and set internal flagsconcerning presence of optional hardware modules

default usually OK

instr::test_instr supports the Run Self Tests menu function in the HardwareSetup window

default is OK (noself-test)

instr::unit_count tells how many units the instrument has necessary

instr::units_configurable reserved for 4142 default is OK

instr::write_units reserved for 4142 default is OK

unit::unit * initializes data members of unit object necessary

~unit::unit * cleans up members of unit object necessary

* denotes a constructor or destructor function for which the actual name is the unit orinstr class name chosen when the mk_unit and mk_instr scripts were run. For example,hp4194::hp4194, in user_meas.cxx.

Programming with C++

This section provides examples of code for common Open Measurement Interfaceprogramming tasks.

Access to Inputs (Sweeps) and OutputsError and Warning MessagesReading from an InstrumentSerial Poll of an InstrumentString HandlingTime DelayUser Input with a Dialog BoxWriting to an Instrument

Access to Inputs (Sweeps) and Outputs

In user_meas.cxx the function cvu_4194::check_sweep demonstrates how to determinesweep properties like Mode (V, for example), Type (LOG, for example), compliance, andstart and stop values.

IC-CAP computes all necessary step values. Do not attempt to compute them from start,stop, and so on, because simulations will use the values IC-CAP computes. Instead,access individual sweep steps with the get_point function.

Following are statements from cvu_4194::check_sweep that determine sweep propertiesand get sweep values. These statements are isolated examples and are not necessarily tobe used in the order shown.

int cvu_4194::check_sweep(sweep* swp)

// header of the function used here

sweep_def *swpdef = swp->get_sweep_def();

// a sweep uses sweep_def for values

switch(swpdef->get_esweep_type())

// to see if it's CON, LOG, LIN, ...

compval = swp->get_compliance();

// compliance

case CON:

val1 = ((con_sweep *)swpdef)->get_value();

// value of CON sweep

case LIN:

val1 = ((lin_sweep *)swpdef)->get_start();

// start value of LIN sweep

val2 = ((lin_sweep *)swpdef)->get_stop();

// stop value

((lin_sweep *)swpdef)->get_stepsize()

// step size

// next 2 are taken from cvu_4194::set_internal_sweep:

linswp = (lin_sweep*)swpdef;

// to enable lin_sweep functions

numpoints = linswp -> get_num_points();

// number of points

if (swp->get_sweep_order() == 1)

// sweep order; 1 => main sweep

switch (swp->get_mode())

// Mode: 'V', 'I', 'F', ...

swp->get_size() // Number of points

swp->get_point(step_num)

// get one point (indexed from 0)

The class named sweep is declared in sweep.hxx. Using a sweep often involves usingfunctions it inherits from the class ds (data set), declared in ds.hxx. The functionget_point is an example of a function inherited from ds. The sweep_type class is insweep_type.hxx.

To save measured data to an IC-CAP Output data set, employ the style incvu_4194::get_data:

dsptr -> keep_point (index++, datapoint, DATA_MEAS);

// datapoint is a double

In the example, dsptr points to a ds object. The class ds declares other forms of thekeep_point function in ds.hxx. These can store complex or 2-port matrix data into theOutput data set.

Error and Warning Messages

The IC-CAP error box appears after a measurement, displays one or more messages, andmust be dismissed by clicking OK if you make one or more statements such as

errbox << "ERROR: HP4194 unsupported internal sweep type."

<< EOL;

errbox << "ERROR: HP4194 sweep produced " << num_points_kept


62

<< "when" << swp_num_points << " were requested" << EOL;

Warnings are displayed in the Status window:

cerr << "WARNING: HP4194 frequency rounded up to 100Hz" << EOL;

The objects errbox and cerr accept any number of arguments, of various types, includingdouble, String, char*, int, and char. Separate them with <<.

Reading from an Instrument

The user_meas.cxx file demonstrates 2 styles. Writing and reading are done with separatecalls. In hp4194::get_id a readstring function is used as follows:

stat=ioport->readstring(ad,id_buf,255);

// below is the code needed to call readstring from

// a unit class function

stat=get_io_port()->readstring(ad,id_buf,255);

// because the instrument owns and maintains the ioport

// object, the unit gets it this way before using it

The first argument above is the GPIB address. The id_buf argument is a buffer guaranteedto be adjusted by readstring to hold 255 bytes, if the read produces that many.

A function is also provided to write a query and then read an answer:

if (ioport->write_n_read(addr,"MKRB?", urbuf, 80) == -1)

The first argument above is the GPIB address. The second argument is a char* to bewritten. The third argument is a buffer guaranteed to be adjusted by readstring to hold 80bytes, if the read produces that many.

The above functions are 2 of many available for an hpib_io_port. Complete declarations ofits functions are in io_port.hxx.

Serial Poll of an Instrument

The following functions are 2 of many available for an hpib_io_port. Complete declarationsof its functions are in io_port.hxx.

Serial polling is done as follows:

int status_byte = ioport->spoll(addr);

// this example not from user_meas.cxx

int status_byte = get_io_port()->spoll(addr);

// call from a unit function

To wait for a particular serial poll bit:

// from cvu_4194::zero_supply:

hpib_io_port *ioport = get_ioport();

// bit-weight 1 below is to await 'measurement complete bit'

if (ioport -> poll_wait(addr, 1, 0, 10.0) == -1)

The arguments are: GPIB address, bit-weight to wait for, a flag reserved for future use,and maximum time that poll_wait should try (10 seconds).

String Handling

C++ offers a substantial improvement over C for handling String type data. In the fileString.h a number of String functions are declared. The following code demonstratesseveral.

String str_hello = "hello"; // declare and initialize a string

String str_world; // just declare

str_world = "world"; // assignment

String hello_world = str_hello + " " + str_world;// concatenation

errbox << hello_world + "0; // writing to errbox

if ("hello world" == hello_world) // test for equality

String instr_cmd = "*RST"; // initialize for next statement:

if ioport->writestring(addr,instr_cmd) == -1) // String to instrument

In the final example, a char* is expected by writestring, and C++ automatically extracts itfrom the String. Do not pass a String to printf or scanf. The declarations of these functionsin /usr/include/stdio.h use the ellipsis notation (...), so C++ does not know that a char*should be passed to them.

Time Delay

An example of a time delay is:

delay (10E-3); // 10 millisecond delay

User Input with a Dialog Box

A number of functions for this purpose are declared in dialog.hxx. Examples to get datafrom dialog boxes are:

// These use the versions of get_double and get_String that

// each take 3 arguments.

double double_result;

String String_result;

int ok_or_cancel; // 0 => OK pressed by user, and -1 => CANCEL

ok_or_cancel = get_double ("Give a double:",default_dbl_val,&double_result);

ok_or_cancel = get_String ("Give a string:",default_string,&String_result);

Writing to an Instrument

An example of writing to an instrument is:

if (ioport->writestring(addr,"TRIG") == -1)

// cvu_4194::zero_supply

The arguments are the GPIB address and a char* string to send. You can also write aquery and read a response with 1 call, write_n_read, discussed in Reading from anInstrument. Writestring and write_n_read are 2 of many functions available for anhpib_io_port. Complete declarations of its functions are in io_port.hxx.


63

Syntax

This section provides help with reading the IC-CAP source code in user_meas.hxx,user_meas.cxx, and the various include files. Follow the example code in user_meas.hxxand user_meas.cxx when implementing a new driver.

NoteFor best results when using the vi editor to browse the source files, execute the command :set tabstop=3

The C++ language introduces several keywords to help understand OMI programming, forexample, class, new, delete, and virtual. Terms that are peculiar to OMI programming, forexample, Measurer, sweep type, sweep order, main sweep, internal sweep, user sweep,unit function, and instrument data, are used in this section and in the source files.

Function declarations in C++ use the improved function prototypes of ANSI/C. Forexample,

int mult_by_2(int input);

// style for forward declaration int y=2;

int y = 2;

int x = mult_by_2(y); // example of invocation

int mult_by_2(int input)

// style for implementation (SAME AS DECLARATION)

{

return 2*input;

}

This is an area of incompatibility with original (Kernighan and Ritchie) C. However, it iseasier to read, and write, and is the emerging new standard. It also gives the compilerinformation with which function call argument lists can be checked, saving run-timeaggravation.

Sometimes in class declarations you will see the function body present:

const char *class_name()

// this code from user_meas.hxx

{ return "cvu_4194"; }

These cases are called inline functions. They behave like normal functions, but the C++compiler emits code inline, without normal function call overhead. For short functions thisreduces both execution time and code size.

New Symbols and Operators

This section defines new symbols and operators in C++.

// A pair of slashes introduces an end-of-line comment. (/* and */ can still be used forC-style comments.)

& Appearing after a type name or class name, & usually indicates that an argument to afunction is passed by reference. Although C can pass arguments by address, the C++notion of reference arguments eliminates many error-prone uses of * (pointer de-reference) and & (address) operators used with pointer handling in C. In the followingexample, the called function increments the callers variable:

// 'input' passed by reference:

void increment(int& input) {

input++; // need not use *input

}

int x=3;

increment(x); // Need not pass &x

// Now x is equal to 4

object. member_function() In C, the . operator is used to access data members in astruct object. In C++, . is also used to access (execute) function members.

ptr_to_object->member_function() In C, the -> operator is used to access datamembers in a struct object to which one holds a pointer. In C++, -> is also used to access(execute) function members of a class type object to which you hold a pointer.

Class Hierarchy for User-Contributed Drivers

The diagram in the following figure depicts the relationships of the classes that are ofprinciple interest to a user creating a driver. The arrows without labels indicate pointersheld in the objects.

Classes Involved in the HP 4194 Example of a User Driver

At the top of the hierarchy are classes named unit, instr, and instr_options. All instrumentdrivers in IC-CAP consist of classes derived from these 3 classes. When AgilentTechnologies adds a driver to IC-CAP, one new class is derived from instr, one or morenew classes are derived from unit, and one new class is derived from instr_options. Theprocess of deriving new classes from these base classes permits the new driver toefficiently reuse generic functionality present in the base classes, while also introducingnew code where necessary to accommodate the specialized needs of the new instrument.

The division of a driver into unit, instr, and instr_options components helps modularity.


64

Generally, the role of each of these parts is as follows:

instr - manages operations associated with the whole instrument, such as self-testand initialization.instr_options - presents a user interface for and stores the values of options uniqueto each Setup in which an instrument is used. For example, the option Use UserSweep determines whether an instrument does or does not use its internal sweepcapability during measurement in a particular Setup.unit - manages operations on a single SMU for example, in the case of a DC analyzer,or on a single oscilloscope input, in the case of a multi-channel digitizing oscilloscopelike the 54120 series. Operations undertaken at this level include the application ofDC and other signals to the DUT, as well as the acquisition of the measured data.

Intermediate in the hierarchy are 3 classes named user_unit, user_instr, anduser_instr_options. These serve the following purposes:

Hide any virtual functions of unit and instr that are unlikely to be necessary tooverride in new driver classes. This allows the critical function declarations to beconcentrated in one location, with comments close at hand.Introduce new member functions provided for Instrument Options management orfor convenience.

At the bottom of the hierarchy are examples of classes introduced by the DriverGeneration Scripts. When a user adds a driver to IC-CAP, the Driver Generation Scriptsadd a class derived from user_instr, one or more classes derived from user_unit, and aclass derived from user_instr_options. The class derived from user_instr_options, which ishp4194_table in the example driver, is completely declared and implemented when theuser runs the mk_instr_ui script. In other words, a programmer using the OpenMeasurement Interface need not become involved with any coding that pertains to thisuser interface component of the driver. The programmer also does not need to providedeclarations for any new classes needed for the driver, since these are completely writtenout when the driver generation scripts mk_unit, mk_instr, and mk_instr_ui are run.However, the programmer is required to fill-in the implementations of several functionsthat ultimately perform the work done by the driver.

Order in Which User-Supplied Functions are Called

The 4 tables below illustrate the following 3 essential instrument operations:

Rebuild (instrument list)CalibrateMeasure

These tables are representative of a typical order of invocation. Some functions may beused more than once, particularly since Measure involves looping through different biaslevels. The column Function Category indicates the location of further information aboutthe function in What Makes up an IC-CAP Driver. Other valuable information is located inthe comments for each function, provided in user_instr.hxx, user_unit.hxx,user_meas.hxx, and user_meas.cxx.

During Rebuild

During this operation, the Hardware Manager locates addresses that respond to a serialpoll. At each such address, available drivers determine if they own the instrument, until 1driver succeeds. They try in the order shown in the Instrument Library list. Note thatunless find_instr() is successful, none of the ensuing functions are called.

The functions called during Rebuild (instrument list) are shown in the following table.

Functions Called During Rebuild (instrument list)

Function Name Function Category

get_addl_addr Hardware Editor Operations

addl_addr_label

find_instr

units_configurable

rebuild_units or build_units

find_units

set_found

unit_count

get_unit

During Calibrate

During this operation the Measurer initiates calibration procedures for each instrument ina Setup that has calibration supported by IC-CAP.

The functions called during Calibrate are shown in the following table.

Functions Called During Calibrate

Function Name Function Category

instr::find_instr Setup Checking

unit::get_int_bias Control and Data Acquisition Functions

unit::can_source Checking of Inputs

unit::can_source_vs_time Checking of Inputs

unit::can_measure_vs_time Checking of Outputs

unit::can_measure Checking of Outputs

unit::bias_compatible Setup Checking

unit::check_sweep Checking of Inputs

unit::check_sync Checking of Inputs

instr::cal_possible Calibration


instr::do_cal Calibration

During Measure

This operation undertakes a potentially complex series of operations on the instrumentsused by a Setup. The exact functions called vary, depending on whether calibration isavailable for particular instruments, and whether the main sweep instrument operates inan internally swept fashion, or in a stepped/spot-mode fashion (the case when theinstrument option Use User Sweep is set to Yes for the main sweep instrument).

The functions called during Measure are shown in the following table (user main sweep)


65

and Functions Called During Measure (with Internal Main Sweep) (internal main sweep).

Functions Called During Measure (with User Main Sweep)

Function Name Function Category Notes


instr::find_units Hardware Editor Operations

instr::set_found Hardware Editor Operations







instr::reset_instr_info Initialization



instr::reset_outptrs Initialization

unit::check_out Checking of Outputs

unit::can_do_second_sweep Control and Data Acquisition Functions


instr::recall_n_chk_calib Calibration

instr::init_instr Initialization

instr::zero_supplies Initialization

BEGIN BIAS LOOP Loop to END BIAS LOOP

unit::set_bias Control and Data Acquisition

unit::enable_output Initialization

unit::set_scalar Control and Data Acquisition

BEGIN USER MAIN SWEEPLOOP

Loop to END MAIN SWEEP

unit::wait_hold_time Control and Data Acquisition


unit::set_sync Control and Data Acquisition

unit::wait_delay_time Control and Data Acquisition

unit::get_data Control and Data Acquisition

unit::get_scalar_data Control and Data Acquisition

END MAIN SWEEP LOOP

END BIAS LOOP


unit::fill_outds Control and Data Acquisition Functions Called During Measure (with Internal Main Sweep)

Function Name Function Category Notes


instr::find_units Hardware Editor Operations

instr::set_found Hardware Editor Operations







instr::reset_instr_info Initialization



instr::reset_outptrs Initialization

unit::check_out Checking of Outputs

unit::can_do_second_sweep Control and Data Acquisition Functions


instr::recall_n_chk_calib Calibration

instr::init_instr Initialization


BEGIN BIAS LOOP Loop through END BIAS LOOP below


unit::enable_output Initialization

unit::set_scalar Control and Data Acquisition

unit::enable_sync Control and Data Acquisition

unit::set_2_internal_sweeps Initialization

unit::set_internal_sweep Initialization


unit::set_sync Control and Data Acquisition

unit::trigger Control and Data Acquisition

unit::wait_data_ready Control and Data Acquisition

unit::get_data Control and Data Acquisition

unit::get_scalar_data Control and Data Acquisition

END BIAS LOOP


unit::fill_outds Control and Data Acquisition

Handling Signals and Exceptions

A variety of conditions may result in termination of a measurement. Among the mostcommon exceptions for a driver is an I/O timeout. Timeouts usually occur when one ormore of the following conditions is present:

Instruments are turned offCabling is incorrectThe driver software makes errors with respect to instrument protocol or the timerequired by the instrument's operations

There are numerous examples in the hp4194 and cvu_4194 functions code thatdemonstrate setting the timeout before making different queries to the instrument. Atimeout is usually detected as a value of -1, returned from the spoll, readstring, orwritestring functions of the hpib_io_port used by the driver software.


66

In addition to instrument I/O problems, either of the following signals may be generated:

SIGFPE This signal occurs when the code executes an operation like a divide byzero. By default, there is no provision in IC-CAP for trapping this signal.

NoteIf this signal occurs during Measure, the default handling of SIGFPE terminates the measurement; ifit occurs during the execution of a transform, the function or macro will continue to execute andupon completion, an error message is displayed indicating a floating point error occurred.

SIGINT This signal is generated when you issue the Interrupt command. Bydefault, there is no provision in IC-CAP for trapping this signal. The measurement isterminated immediately. Note: For complex operations, it may take several minutesbefore control is returned.

If your application requires special error recovery for these signals, it is possible to trapthem. For details, refer to Handling Signals and Exceptions (extractionandprog) inTransforms and Functions (extractionandprog) .

NoteDo not alter the handling of SIGUSR1 and SIGUSR2; both signals are used internally by IC-CAP for errortrap and recovery purposes.


67

Prober Drivers in IC-CAPA prober driver is a set of USERC functions designed to control an IC wafer prober from anIC-CAP macro program. There are 3 types of probers (an initial call declares which type isin use) with these symbolic names:

EG1034X (ElectroGlas 1034X), EG2001X (ElectroGlas 2001X)APM3000A (and APM6000A and APM7000A) (TSK APM models)SUMMIT10K (Cascade SUMMIT 10000)

These probers share the same driver functions. External user functions and internal designfunctions, as well as prober settings and commands, are described in this section.

Additional TIS prober drivers required the renaming of native IC-CAP prober functionsfrom prober_xxxx() to icprober_xxxx(). This only affects systems where the prober.c filehas been customized in the OMI environment, and will not affect previously-writtenmacros. See the Readme file in $ICCAP_ROOT/src/README for information about thesefunctions.

To provide easier manipulation of a raw GPIB device file, IC-CAP offers a set of low-levelI/O functions named ice_hpib_xxxx. The declarations of these functions are found inicedil.h; their definitions are in icedil.c. Both files are provided as C source files. For moreinformation on these I/O functions, see icedil Functions (measurement). Driver functionsare contained in the directory $ICCAP_ROOT/src in the files shown in the following table.

Prober Driver Source Files

File Name Description

prober.h Prober call prototypes for userc.c

prober.c Actual code for each prober function

icedil.h Low level I/O call prototypes for prober.c

icedil.c Actual code for each ice_hpib_xxxx call

testprob.c Small interactive program to test the driver

run_testprob Properly sets your shared library lookup path and runs ./testprob if it exists, otherwise it runs$ICCAP_ROOT/bin/testprob. ICCAP_ROOT must be properly set in your environment forrun_testprob to work.

A custom driver can be added by editing prober.c in $ICCAP_ROOT/src and generating anew shared library file, libicuserc.<ext> (where ext is a platform-specific extension)because all prober drivers are written in C and treated as library functions.

For information on libicuserc, refer to Creating a New Shared Library (measurement).For details regarding adding library functions, refer to Creating C Language Functionsin IC-CAP (extractionandprog).

Source code is provided with this open interface. Recompilation and relinking arenecessary if this driver is user-modified.

External Prober User Functions

This section describes the external user functions.

Prober_debug This function takes 2 arguments and sets the internal flags. The firstargument defines the debug flag; when it is 1, all debugging information is displayed inthe Status window. The second argument defines the stop flag; when it is 1, the Macroexecution stops when an error is detected. After Prober_init(), the debug flag is off (0)and the stop flag is on (1). This function does not exist in TIS. This function alwaysreturns 0. An example call is:

x = Prober_debug (1, 0);

Every function looks at this internal debug flag and prints out any GPIB commands it isgoing to send, or a string it just received from the prober.

Prober_init This function must be called before any other prober calls are made in aMacro program. This function takes a GPIB address of the prober, flat orientation, probertype name, and a raw GPIB interface name to which the prober is connected. The flatorientation is usually 0, 90, 180, or 270. The function returns 0 when prober initializationis successful and -1 when it fails. The following table lists the GPIB configurationrecommended for HP 4062UX.

NoteA raw GPIB interface name is different for each platform. Refer to the following table for this name. Aseparate GPIB interface may be necessary if the given prober does not conform to IEEE 488 standard.

Standard HP 4062UX Configuration

Select Code Devices

7 or 27 Instruments and Switching Matrix

25 Wafer Prober

For a Sun SPARC computer, use the following call because a National Instruments GPIBcard has this name by default:

x = Prober_init (1, 0, "EG1034X", "/dev/gpib0");

For an HP 700 Series computer, use the following call because it involves a symbolic namerather than a GPIB interface filename:

x = Prober_init (1, 0, "EG1034X", "hpib");

This function also checks the prober type and sets the internal prober type flag forsubsequent driver calls. It closes its private unit descriptor from any previous proberaccess, opens the given GPIB interface file and keeps a new entity id. It then calls anappropriate subfunction, which does the prober-dependent initialization.

Prober_reset This function takes no arguments and sets the prober to Local mode. Itreturns 0 when successful and -1 when it fails. This function is not available for EG1034X(for which it is a no-operation). An example call is:

x = Prober_reset ();

This function clears the interface file and sends a selected device clear command to theprober.

Prober_status This function takes no arguments and returns 3 Real values in 1 array.


68

The first element of the array indicates whether the prober is Remote (1) or Local (0). Thesecond element indicates whether the edge contact is detected (1) or not (0). The lastelement indicates whether the Cassette is empty (1) or not (0). An example call is:

status = Prober_status ();

if (status[0] == 1) then ...

This function sends a query command to the prober and receives information aboutRemote/Local state as well as the edge sensor output. The Cassette Empty error isdetected in the function Phome and referred by this function, which keeps these statesand returns them back in an array of Real values.

Pdown This function takes no arguments and lowers the chuck of the wafer prober. Itreturns 0 when successful and -1 when it fails. An example call is:

x = Pdown ();

Phome This function takes no arguments and performs several tasks depending on theprober type. It returns 0 when successful and -1 when it fails. When it detects a CassetteEmpty error, it returns 1. An example call is:

x = Phome ();

This function calls a subfunction based on the prober type. A subfunction actually does theprober-dependent operation appropriate for Phome.

NoteSet the SUMMIT 10000 prober to Remote manually after this function to move the chuck to its Loadposition and turn the mode to Manual for wafer alignment.

Pimove Like Pmove, this function takes 2 arguments and moves the chuck relative to thecurrent position. It returns 0 when successful and -1 when it fails. An example call is:

x = Pimove (1, 0);

Pink This function takes 1 argument and triggers the specified inker. It returns 0 whensuccessful or -1 when it fails. EG1034X and SUMMIT10K probers do not support an inker,so this function is a no-operation for them. An example call is:

x = Pink (1);

Pmove This function takes 2 arguments and moves the chuck to the specified absolutecoordinates established by Pscale and Porig. The first argument specifies the new Xposition and the second specifies the new Y position. It returns 0 when successful and -1when it fails. An example call is:

x = Pmove (2, 4);

This function calculates how many machine units the chuck must move relative to thecurrent position, and sends an appropriate GPIB command to move the chuck. It alsoupdates its internal variables to keep track of the position.

Porig This function takes 2 numbers and defines these numbers as X and Y coordinatesof the current chuck position. This function must be called before any Pmove or Pimovefunctions. It always returns 0. An example call is:

x = Porig (0, 0);

This function stores the given numbers in its private variables.

Ppos This function takes no arguments and returns 2 Real values in an array, indicatingthe current die X and Y position being probed. The first element of the array is the Xcoordinate and the second is the Y coordinate. An example call is:

position = Ppos ();

print "X = "; position[0], "Y = "; position[1];

This function copies its private variables (which indicate the current position) and returnsthem.

Pscale This function takes the die X and Y dimensions in micrometers. These numbersare later used in Pmove, Porig, and Pimove functions. It always returns 0. An example callis:

x = Pscale (5000, 5000);

This function stores the given numbers in its private variables.

Pup This function takes no arguments and raises the chuck of the wafer prober so thatprobe pins come in contact with the wafer. It returns 0 when successful and -1 when itfails. An example call is:

x = Pup ();

Internal Prober Functions

Several internal functions support the user functions to customize the prober driver. Foreach algorithm, refer to the prober.c source file.

prober_get_err This function takes 1 argument and calls a subfunction depending onthe prober type. Each subfunction reads any error status from the prober. If it encountersan unknown error, it prints out the given number with an error message to the Statuswindow. It always returns 0. An example call is:

ret = prober_get_err(n);

prober_get_srq This function takes no arguments and returns 0 (no SRQ) or 1 (SRQ)depending on the SRQ line of the device file. An example call is:

ret = prober_get_srq();

prober_message This function takes 1 argument, a pointer to a string, and prints anerror message to the Status window such as <name>: unknown prober type, where <name> is replaced with the given string. An example call is:


69

ret = prober_message("Prober_reset");

prober_precheck This function takes no arguments and checks prober state such asRemote/Local and SRQ. It returns 0 when successful and -1 when it fails. An example callis:

ret = prober_precheck();

prober_response This function takes 1 argument that is either a pointer to a characterarray or null. It calls a subfunction depending on the prober type and each subfunctionreads any status information from the prober. Internal flags are set according to thestatus and any errors are reported. It returns 0 if there is no error. If a non-null pointer isgiven, a received string from the prober is returned using this pointer. An example call is:

char buffer[PSIZE];

ret = prober_response(buffer);

prober_spoll This function takes no arguments and performs serial polls in a prober-dependent way that may be different from the standard IEEE 488 implementation. Itreturns a status byte from the prober. An example call is:

ret = prober_spoll();

prober_wait_srq This function takes 1 argument that is a timeout value in seconds, andwaits for SRQ to be asserted. It returns 0 when SRQ is detected and -1 when a timeout orerror occurs. An example call is:

ret = prober_wait_srq(60.0); /* 60 sec */

Prober Settings and Commands

This section describes the correct IC-CAP wafer prober settings and their associated GPIBcommands.

EG1034X

This simple manual prober uses 2 settings. (Note that IC-CAP uses SRQ whereas HP4062UX does not.)

GPIB Address: AnySRQ Switch: Enabled

The following table lists the EG1034X GPIB commands. (Note that IC-CAP uses the MMcommand to move the chuck; HP 4062UX uses the MO commands for the EG1034X.)

EG1034X GPIB Commands

Item Command Reply Item Command Reply

Move Chuck MM MC Chuck Home HO MC

Chuck Up ZU MC Chuck Status ?S SZ...

Chuck Down ZD MC

EG2001X

This driver is tested with a prober software version called AC. The parameters listed in thefollowing table must be set to control this prober. Note that the I/O PROTOCOL is differentfrom the one for HP 4062UX. The Die Size is optional, but is included because IC-CAP doesnot set the size for manual operations.

EG2001X Settings

Parameter Value Parameter Value

METRIC/ENGLISH METRIC AUTO LOAD ENB if available

DIE X and Y SIZE Any AUTO ALIGN ENB if available

AUTO PROBER PAT. EXTERNAL AUTO PROFILE ENB if available

AUTO DIAMETER ENB MF/MC on X-Y ENB

Z-TRAVELING MODE EDGE-SEN MF/MC on Z DIS

I/O PROTOCOL ENHANCED MF/MC on OPT. ENB

I/O PORT GPIB-SP MF/MC onothers

DIS

GPIB ADDRESS Any

SRQ SWITCH ENB

The following table lists the EG2001X GPIB commands. Note that Chuck Home uses bothUL and LO commands (HP 4062UX uses LO).

EG2001X GPIB Commands


Move Chuck MM MC or MF Auto Profile PZ MC or MF

Chuck Up ZU Auto Align AA MC or MF

Chuck Down ZD Trigger Inker IK MC or MF

ChuckHome

UL/LO MC or MF Chuck Status ?S SZ...

APM3000A, APM6000A, APM7000A

This prober uses the following settings:

GPIB Address: AnyMode Switches: 3-4 OFF, 3-5 ON, 23-4 ON

The following table lists the commands.

APM3000A, APM6000A, and APM7000ACommands


Move Chuck A 65 CPU Halt T

Chuck Up Z 67 or73

TriggerInker

M 69

Chuck Down D 68 Chuck Home L 70 or 76

SUMMIT10K

This driver waits for an SRQ for an operation completed. With Summit Software version


70

2.10, the F10 key enables Remote mode. This prober uses the following settings.

COMMUNICATION PROTOCOL: GPIBCOMMAND SET: nativeDISP REMOTE CMDS: offBUS ADDRESS: anyTIMEOUT: 30.0CONTROL MODE: remote

SUSS PA 150, PA 200

The SUSS ProberBench Interface developed by Karl Suss for IC-CAP is provided as aconvenience, but is not supported by Agilent Technologies. The prober driver supports allfunctions described in External Prober User Functions and Internal Prober Functionsexcept prober_spoll (), prober_get_srq (), and prober_wait_srq (). In addition to theseIC-CAP functions, you can use the complete ProberBench command set (150 functions) toenhance operation. For information on these functions, refer to the ProberBench UserManual. For information on writing macros to control the prober, refer to Writing a Macro(measurement).

The SUSS PA 150 and PA 200 Semiautomatic Probers utilize a Microsoft Windows-baseduser interface running on an IBM-compatible PC. The IC-CAP environment communicateswith the prober via a macro over the IEEE 488 bus.

The required PC IEEE488 control hardware is: IOtech Personal488/AT.

The PC configuration must use the following values for the settings shown; all othersettings use default values:

Interface Type: GP488B

Name: IEEE

IEEE Bus Address: 22

System Controller: Off

Time-out (ms): 3000

Interface BusAddress:

02E1

DMA Channel: None

Interrupt: None

Prober Driver Test Program

This section describes the prober test program testprob, which is provided with C sourcecode. This program runs independently from IC-CAP and interactively calls driver functionsto test an Agilent-supplied or a custom driver.

The file for this program is located in $ICCAP_ROOT/src and is called testprob.c. Itincludes the test program main. The Makefile offers an option to build this test program.This program is linked with prober.o, iceswn.o, icedil.o, a GPIB library to exercise bothprober and switching matrix drivers. If testprob has been rebuilt with a custom driver, usean absolute path to specify the new testprob because $ICCAP_ROOT/bin has another,original testprob executable.

The testprob executable is an interactive program that gets user input from its stdin andcalls an appropriate driver function, then prints out the return value(s) of the driverfunction to the Status window.

The run_testprob script properly sets your shared library lookup path and runs ./testprobif it exists, otherwise it runs $ICCAP_ROOT/bin/testprob. Therefore, you should use therun_testprob script to run testprob. Make sure $ICCAP_ROOT is properly set in yourenvironment, then type run_testprob.

An actual prober (matrix) must be connected to a raw GPIB device file in order to performdriver (matrix) tests. Off-line testing is not available with this program.

This program expects to see a function name and its arguments as if they appeared in anIC-CAP Macro program. However an argument list cannot include another function, that is,nesting is not allowed.

A command example is:

Prober_init(1, 0, "EG1034X", "hpib")

The currently supported functions are shown next.

Connect Prober_debug

FNPort Prober_init

Pdown Prober_reset

Phome Prober_status

Pimove Pscale

Pink Pup

Pmove SWM_debug

Porig SWM_init

Ppos Wait

NoteAny line starting with # is treated as a comment and is ignored. A blank line is skipped (this is helpfulwhen a file is used to supply input to this program).

Because this test program is not a real Macro interpreter, it has the following restrictions:

No control constructsNo variablesNo nesting of functionsNo function library other than the prober and matrix driverNo capability to execute IC-CAP Macro programs.

Because nesting is not supported, the Connect function needs a port address such as32701 instead of FNPort(1). Refer to the HP 4062UX Programming Reference for moreinformation about port addresses.


71

Matrix Drivers in IC-CAPA matrix driver is a set of USERC functions designed to control the switching matricesthrough an HP 4084 controller from an IC-CAP Macro program. The matrix driver supportsthe matrices listed in the following table.

External user functions and internal design functions are described in this section. Theyare designed to be compatible with HP 4062UX TIS where possible.

Source files for this matrix driver are iceswm.h and iceswm.c. The header file iceswm.h isincluded in userc.c so that the function names can appear in the Function List of IC-CAP.

Source code is provided with this open interface.

Types of Matrix Drivers

Matrix Controller Pins Device

HP 4085A HP 4084A 48 HP 4062A and HP 4062B

HP 4085B HP 4084B 48 HP 4062C and HP 4062UX

HP 4089A HP 4084B 96 same as above, with 2 controllers

External Matrix Driver User Functions

This section describes the matrix driver external user functions.

SWM_debug This function takes 1 argument and sets the internal debug flag. When theargument is 1, debugging information is printed out to the Status window; when theargument is 0, printing is turned off. It always returns 0. This function does not exist inTIS. An example call is:

x = SWM_debug(1)

Every function looks at this flag and prints out any GPIB commands it is going to send, ora string it just received from a matrix controller.

SWM_init This function takes 2 GPIB addresses, a matrix name, and a raw GPIBinterface name to which the matrix is connected. The first GPIB address is for the block 1(usually 19) and the second is for the block 2 (22). However, a different address can beassigned for each matrix controller. For the HP 4085A and HP 4085B (both 48-pinsystems), the second address is used as the controller address, and the first address isignored. It returns 0 when successful and -1 when it fails. This function does not exist inTIS. An example call is:

x = SWM_init (19, 22, "HP4089A", "hpib"); ! for 96-pin

or

x = SWM_init (0, 22, "HP4085B", "hpib"); ! for 48-pin

This function checks the matrix type and sets the internal type flag for subsequent matrixcalls. It closes its private entity id from a previous matrix access (when it exists), andopens the given raw GPIB device file. Then it calls an internal function swm_init_unit toreset a controller. This clears all pins and ports.

Connect This function takes a port address and a pin number and connects the givenport to the pin. The port address is either 0 or from 32701 to 32711, inclusive. The pinnumber is 0 or from 1 to 48/96 inclusive. When a pin card does not exist for the given pinnumber, it gives an error message and aborts the Macro execution.

An example call is:

x = Connect(32701, 25);

This function sends GPIB commands to the matrix controller and either connects ordisconnects the specified port and pin. The following table lists argument combinations.

Argument Combinations

Port Address Pin Number Description

0 0 Disconnect all pins from all ports.

0 X Disconnect pin X from its connected port.

X 0 Disconnect all pins connected to port X.

X Y Connect port X to pin Y.

As in TIS, multiple pins can be connected to 1 port by calling this function several times.Pin numbers 1 through 48 belong to block 1; pin numbers 49 through 96 belong to block2. When a 96-pin matrix is used, do not connect block 1 and block 2 pins to 1 single port.Because this function does not include switching delay, allow enough wait time before andafter measurement to prevent relay damage. Virtual Front Panel (VFP) is not supported.

FNPort This function takes a port number and returns a port address for Connect. Thisallows compatibility with the HP 4062UX. An example call is:

port = FNPort(1);

Wait This function takes a wait time, in seconds, to give a necessary delay to wait untilSMU outputs become zero for dry switching. This function does not exist in TIS. It returns0 when successful or -1 when it fails. An example call is:

x = Wait (0.1) ! 100ms delay;

Internal Matrix Driver Prober Functions

The internal functions described next support the user functions. Refer to the source filefor each algorithm.

swm_connect_pin This function takes a GPIB address of a controller, a port number,and a pin number. It sends a Pin Connect command to the controller, and is called fromswm_connect (Connect) to actually perform the pin connection and disconnection.

swm_connect_port This function takes a GPIB address and a port number to send aPort Connect command to the controller, which manages input relays of an HP 4089Amatrix.

swm_cut_port_pin This function takes a GPIB address and a pin number to cut the pin


72

connection when a 96-pin matrix is used. It also checks if the port to which the pin wasconnected can be turned off; if it can (both Force and Guard are off), it turns off this port.

NoteWhen a switching matrix controller shares a single GPIB with other instruments, set the system variableINST_START_ADDR to a value greater than the matrix controller's GPIB address. This prevents IC-CAPfrom accessing the controller while performing Rebuild (instrument list).

swm_init_unit This function takes a file designator (or eid, a small integer usuallyobtained by calling the open system function) and a GPIB address of a matrix controller. Itis called from swm_init to initialize a controller and clear all pins and ports for which thecontroller is responsible.

swm_parse_err This function takes a status byte sent from a matrix controller anddetermines the cause of an SRQ. If there is no error, it returns 0 to allow the caller tokeep running. If there is an error, it returns -1 to abort the execution of the caller.

swm_release_port This function takes a GPIB address and a port number to send aPort Disconnect command to the controller only when a 96-pin matrix is used. Becauseblock 1 and block 2 pins should not be connected to a single port, a disconnect requestsuch as Connect(32701, 0) not only cuts the connection between a port and a pin, butalso disconnects the input relays of the port.


73

Driver ExamplesUsing IC-CAP with B2200A and B2201 Low-Leakage Mainframe Driver(measurement)Using IC-CAP with the HP 5250A Matrix Driver (measurement)Using IC-CAP with HP 4062UX and Prober and Matrix Drivers (measurement)

Using IC-CAP with B2200A/B2201 Low-Leakage Mainframe Driver

This section describes the transforms implemented for the B2200A/B2201 Low-LeakageMainframe Driver.

List of the transforms:

B2200_bias_card_enableB2200_bias_ch_enableB2200_bias_enableB2200_bias_initB2200_close_interfaceB2200_connectB2200_couple_enableB2200_couple_setupB2200_debugB2200_disconnect_cardB2200_GPIB_handlerB2200_ground_card_enableB2200_ground_enableB2200_ground_initB2200_ground_outch_enableB2200_ground_unused_inputsB2200_initB2200_open_interface

The following sections describe these transforms. For more details about the AgilentB2200A/B2201A, see its User Guide.

Utility Functions

B2200_debug

When set to 1, prints out all command strings sent to the instrument. This flag is commonto all B2200A's on the bus, regardless of their GPIB address.

B2200_debug(<flag>)

where:

<flag> is "1", "0", "Yes", or "No".

B2200_close_interface

Closes the current interface, which was opened by calling B2200_open_interface().

B2200_GPIB_handler

Returns -1 if the interface has not been initialized (invalid handler). Returns a positiveinteger (handler) if the interface has been opened.

Returns the current interface handler. The function is provided as a utility function, whichenables you to write advanced PEL code to write and read data to the B2200A using theHPIB_write and HPIB_read functions. Initializing the handler using B2200_open_interfaceenables you to use B2200A's built-in driver functions as well as writing PEL code tosupport other features that are not currently supported by the built-in functions, all in thesame PEL code.

Initialization and General Configuration

B2200_open_interface

Opens and initializes the GPIB interface and must be run first in the PEL program. Theinterface handler is saved in a static variable so that the interface will be shared by all theother B2200's function calls. You can drive multiple B2200 instruments as long as they areon the same interface bus (obviously, they must have different addresses).

B2200_open_interface(<Interface Name>)

where:

<Interface Name> is the name of the GPIB interface.

B2200_init

Must be run first in the PEL program to initialize the instrument and set the configurationmode. When the instrument is in AUTO configuration mode and multiple plug-in cards areinstalled in the B2200 slots from slot 1 continuously, the installed cards are then treatedas one card (numbered 0). This function resets all the settings to factory default beforesetting the configuration mode.

This function also sets the default connection rule for the specified card. When theconnection rule is FREE (default mode), each input port can be connected to multipleoutput ports and each output port can be connected to multiple input ports. When theconnection is SINGLE, each input port can be connected to only one output. Connectionsequence specifies the open/close sequence of the relays when changing from an existingconnection to a new connection.

B2200_init(<addr>,<cardNumber>,<config>,<connectionRule>,<connectionSequence>)

where:

<addr> is the GPIB address of the Mainframe (must be a positive number from1 to 30).<cardNumber> is 0(auto), 1, 2, 3, or 4.<config> is "AUTO" or "NORMAL" (string input).<connectionRule> is "FREE" or "SINGLE".<connectionSequence> is "NSEQ", "BBM", or "MBBR".

NSEQ (No SEQuence): Disconnect old route, connect new route.BBM (Break Before Make): Disconnect old route, wait, connect new route.


74

MBBR (Make Before BReak): Connect new route, wait, disconnect old route.

Transforms Governing the Bias Mode

B2200_bias_init

Selects the Input Bias Port for the specified card. The Input Bias Port is the dedicated biasport.

B2200_bias_init(<addr>, <CardNumber>, <InputBiasPort>)

where:

<addr> is the GPIB address of the Mainframe (must be a positive number from1 to 30).<CardNumber> is 0(auto), 1, 2, 3, or 4.<InputBiasPort> is 1 to 14 (numeric input) or -1 to disable bias port.

B2200_bias_ch_enable

This function bias-enables specific output ports in the channel list for the specified card.The input ports specified in the channel list are ignored since the input port is always theBias Input Port. By default, all the outputs are bias-enabled after a reset.

B2200_bias_ch_enable(<addr>,<CardNumber>,<State>,<Channel list>)

where:

<addr> is the GPIB address of the Mainframe (must be a positive number from1 to 30).<CardNumber> is 0(auto), 1, 2, 3, or 4.<State> is the output port's state (allowed values are "ENABLE", "DISABLE","E", or "D")<Channel list> is the list of channels, known as connection routes. Examplechannel list: (@10102, 10203, 10305:10307)

B2200_bias_card_enable

This function bias-enables all the output ports of the specified card. By default, all portsare bias-enabled after a reset.

B2200_bias_card_enable(<addr>, <CardNumber>, <CardState>)

where:

<addr> is the GPIB address of the Mainframe (must be a positive number from1 to 30).<CardNumber> is 0(auto), 1, 2, 3, or 4.<CardState> is the card output port's state (allowed values are "ENABLE","DISABLE", "E", or "D").

B2200_bias_enable

Enables the bias mode for the specified card once Input Bias Port and Enabled Outputports are specified. When Bias Mode is ON, the Input Bias Port is connected to all BiasEnabled output ports that are not connected to any other input ports. Bias Disabled outputports are never connected to an Input Bias Port when Bias Mode is ON.

If another input port is disconnected from a bias enabled output port, this port isautomatically connected to the Input Bias Port.

If another input port is connected to a Bias Enabled output port, the output port isautomatically disconnected from the Bias Input Port. When Bias Mode is OFF, the InputBias Port is the same as the other ports.

B2200_bias_enable(<addr>, <CardNumber>, <mode>)

where:

<addr> is the GPIB address of the Mainframe (must be a positive number from1 to 30).<CardNumber> is 0(auto), 1, 2, 3, or 4.<mode> is "On", "Off", "1", or "0".

Transforms Governing the Ground Mode

B2200_ground_init

Selects the input Ground Port for the specified card. For each card, you can specify thesame or a different Ground Port. By default, the input Ground Port is port 12. The groundport should be connected to 0 V output voltage. See the Agilent B2200 User's Guide fordetails.

B2200_ground_init(<addr>,<CardNumber>,<InputGroundPort>)

where:

<addr> is the GPIB address of the Mainframe (must be a positive number from1 to 30).<CardNumber> is 0(auto), 1, 2, 3, or 4.<InputGroundPort> is 1 to 14 (numeric input) or -1 to disable ground port.

B2200_ground_outch_enable

Ground-enables or ground-disables output ports. When Ground Mode is turned ON, theground-enabled output ports that have not been connected to any other input port areconnected to the input ground port. The input ports specified in channel lists are ignoredsince the input port is always the Input Ground Port. By default, all the outputs areground-disabled after a reset.

B2200_ground_outch_enable(<addr>,<CardNumber>,<State>,<Channel list>)

where:

<addr> is the GPIB address of the Mainframe (must be a positive number from1 to 30).<CardNumber> is 0(auto), 1, 2, 3, or 4.<State> is the port's state (allowed values are "ENABLE", "DISABLE", "E", or


75

"D").<Channel list> is the list of channels, known as connection routes. Examplechannel list: (@10102, 10203, 10305:10307)

B2200_ground_unused_inputs

Specifies the ground-enabled (or unused) input ports for the specified card. When GroundMode is turned ON, the ground-enabled input ports that have not been connected to anyother port are connected to the input Ground Port. By default, all the inputs are ground-disabled after a reset.

B2200_ground_unused_inputs(<addr>,<CardNumber>,<Input Channels>)

where:

<addr> is the GPIB address of the Mainframe (must be a positive number from1 to 30).<CardNumber> is 0(auto), 1, 2, 3, 4.<Input Channels> is the list of input channels (e.g., "1, 2, 5"). Only input ports1 to 8 can be defined as unused (these are the input Kelvin Ports).

B2200_ground_card_enable

Enables ground-enabling for all the output ports of the specified card. By default, all portsare ground-disabled.

B2200_ground_card_enable(<addr>,<CardNumber>,<CardState>)

where:

<addr> is the GPIB address of the Mainframe (must be a positive number from1 to 30).<CardNumber> is 0(auto), 1, 2, 3, or 4.<CardState> is the card output port's state (allowed values are "ENABLE","DISABLE", "E", or "D").

B2200_ground_enable

Enables the bias mode for the specified card. When Ground Mode is turned ON, the InputGround Port (default is 12) is connected to all the Ground Enabled input/output ports thathave not been connected to any other port. At Reset, Ground Mode is OFF. Ground Modecannot be turned ON when Bias Mode is ON.

See the Agilent B2200 User's Guide for additional comments and restrictions.

B2200_ground_enable(<addr>, <CardNumber>, <mode>)

where:

<addr> is the GPIB address of the Mainframe (must be a positive number from1 to 30).<CardNumber> is 0(auto), 1, 2, 3, 4.<mode> is "On", "Off", "1", or "0".

Transforms Governing the Couple Mode

B2200_couple_enable

Use this function to enable or disable Couple Port mode. Couple Port mode allowssynchronized connection of two adjacent input ports to two adjacent output ports.

B2200_couple_enable(<addr>, <CardNumber>, <Mode>)

where:

<addr> is the GPIB address of the Mainframe (must be a positive number from1 to 30).<CardNumber> is 0(auto), 1, 2, 3, or 4.<mode> is "On", "Off", "1", or "0".

B2200_couple_setup

Selects the couple ports for Kelvin connections. At Reset, no input ports are coupled.

B2200_couple_setup(<addr>,<CardNumber>,<ListOfCoupledPorts>)

where:

<addr> is the GPIB address of the Mainframe (must be a positive number from1 to 30).<CardNumber> is 0(auto), 1, 2, 3, or 4.<ListOfCoupledPorts> is the list of odd number input channels (e.g., "1, 3, 5"means coupled ports are 1-2, 3-4, 5-6).

Transforms Governing the Switching

B2200_connect

Connects or disconnects specified channels. Bias Mode and coupling Mode are also takeninto account when a channel is closed or opened.

For example, in the list (@10102, 10203:10205), the following channels are connected ordisconnected on card 1. Input port 1 to output port 2. Input port 2 to output port 3 and 5.

B2200_connect(<addr>,<Connect/Disconnect>,<ChannelList>)

where:

<addr> is the GPIB address of the Mainframe (must be a positive number from1 to 30).<Connect/Disconnect> is C or D.<ChannelList> is the list of connections to close.

B2200_disconnect_card

Opens all relays or channels in the specified cards.

B2200_disconnect_card(<addr>, <CardNumber>)


76

where:

<addr> is the GPIB address of the Mainframe (must be a positive number from1 to 30).<CardNumber> is 0(auto), 1, 2, 3, or 4.

Using IC-CAP with HP 4062UX and Prober/Matrix Drivers

This section describes how to use HP 4062UX instruments and the prober/matrix from IC-CAP for wafer device characterization. Also included in this section is information aboutwriting a macro, controlling the prober, and conditions of which to be aware.

While the HP 4062UX is an ideal instrument for performing device characterization withIC-CAP, it is necessary to understand IC-CAP, probers, matrices, and the instrumentsunder control. IC-CAP is an independent program from HP 4062UX TIS or VFP. It is notnecessary, and can be damaging, to run the START program before running IC-CAP. Torun IC-CAP after running the START program, the HP/Agilent 4142B must first be resetmanually.

After running the HP 4062UX START program, the HP/Agilent 4142B is put into its binarymode. Because IC-CAP assumes that all the instruments to which IC-CAP is connectedaccept ASCII commands, IC-CAP cannot recognize the 4142B. Reset the 4142B bysending a Device Clear or by turning the instrument off and on again. To send a DeviceClear to the 4142B, use the IC-CAP GPIB analyzer (Tools menu):

In the Instrument Setup Window, choose Tools > Send Byte.1.Enter the default value 20 and choose OK.2.

NoteExecute the START program to run TIS applications on the HP 4062UX, similar to a normal power-up.

Writing a Macro

While instruments like the HP/Agilent 4142B and the HP 4280A are controlled by IC-CAPwith Setup tables, both the wafer prober and the switching matrix must be controlledthrough macro programs using the Pxxxxx() and Connect() functions. The Setup tabledefines which measurement unit is going to force certain output. Users must perform thefollowing actions:

Determine which matrix port needs to be connected to which matrix pin.1.Write several Connect() functions in a macro program that invokes this Setup2.measurement with a iccap_func() statement.

The example shown in the following figure involves 4 SMUs of an HP/Agilent 4142B andmeasures Id_vs_Vg characteristics of an NMOS device on a wafer.

Sample Wafer Test Program

! Prober and Matrix Test Program

x = swm_init(19, 22, "HP4085B", "/dev/ice_raw_hpib")

x = connect(fnport(1), 15) ! SMU1 - Drain

x = connect(fnport(2), 7) ! SMU2 - Gate

x = connect(fnport(3), 8) ! SMU3 - Source

x = connect(fnport(4), 6) ! SMU4 - Bulk

x = prober_init(2, 0, "EG2001X", "/dev/ice_raw_hpib")

!

linput "Load Cassette and Press OK", msg

status = prober_status() ! wait until Remote

while (not status[0])

status = prober_status()

endwhile

iccap_func("/nmos2/large/idvg", "Display Plots")

!

x = pscale(8200, 8200) ! test chip die size

x = phome() ! goes to the first die

while (x == 0)

x = porig(0, 0) ! first die coordinates

i = 0

while (i < 5) ! test diagonal 5 dies

x = pdown()

x = pmove(i, i)

x = pup()

print

print "Die Position X=";i;" Y=";i;

iccap_func("/nmos2/large/idvg", "Measure")

iccap_func("/nmos2/large/idvg", "Extract")

i = i + 1

endwhile

x = phome() ! load next wafer

endwhile

if (x == 1) then linput "Cassette Empty. Test End.", msg

x = connect(0, 0) ! disconnect matrix pins

Prober Control

Prober control is determined by the number of test modules, which is either single ormultiple per die.

With the Pxxxx functions, it is assumed that there is a single test module on each die andevery test module exists in the same place relative to its die origin. In this case, it is easyto control the wafer prober.

The example wafer test program illustrated above shows the size of each die to be 8200μm × 8200 μm. The operator first indicates to the prober where the test module is on thefirst die. Once the prober is set to find this test module, Pmove() or Pimove() can step toany die and probe the same test module.

When there are multiple modules per die, every module position must be calculated inmicrons and Pscale(1,1) must be called. You must know each module position relative toits die origin, and each die position relative to its wafer origin. You must calculate thesenumbers to move the wafer chuck to its correct probing position.

Special Conditions

When using probers and matrices, be aware of the following conditions:

Interface File A dedicated GPIB interface for a prober is recommended to avoidunknown effects on other instruments. However, if the given prober conforms to the IEEE488 standard, it is possible to put the prober on the same GPIB with other instruments.Set the INST_START_ADDR system variable high enough to protect the prober from beingaccessed by the Rebuild (instrument list) operation.


77

Interrupt Both prober and matrix functions are simple C functions called from the Macrointerpreter of IC-CAP. It is possible to interrupt any one of these functions during its GPIBcommunication. Therefore, whenever you interrupt the execution of a Macro program thatinvolves prober or matrix control, it might be necessary to reset the bus. Prober_init()resets its interface bus to clear any pending GPIB communications with the prober.However, SWM_init() only sends a Selected Device Clear to the matrix controller. Ifnecessary, you can reset the measurement bus by choosing Tools > Interface > Reset.

Bus Lock The HP 4062UX can lock the measurement bus even when a TIS program isnot running. Be sure that the GPIB for measurement instruments is unlocked when IC-CAPstarts up. The easiest way to ensure this unlocked condition is to exit the HP BASICprocess from which any HP 4062UX program has been executed. IC-CAP also locks themeasurement bus only during a measurement, which is similar to "Implicit Locking" of theHP 4062UX.

Measurement Accuracy While the HP 4062UX performs certain error corrections for its48- and 96-pin matrices, IC-CAP does not know about these internal parameters.Therefore the capacitance measurement accuracy is not specified when IC-CAP measuresa capacitance through a switching matrix. However, performing a calibration at the matrixpins should reduce these errors introduced by the matrix.

NoteHCU and HVU are not supported by HP 4062UX. Do not use HCU or HVU with HP 4062UX because theiroutput range exceeds the maximum ratings of the switching matrix and may cause damage to theswitching matrix.

Using IC-CAP with the HP 5250A Matrix Driver

This section describes the transforms implemented for the HP 5250A Switching Matrix.

NoteThe old switching box transforms that were implemented for the HP 40XX series are not compatible withthe new ones. The instruments have different commands for switching and the 5250A has new featuressuch as BIAS and COUPLE modes, which were not available for the old 40XX series.

List of the transforms:

HP5250_debugHP5250_initHP5250_card_configHP5250_bias_initHP5250_bias_cardHP5250_bias_channelHP5250_bias_setmodeHP5250_couple_setupHP5250_couple_enableHP5250_connectHP5250_disconnect_cardHP5250_compensate_capHP5250_show()

The following sections describe these transforms. For more details about the HP 5250A,see its User Guide.

Utility Functions

HP5250_debug

This transform is only for debugging. When the debug flag is set to 1, all the functionsprint out all the command strings that are sent to the instruments. Set flag using thevalues 1 or 0, or use YES or NO.

HP5250_debug(<flag>)

HP5250_compensate_cap

This transform is the equivalent IC-CAP C routine for the HP BASIC capacitancecompensation routine called Ccompen_5250 supplied with the HP 5250A. It returns a 2 by1 matrix (2 rows, 1 column) defined as follows:

output.11 represents compensated capacitance data [F].output.21 represents compensated conductance data [S].

HP5250_compensate_cap (RawCap, RawCond, Freq, HPTriaxLenght,

UserTriaxLenghtHigh, UserTriaxLenghtLow, UserCoaxLenghtHigh,

UserCoaxLenghtLow)

where:

RawCap is Input Dataset containing raw capacitance data [F]RawCond is the Input Dataset containing raw conductance data [S]Freq is the measured frequency [Hz]HPTriaxLenght is the HP Triax Cable Length [m]UserTriaxLenghtHigh is the user Triax Cable Length (High) [m]UserTriaxLenghtLow is the user Triax Cable Length (Low) [m]UserCoaxLenghtHigh is the user Coax Cable Length (High) [m]UserCoaxLenghtLow is the user Coax Cable Length (Low) [m]

HP5250_show()

This transform has no inputs. It returns to the standard output (screen or file) thefollowing data about the instrument status:

Instrument NameInstrument Configuration (AUTO/NORMAL).

The following information is output for each card installed in the instrument (card 0 if theinstrument is in auto configuration mode):

Connection modeConnection sequenceInput Bias PortEnabled Output Bias PortsBias Sate (ON/OFF)Coupled Input Ports (only lower number is listed, e.g., "3,5" means ports 3 and 4 arecoupledCouple Port Mode (ON/OFF)Connection Matrix Inputs(10)xOutputs(12,24,36, or48). The following table shows an


78

output example of the Channel Matrix State where Card 1 is a 10x12 matrix switch. A"1" in a matrix cell means the connection is closed.Output Ports > 1 2 3 4 5 6 7 8 9 10 11 12

Input Ports 1 0 0 0 1 0 0 0 0 0 0 0 0

2 1 1 1 1 1 0 0 0 0 0 0 0

3 0 1 0 0 0 0 0 0 0 0 0 0

4 0 0 0 0 0 0 0 0 0 0 0 0

5 0 0 0 0 0 0 0 0 0 0 0 0

6 0 0 0 0 0 0 0 0 0 0 0 0

7 0 0 0 0 0 0 0 0 0 0 0 0

8 0 0 0 0 0 0 0 0 0 0 0 0

9 0 0 0 0 0 0 0 0 0 0 0 0

10 0 0 0 0 0 0 0 0 0 0 0 0

Initialization and General Configuration

HP5250_init

This transform must be run first to initialize the instrument with the address and interface.Using this transform the configuration mode can be set to AUTO. When the instrument isin AUTO configuration mode the same type of card must be installed in the HP 5250 slotsfrom slot 1 continuously. The installed cards are then treated as 1 card (numbered 0).

HP5250_init (BusAddress, "Interface", "Configuration")

where

BusAddress is interface bus address (default is 22)"Interface" is interface name (default is hpib)"Configuration" is AUTO/NORMAL A/N (default is NORMAL)

HP5250_card_config

This transform is used to change the default configuration for the specified card. When theconnection rule is FREE (default mode), each input port can be connected to multipleoutput ports and each output port can be connected to multiple input ports. When theconnection is SINGLE, each input port can be connected to only 1 output. Connectionsequence specifies the open/close sequence of the relays when changing from an existingconnection to a new connection.

HP5250_card_config (CardNumber, "ConnRule", "ConnSequence")

where

CardNumber specifies the card (0 for AUTO configuration mode)"ConnRule" is FREE/SINGLE (default is FREE)"ConnSequence" is NSEQ/BBM/MBBR (default is BBM)

NSEQ (No SEQuence): Disconnect old route, connect new route.BBM (Break Before Make): Disconnect old route, wait, connect new route.MBBR (Make Before BReak): Connect new route, wait, disconnect old route.

Transforms Governing the Bias Mode

HP5250_bias_init

This transform selects the bias port. When using the HP/Agilent E5255A card, the InputBias Port is the dedicated bias port; however, for the HP/Agilent E5252A the Input BiasPort must be selected using this function.

HP5250_bias_init(CardNumber, InputBiasPort)

where

Card Number specifies the card (allowed values 0-4, 0 = auto configurationmode)InputBiasPort specifies the input bias port number (allowed values are 1-10)

HP5250_bias_card

This transform bias-enables all the output ports for the specified card.

HP5250_bias_card(CardNumber, "CardState")

where

CardNumber specifies the card (allowed values 0-4, 0 = auto configurationmode)"CardState" is the card's state (allowed values are ENABLE/DISABLE or E/D)

HP5250_bias_channel

This transform bias-enables the specified output ports in the channel list. Note that theinput ports are ignored since the input port is always the Bias Input Port.

HP5250_bias_channel ("State", "Channel list")

where

"State" is the output port's state (allowed values are ENABLE/DISABLE or E/D)"Channel list" is the list of channels, known as connection routesExample channel list: (@10102, 10203, 10305:10307)

HP5250_bias_setmode

This transform enables the bias mode for the specified card once Input Bias Port andEnabled Output ports have been specified.

HP5250_bias_setmode (CardNumber, "BiasMode")

where

CardNumber specifies the card (allowed values 0-4, 0 = auto configurationmode)"BiasMode" sets the bias mode on or off (allowed values are ON/OFF or 1/0)


79

When Bias Mode is ON, the Input Bias Port is connected to all the Bias Enabled outputports that are not connected to any other input ports. Bias Disabled output ports arenever connected to an Input Bias Port when Bias Mode is ON.

If another input port is disconnected from a bias enabled output port, this port isautomatically connected to the Input Bias Port.If another input port is connected to a Bias Enabled output port, the output port isautomatically disconnected from the Bias Input port.

When Bias Mode is OFF, the Input Bias Port is the same as the other ports.

Transforms Governing the Couple Mode

HP5250_couple_setup

This transform sets up couple ports for making kelvin connections.

HP5250_couple_setup (CardNumber, "InputPorts")

where

CardNumber specifies the card (allowed values 0-4, 0 = auto configurationmode)"InputPorts" is the list of coupled portsExample: In the list "1,3,5,7,9" the coupled ports are 1-2, 3-4, 5-6, 7-8, 9-10

HP5250_couple_enable

This transform enables couple port mode. Couple port allows synchronized connection of 2adjacent input ports to 2 adjacent output ports.

HP5250_couple_enable (CardNumber, "CoupleState")

where

CardNumber specifies the card (allowed values 0-4, 0 = auto configurationmode)"CoupleState" is the coupled state (allowed values are ON/OFF or 1/0)

Transforms Governing the Switching

HP5250_connect

This transform connects or disconnects specified channels. Note that Bias Mode and/orcoupling Mode are also taken into account when a channel is closed or opened.

HP5250_connect ("Action", "Channel list")

where

"Action" connects or disconnects channels (allowed values are C and D)"Channel list" is the list of connection routes to be switchedExample: In the list (@10102, 10203:10205), the following channels areconnected or disconnected on card 1:Input port 1 to output port 2.Input port 2 to output port 3, 4, and 5.

HP5250_disconnect_card

This transform simply opens all relays or channels in the specified cards.

HP5250_disconnect_card (CardNumber)

where

CardNumber specifies the card (allowed values 0-4, 0 = auto configurationmode)


80

Handling Signals and Exceptions in Prober and MatrixDriversA variety of conditions may result in termination of a measurement. Among the mostcommon exceptions for a driver is an I/O timeout. Timeouts usually occur when one ormore of the following conditions is present:







If your application requires special error recovery for these signals, it is possible to trapthem. For details, refer to Handling Signals and Exceptions (extractionandprog) inTransforms and Functions (extractionandprog)


Instrument Drivers

A variety of conditions may result in termination of a measurement. Among the mostcommon exceptions for a driver is an I/O timeout. Timeouts usually occur when one ormore of the following conditions is present:







If your application requires special error recovery for these signals, it is possible to trapthem. For details, refer to Handling Signals and Exceptions (extractionandprog) inTransforms and Functions (extractionandprog)



81

Performing a MeasurementThe general procedure to perform a measurement in IC-CAP is:

Physically connect the hardware1.Specify an interface file for each GPIB card2.Build an active instrument list3.Assign unit names4.Use unit names in setups5.Specify instrument options6.Perform calibration7.Measure8.

This section describes the measurement procedure under the following topics:

A Typical IC-CAP Test SetupConnecting the HardwareVerify InstrumentsViewing Hardware SetupViewing Instrument LibrarySpecifying Interface FilesBuilding an Active Instrument ListAssigning Unit NamesUsing Unit Names in a SetupAdding a Ground UnitUsing Multiple InstrumentsSpecifying Instrument OptionsSaving Instrument OptionsCalibrationPerforming a MeasurementAborting a MeasurementViewing Measurement ResultsAccessing Data from a Previous MeasurementClearing Data from Memory

A Typical IC-CAP Test Setup

A typical IC-CAP test setup consists of a test fixture (such as the HP 16058A) on which theDevice Under Test (DUT) is mounted, along with any additional test circuitry.Measurement instruments are connected to the test fixture terminals using triaxial orcoaxial cables, and to the GPIB bus on the computer using a GPIB connector and a cable.

Notes

Hardware connections vary depending on the test fixture being used. For information on availabletest fixtures, refer to the operating manual for the specific instrument.For a list of supported instruments and configuration information, refer to Supported Instruments(measurement).

Connecting the Hardware

You can connect the hardware by making the appropriate GPIB connections between theinstruments and your computer. Connections must also be made between the individualunits of the instruments and the test fixture. The test device should be inserted into thetest fixture with the appropriate unit to device connections.

DC Connections

DC connections usually require connecting the SMUs and VS/VMUs of a DC Analyzer to thetest fixture in which the DUT is mounted. The test fixture can be the HP 16058A TestFixture, a probe station, or a switching matrix.

CV Connections

You can take the CV measurements using either an internal DC bias or an external DCbias. Usually, a DC analyzer is used to provide the external DC bias. When using internalor external biasing sources, note the hardware connection differences.

When using the DC bias internal to the CV meter, you only need to connect thecapacitance meter high and low units to the fixture in which the DUT is mounted. Theinternal DC bias is automatically available through this connection.

When using an external DC bias, this external source is used by making a connection tothe rear panel of the CV meter.

AC Connections

AC connections require connecting the 2-port units of a network analyzer to the fixture inwhich the DUT is mounted, as well as connecting a DC analyzer to provide a DC source.Connect the external DC source to the rear panel of the network analyzer.

Time Domain Connections

Time domain connections involve cabling between a generator, an oscilloscope, and theDUT. Another cable can be used to pass a trigger pulse from a generator to theoscilloscope's trigger input. For more information on these connections, refer to the54120.help text file provided in the iccap/lib directory.

Connecting a DC Analyzer and a Test Fixture

In the example procedure, the DC Analyzer contains two SMUs (Source/MeasurementUnits), each capable of sourcing and measuring voltage and current. For each SMU, asingle triaxial cable carries the input signal to the device and the output signal from thedevice.

To connect a DC Analyzer and a test fixture:

Connect a cable from each of the connectors (marked SMU1 - 2) at the back of the1.DC Analyzer to each of the connectors (marked SMU1 - 2) at the back of the testfixture.Insert the (DUT) into the test fixture and make the appropriate SMU to device lead2.connections.Connect a GPIB cable from the GPIB connector at the back of the DC Analyzer to the3.


82

GPIB bus connector on the computer.Ensure that the GPIB address is set to a value that does not conflicts with other GPIB4.addresses on the bus.Turn on the DC Analyzer.5.

Verify Instruments

You must configure IC-CAP so that it can recognize the system instruments on the GPIBand the individual source/monitor units (SMUs) in the measurement instruments. Thecomplete configuration is performed after an initial system installation or any time thesystem hardware is changed.

ImportantWhenever you change the device types, such as FET to BJT, you must rename the SMUs in theconfiguration.

A configuration file .icconfig containing hardware information and system level variablesis generated during the installation procedure and read when the program starts. Whenyou exit IC-CAP, the current configuration is saved in the .icconfig file in your homedirectory. For details, refer to Windows Installation (appendixa) or UNIX Installation(appendixa), as appropriate.

Viewing Hardware Setup

To view the hardware setup, from the IC-CAP Main Window, select Tools > Hardware

Setup or click Hardware Setup icon on the toolbar.

Viewing Instrument Library

The Instrument Library lists all the instruments for which IC-CAP drivers are provided. Youcannot edit an Instrument Library. For a list of supported instruments and configurationinformation, refer to Supported Instruments (measurement).

Specifying Interface Files

You must specify an interface file for each GPIB card being used for the measurement. Forinformation on configuring your interface, refer to the IC-CAP Customization andConfiguration (customization) sections. If your computer do not have access to aninterface file, and you need to specify instrument options in a Model file, you can add adummy interface. When you enter the name of an interface, begin the interface name witha prefix as dummy, such as dummy_gpib. IC-CAP interprets such type of interface as adummy interface.

To add an interface file:

In the IC-CAP/Main window, choose Tools > Hardware Setup or click Hardware1.

Setup. Click Add Interface.2.In the Add HP-IB Interface dialog box, type the name of the interface.3.

Click OK. The following screenshot displays a dummy GPIB interface added to the1.HB-IB Interface list and made the currently selected interface.


83

Building an Active Instrument List

The IC-CAP program must recognize the instruments that are physically connected to thesetup. When an instrument is added to the active list, the program identifies theinstrument by an instrument name, interface name, and GPIB address.

For example,an HP 4141 added to an active instrument list:

HP4141(gpib0, 23)

where,

HP4141 is the instrument name.

gpib0 is the gpib interface symbolic name.

23 is the HP 4141 gpib address.

If an instrument is powered up and connected to the GPIB bus, you can have the programadd it to the active instruments list automatically.

To add an instrument automatically in the IC-CAP/Main window, chosse Tools >Hardware Setup and click Rebuild.

All active instruments, with their respective addresses and interface name, are added tothe list. The status of the setup is displayed in the Status window.

NoteThe hardware displayed in the Instrument List may not reflect the physical instruments actuallyconnected. See adding instruments manually below.

Alternatively, you can add an instrument to the active instruments list manually, whetherthe instrument is physically connected to the system or not.

To add an instrument manually:

In the IC-CAP/Main window, choose Tools > Hardware Setup.1.Select the instrument in the Instrument Library list.2.Click Add to List.3.

NoteYou cannot manually add an HP/Agilent 4142 or HP/Agilent4155/6 instruments to the list of the activeInstruments List as the units of these instruments are configurable. The program finds the units theseinstruments have when you execute Rebuild Active List.

Certain IC-CAP error messages include the internal instrument or GPIB identifier. It ishelpful for you to understand the address syntax. The format of this internal id is:

INSTR_TYPE SELECTCODE ADDR or INSTR_TYPE.SELECTCODE.ADDR

where:

INSTR_TYPE is the instrument model number as listed in the Instrument Library.

SELECTCODE is the gpib interface's Logical Unit Number or Board number of theGPIB interface.

ADDR is the address (in decimal notation) of instrument, as set on theinstrument's address selector switch.

NoteOn Sun SPARC workstations, the selectcode is the GPIB board number and defaults to 1 for the first GPIBSBus or PCI GPIB board and 2 for the second. On Linux workstations, the selectcode is the Logical UnitNumber assigned to the GPIB interface. By default this is 7. See your GPIB interface documentation formore information.


84

Assigning Unit Names

When instruments are added to the active instrument list, the corresponding units areadded to the Unit Table. The Unit Table contains an entry for each active unit.

The information listed for each unit consists of a unit's physical name matched to theunit's user-defined name. By default the user-defined name is the same as the unit'sphysical name.

The unit names assigned by IC-CAP to the physical units are listed in the description ofindividual supported instruments (see Supported Instruments (measurement)). Forexample, the physical name of the first SMU of an HP 4145 appears in the Unit Table as

SMU1

The user-defined name defaults to the IC-CAP defined unit name:

SMU1

This value appears under the Unit Name column. All Unit Name fields can be edited. Youmay set this unit name to any name, but the user-defined name must be used whenspecifying the units in the Inputs and Outputs of a setup.

When a duplicate unit name is specified in the Unit Name field, a warning that a duplicatename exists is displayed on the window running the IC-CAP process. For example, if youhave two unit names called SMU1, the following warning is issued:

WARNING: Unit name <SMU1> used 2 times.

To assign the unit names:

In the IC-CAP/Main window, choose Tools > Hardware Setup.1.Select the instrument to configure in the Instrument List.2.Click Configure.3.Enter the new name(s) in the Unit Table and click OK.4.

Using Unit Names in a Setup

Each IC-CAP model includes setup specifications. The unit names assigned by IC-CAP tothe physical units are listed in the Unit field of the Input and Output tables displayed inthe Measure/Simulate folder.

IC-CAP must be able to recognize the instruments and their corresponding units, so theunit names in the hardware configuration must match the unit names assigned by theprogram.

For example, to take a CV measurement using the Capacitance Meter and an HP 4141 DCAnalyzer for an external DC bias, you specify the unit names CM and SMU1 in the Unitfields of the Setup. Since the CM unit is from the HP 4271 and the SMU1 unit is from theHP 4141, both the HP 4271 and HP 4141 Instrument Options tables are available for thissetup.

The options listed in the Instrument Options table vary for each instrument. Refer toConfiguring Hardware and Performing Measurement (measurement) for a list of allavailable instrument options, along with their descriptions, for each instrument supportedin IC-CAP.

To specify unit names in a setup:

In the Model window, select DUTs-Setups.1.Select the setup.2.Select Measure/Simulate.3.Select the Input or Output table.4.Click Edit.5.In the dialog box, edit the Unit Table as necessary to match the unit names specified6.in the hardware configuration.


85

Adding a Ground Unit

A ground unit, which does not appear in the Unit Table of the hardware configuration, canbe added to a setup. The ground unit is non-programmable and is available so that groundconnections can be specified in a setup to reflect actual physical connections to ground.

The ground unit name is case insensitive and can be entered as GND, GNDU, or GROUND inthe Unit field of the Input and Output tables. Do not assign these reserved names to anyof the units in the Unit Table. For example, entering the name GROUND into a Unit Namefield in the Unit Table causes a warning message to appear:

WARNING: 'GROUND' is a reserved Unit Name

NoteAlternatively, you can edit the Unit field directly in the Input or Output table.

Using Multiple Instruments

Some measurements may require more than one instrument. For example, a capacitancemeasurement using a CV meter and an external bias source requires a CV meter and a DCanalyzer. In this case, both the CV meter and the DC analyzer must be connected andrecognized by IC-CAP.

IC-CAP also allows measurements using multiple instruments of the same type. Forexample, you may perform a DC measurement using SMUs from two different HP/Agilent4142 instruments. To do this, the two 4142 instruments must be set to different GPIBaddresses, connected to the system, and recognized by IC-CAP. Assign unique unit namesto the SMUs and VS/VMs for each instrument since this unit name, which is entered in theSetup specification, determines the instrument and unit to be used to bias an input ormonitor an output.

NoteBecause an GPIB interface is locked by IC-CAP while making measurement and calibration, it is possible toshare a single GPIB interface with multiple users on an HP workstation. However, the GPIB interface on aSun workstation is not sharable since this interface does not offer a device locking mechanism.Simultaneous access of the GP-IB interface on a Sun workstation is not supported by IC-CAP.

Specifying Instrument Options

After the unit names are specified in the Input and Output tables of the setup, you canedit options for each instrument.

Measurement instruments use both internal (system) sweep and user sweep modes. For adescription of each mode and the instruments that support sweep modes, see SweepModes and Input/Output Types (measurement).

To view or edit instrument options:

In the Model window, select the DUTs-Setups folder.1.Select the setup.2.Select Instrument Options.3.Edit the option fields directly in the table by selecting the field and typing the new4.option.

Saving Instrument Options

You can save the instrument options in a file. The Save As command saves any activeoptions tables. If no active instrument of the same type is available, IC-CAP keeps theoptions table information in memory. All inactive options tables are cleared when an activeinstrument is added to the Instrument Setup or when a measurement or calibration ismade. The instrument options file is assigned the default suffix, .iot (Instrument OptionTables).

data

{

TABLE "HP4141.7.17"

{

element 0 "Use User Sweep" "No"

element 0 "Hold Time" " 0.000 "

element 0 "Delay Time" " 0.000 "

element 0 "Integ Time" "S"

element 0 "Init Command" ""

}

}

To save the instrument options to a file:

In the Model window, select the DUTs-Setups folder.1.Select the setup and Instrument Options.2.


86

Select File > Save As and choose (.iot) Instrument Options.3.Type a file name in the File Name field and choose OK.4.

Calibration

Most instruments have internal or hardware calibration capability. For example, CV metersand network analyzers have an internal calibration menu to perform appropriatecalibration for each test condition. Refer to Supported Instruments (measurement) forspecific information regarding instrument calibration.

NoteCalibration data for CV meters remains in memory until IC-CAP terminates. At start up, you must calibratea CV meter for the first measurement.

An HP 54120 series digital oscilloscope requires manual calibration on its front panel.Manual calibration means that an operator must be present to measure calibrationstandards. Refer to the description of this instrument family in this section for moreinformation on calibration.

Software calibration is also supported by IC-CAP for several instruments. For example, asimple offset error reduction is possible with the HP 4271. More elaborate 12-termcalibration is provided for all network analyzers.

One-port calibration is not supported for network analyzers because the 2-port conversionfunction used in extraction functions needs all four S-parameters. You can work aroundthis by measuring uncalibrated data and then performing 1-port calibration in PEL.

To perform a calibration of the applicable instruments being used in the Setup:

In the Model window, select the DUTs-Setups folder.1.Select the setup and Measure/Simulate.2.Click Calibrate.3.

Performing a Measurement

After you have entered the instrument configuration and have done the necessarycalibrations, you are ready to perform measurements.

To perform measurements for the active setup only:

In the Model window, select DUTs-Setups.1.Select the DUT and the setup.2.Select Measure/Simulate.3.

Click Measure Setup. 4.The system status line in the IC-CAP/Status window displays:Measure in progress...When the measurement is done, the status line displays:IC-CAP ReadyThe IC-CAP measurement is complete.

To perform measurements for all setups in the active DUT:

In the Model window, select DUTs-Setups.1.Select the DUT.2.

Choose Measure DUT. 3.The system status line in the IC-CAP/Status window displays:Measure in progress...When the measurement is done, the status line displays:IC-CAP ReadyThe IC-CAP measurement is complete.

To clear measured data for a selected setup:

In the Model window, select DUTs-Setups.1.Select the DUT and the setup.2.Select Measure/Simulate.3.Click Clear and choose the type of data to clear.4.

Aborting a Measurement

You can interrupt a measurement from the Status window. If you abort a measurementwhile an internal system sweep is in progress, the measurement in IC-CAP is aborted, butthe instrument continues to step through its sweep values until the sweep is completed. Ifanother IC-CAP measurement using this instrument is attempted before the sweep iscompleted, IC-CAP waits until the sweep is done before performing the measurement.

To abort a measurement:

In the IC-CAP/Status window, click Interrupt IC-CAP Activity.


87

You can use the Tools menu in the Hardware Setup window to control some measurementactivities. For example, you can stop an internal sweep by sending a command byte toinstruments on the bus.

To stop an internal sweep:

Open the Hardware Setup window.1.Select Tools > Send/Receive.2.Select Send Byte.3.In the dialog, enter the low-level GPIB command 4 to clear a single device or a 20 to4.clear all instruments.Click OK.5.

NoteDo not send SIGKILL to the IC-CAP process unless that is the only known way to abort a measurement.

Viewing Measurement Results

You can view results of both the measured and simulated data in a graphic display.Measured data is represented by solid lines and simulated data is represented by dashedlines. For details on viewing results, see Printing and Plotting (printandplot).

To view the results of the measurement:

In the Model window, select DUTs-Setups.1.Select the DUT and the setup.2.Select Plots.3.Click Display Plot or Display All.4.

Accessing Data from a Previous Measurement

If you wish to access data from a previously stored measurement or parameter extraction,you will need to perform this additional step to let IC-CAP know where to find the data andbring it into this model file:

From the IC-CAP Main menu, select File > Change Directory.1.Type in the path name of the file where your data is stored.2.Click OK.3.

Clearing Data from Memory

The Clear command allows you to clear from memory the data for the current setup. Youcan clear measured data, simulated data, or both. This command is useful when you havealready measured all setups of a DUT, or all DUTs that make up the model, and need tomake a change to one setup. You can clear the measured data for that setup and re-measure that setup.


88

A Sample Measurement ExampleThis example provides a general overview for performing a measurement. Using thesupplied model bjt_npn.mdl, the example measures the forward early voltage of an npnbipolar device using the SMUs of the HP 4141 DC Analyzer.

NoteBefore starting the measurement example, follow the procedures in Opening a Model File to open themodel bjt_npn.mdl.

Hardware Setup

The HP 4141 contains four SMUs (Source/Measurement Units), each capable of sourcingand measuring voltage and current. For each SMU, a single triaxial cable carries the inputsignal to the device and the output signal from the device.

To connect the hardware:

Connect a cable from each of the four connectors (marked SMU1 - SMU4) on the1.back of the HP 4141 to each of the four connectors (marked SMU1 - SMU4) on theback of the HP 16058 Test Fixture.Insert the bipolar npn test device into the HP 16058 Test Fixture and make the2.appropriate SMU to device lead connections.Connect an GPIB cable from the GPIB connector on the back of the HP 4141 to the3.GPIB bus connector on the computer.Make sure the GPIB address is set to a value that does not conflict with other GPIB4.addresses on the bus.Turn on the HP 4141.5.

To setup the hardware:

In the IC-CAP/Main window, click Hardware Setup.1.Select Add Interface.2.A dialog box opens. In the Name field, type the name of the interface, hpib.3.Choose OK.4.IC-CAP finds the device and adds HP 4141 to the Active Instruments list.

When the HP 4141 is added to the active instrument list, the corresponding units areadded to the Unit Table. The Unit Table contains an entry for each active unit.

To view unit names:

Select Configure. A dialog box opens, displaying the Unit Table and Instrument1.Address.For this example, the units of the HP 4141 are: HP4141.7.23.SMU1,HP4141.7.23.SMU2, HP4141.7.23.SMU3, HP4141.7.23.SMU4, HP4141.7.23.VS1,HP4141.7.23.VS2, HP4141.7.23.VM1, and HP4141.7.23.VM2.No changes are made. Choose Cancel.2.

Assigning Units to a Setup

The next step explains how to use these unit names in a setup to specify a particularmeasurement. The bjt_npn model already includes setup specifications for the inputs vb,vc, ve, and vs, and the output ic. The assigned unit names are entered in the Unit field ofthe Input and Output tables. By specifying the unit names, you assign the HP 4141 to thesetup fearly.

To specify unit names in a setup:

In the Model window, select DUTs-Setups.1.Select the DUT dc and the setup fearly.2.In the Measure/Simulate folder, select the Input table vc.3.Click Edit.4.In the dialog box, edit the Unit field by entering SMU1.5.Choose OK.6.Repeat steps 3 through 6, assigning SMU2 to the Unit field of the Input vb, SMU3 to7.the Unit field of the Input ve, and SMU4 to the Unit field of the Input vs. To monitorthe output, assign SMU1 to the Unit field of Output ic.

Specifying Instrument Options

The HP 4141 has four instrument options that may be set before taking a measurement.

To view instrument options:

Select Instrument Options.1.

For this example, the default values of these options are used.2.Since the option Use User Sweep is set to No, the main sweep Input vc (SweepOrder = 1) is swept from 0.000 to 5.000 volts in 21 steps using an internalinstrument sweep. When Use User Sweep is set to Yes, the measurement istaken from 0.000 to 5.000 volts in 21 steps, with each voltage being setseparately or point-by-point. A measurement taken with a user sweep is slowerthan the same measurement taken with an internal instrument sweep. However,the advantage for using a user sweep is increased flexibility in the types ofmeasurements that can be taken.The Hold Time option is set to 0.000. This means that the instrument waits0.000 seconds before starting the main sweep.The Delay Time option is set to 0.000. This means that the instrument waits0.000 seconds before the measurement is taken at each step in the sweep.The Integration Time is set to S. This means that the integration time of the HP4141 is short.

Since no changes are made, choose Cancel.3.

Measurement

To take a measurement:


89

Select Measure/Simulate1.Click Measure.2.The system status line in the IC-CAP/Status window displays:Measure in progress...When the measurement is done, the status line displays:IC-CAP ReadyThe IC-CAP measurement is complete.

Viewing Results

To view results:

Select Plots1.Click Display Plot. A plot of the measured data displays in a separate window.2.


90

Sweep Modes and Input/Output TypesMeasurement instruments use both internal (system) sweep and user sweep modes. Thissection describes each of these modes and the instruments that support sweep modes.For additional information on defining setups, refer to Simulation Types.

Sweep Types

The sweeping of a source for an instrument is controlled by the instrument or by IC-CAP.This applies only for the instrument to which the primary unit belongs. The primary unit isa unit with a sweep order 1. The instrument with the primary unit is called the primaryinstrument.

Internal (System) Sweep

A primary instrument can perform its internal sweep when the Use User Sweep option ofthat instrument is set to No. Some instruments, such as the HP 4271, cannot perform asweep measurement and do not have this option. A spot measurement with the internalsweep option enabled is converted to a single point measurement with the user sweep. Itis impossible to perform a single spot internal sweep. Internal sweep is much faster thanthe user sweep (described in the next section), but not all sweep types are supported bythe internal sweep of a particular instrument.

User Sweep

When the Use User Sweep is Yes for a primary instrument, IC-CAP performs a set of spotmeasurements to make up a single sweep measurement. Even though all supportedinstruments except time-domain instruments perform spot measurement, instruments likeNetwork Analyzers need to use the internal sweep for calibrated data. Most sweep typesare possible with user sweep because IC-CAP controls each point directly. However, a usersweep is much slower than an internal sweep.

Multiple Instruments

When multiple instruments are involved in a measurement setup, non-primary sweepinstruments use the user (spot-mode) sweep regardless of how the Use User Sweepoption is set. The sweep capabilities of the primary sweep instrument and the nature ofthe measurement determine whether internal or user sweep is appropriate.

When the primary instrument has internal sweep capability and other instruments are onlyused for non-primary sweeps or constants, the internal sweep for the primary instrumentis possible. This includes the case where a network analyzer sweeps its frequency as aprimary sweep while a DC bias is given as a secondary sweep from some DC instruments.

When multiple instruments have to synchronize at each measurement point, the usersweep must be used because these instruments don't know about each other. Only in thisfashion can IC-CAP control them properly. An example is to measure both S parametersand DC currents at each frequency point.

Supported Internal Sweeps

The following tables list the inputs, outputs, and internal sweeps that are possible witheach instrument, with the following exceptions:

Several time-domain pulse parameters can be extracted with Output T, likeRISETIME.The 8510A supports only LIN sweep.54120 Series includes HP 54121T/122T/123T.The 54122 does not support V-TDR Input, because the necessary pulse generator isnot available in this instrument.Two-port data is taken as S parameters, then converted by IC-CAP to otherparameters.The pulse generators have no measurement capability, thus no Output modes.

For more information, refer to the individual instrument descriptions in SupportedInstruments (measurement).

Internal Sweeps for DC and CV Instruments


91

InstrumentType:

DCAnalyzer>

CVMeter>

ModelNumber:

4140 4141/42/45 4155/56 4194 4271 4275 4280 4284/85 E4980A

InputMode:

V x x x x x x x

InputMode:

I x x x

InputMode:

F

InputMode:

T

InputType:

LIN x x x x x

InputType:

LOG x

InputType:

SYNC x x

InputType:

LIST

InputType:

CON x x x x x x x x

InputType:

TDR

OutputMode:

V x x

OutputMode:

I x x x

OutputMode:

C † x x x x x x

OutputMode:

G † x x x x x x

OutputMode:

R † x x

OutputMode:

X † x x

OutputMode:

SHYZKA

OutputMode:

T

†If Output Mode is... then Measurement is...

C Cp-Gp (only Cp read)

G Cp-Gp (only Gp read)

C, G Cp-Gp (both Cp and Gp read)

R Cs-Rs (only R read)

C, R Cs-Rs (both Cs and Rs read)

X (type Y) G-B (complex data of form G +j*B read)

X (type Z) R-X (complex data of form R + j*Xread)

Internal Sweeps for Noise Instruments

HP/Agilent 35670ADynamic SignalAnalyzer (Source)

HP/Agilent 35670ADynamic SignalAnalyzer (Channel)

Input Mode: V x

Input Mode: I

Input Mode: F x

Input Mode: T

Input Type: LIN x x

Input Type: LOG x

Input Type: SYNC

Input Type: LIST x

Input Type: CON

Input Type: TDR

Output Mode: V

Output Mode: I

Output Mode: C

Output Mode: G

Output Mode: SHYZKA

Output Mode: T x Internal Sweeps for Network Analyzers and Time-Domain Instruments

Instrument Type: NWA> Oscilloscope> Pulse Gen

Model Number: 3577 8510, 8702,8719, 8720,8722, 8753

54120 Series 54510 8130, 8131

Input Mode: V x

Input Mode: I

Input Mode: F x x

Input Mode: T x x

Input Type: LIN x x x x

Input Type: LOG x

Input Type: SYNC

Input Type: LIST x

Input Type: CON x x

Input Type: PULSE x x

Input Type: TDR x x

Output Mode: V x x

Output Mode: I

Output Mode: SHYZKA x x

Output Mode: T x x


92

Repetitive MeasurementsTo ensure the safest possible instrument operation, IC-CAP performs checking andinstrument initialization at the beginning of each measurement. When the samemeasurement is performed repeatedly, this checking and initialization is unnecessary.Repeated measurements, such as might be programmed within an IC-CAP Macro, can beaccelerated if only the first such measurement is subject to this checking and initialization.

Ann IC-CAP feature called Fast Measurements (measurement) improves the speed ofrepetitive measurements. The Fast Measurement techniques minimize the use of I/O andinstrument operations that are unnecessary when a measurement is repeated.

See AlsoFast Measurements (measurement)


93

Fast MeasurementsFast Measurement in IC-CAP allows you to eliminate a sizable amount of overheadassociated with an instrument setup and initialization. This feature can only be used whencertain criteria (listed below) are met. Although these set of conditions may seemsomewhat restrictive, they are necessary to provide reasonable reliability when benefitingfrom the flexibility of certain IC-CAP features, such as expression evaluation and theavailability of different Instrument Option values in different Setups.

Criteria for Fast Measurement

In order to enable Fast Measurement and ensure reliable operation, the following criteriamust be met before measuring:

Create a variable named MEASURE_FAST (a reserved variable name) and set itsvalue to Yes. Another variable, NO_ZEROING, is associated with a higher level ofoptimization and can be used for second level speedup.The preceding measurement must have succeeded so that complete instrumentinitialization has taken place.The setup being measured must be the same as in the preceding measurement.The Instrument Options values must generally evaluate to the same values as theydid during the preceding measurement.The functions offered by the Hardware Manager must not have been used since thepreceding measurement. For example, deleting an instrument from the Active Listdisables Fast Measurement for the first measurement that follows the deletion.

When these criteria are not met, IC-CAP reverts to its normal manner of completeinstrument initialization prior to each measurement. Note that Fast Measurement is notdisabled by a change in Input specification.

Enabling Fast Measurement

This section explains the steps necessary for enabling Fast Measurement and explains howto temporarily disable Fast Measurement and force IC-CAP to fully initialize instrumentsduring the next measurement. Two types of Fast Measurement are available: First Leveland Second Level.

To enable first level speedup:

Create an IC-CAP variable named MEASURE_FAST1.To perform Fast Measurement for a particular setup, create the variable at thesetup level.To perform Fast Measurement globally for all setups you repeatedly measure,create the variable at the system level.

Set the value of MEASURE_FAST to YES2.

The functions offered by the Hardware Manager must not have been used since thepreceding measurement. For example, deleting an instrument from the Active List disablesFast Measurement for the first measurement that follows this operation. The sectionRequesting Complete Initialization on the Next Measurement explains a simple,recommended way to use the Hardware Manager for ensuring the next measurementundertakes complete instrument initialization.

NoteMEASURE_FAST skips instrument initialization. This could be useful if an instrument is controlledadditionally with a macro program. For simple GPIB operations with library functions, refer to UsingTransforms and Functions." The Init Command instrument option is another method to send an arbitrarycontrol command per setup to an instrument.

Second level speedup applies only when the conditions necessary for first level are alsomet. Second level speedup does not yield speed improvements as substantial as thosefrom first level speedup.

To enable second level speedup:

Create an IC-CAP variable named MEASURE_FAST1.To perform Fast Measurement for a particular setup, create the variable at thesetup level;To perform Fast Measurement globally for all setups you repeatedly measure,create the variable at the top level.

Set the value of MEASURE_FAST to YES2.

Create an additional variable named NO_ZEROING3.

Set the value of NO_ZEROING to YES.4.


94

NoteSetting the value of NO_ZEROING to YES prevents the program from disabling or zeroing instrumentsprior to each measurement. As a safety measure, IC-CAP still ensures that each instrument involved in themeasurement ceases sourcing bias or other types of signals after the measurement concludes.

Forced Instrument Initialization

In some cases it might be inappropriate for IC-CAP to provide Fast Measurement, but itdoes so nonetheless. This section lists the cases where this can happen and explains howto temporarily override Fast Measurement and ensure complete instrument initialization.

IC-CAP may not detect certain instances of calibration that have been invalidated bychanged Input specification. Input specifications can be changed explicitly (with themouse and keyboard), or implicitly by a macro (such as, a macro that alters the values ofIC-CAP variables used within expressions in the Input editors). In such cases, when FastMeasurement has been requested, IC-CAP may proceed with the measurement withoutwarning about the changes in the Input specification.

IC-CAP always detects and downloads changes in Instrument Options settings. Changeshere result in warnings about invalid calibration when appropriate. When an Input orOutput is added to, or removed from a Setup, it might be appropriate for IC-CAP to fullyre-initialize the instruments used by that Setup, even if Fast Measurement is requested.However, IC-CAP will not do so, unless you use a method such as the one described in thenext section.

Requesting Complete Initialization on the Next Measurement

Here is an easy way to ensure that IC-CAP undertakes complete instrument initializationat the beginning of the next measurement. From the Hardware Manager menu, executethe function Disable Supplies. This function ensures that all instruments listed in the activelist will cease sourcing bias and other signals to the DUT.

You must request complete re-initialization with this method any time a setup has beenmodified in any way except when changing sweep end points. You should also use thismethod when you have altered an instrument's settings though any of the following:

The instrument's front panelIC-CAP's GPIB AnalyzerArbitrary instrument I/O in Programs or TransformsAdditional software programs

See AlsoRepetitive Measurements (measurement)


95

GPIB AnalyzerThe GPIB analyzer offers basic capabilities for communicating with instruments via theGPIB. It can be used to debug an instrument driver or to manually set an instrument to acertain state not supported by IC-CAP. The analyzer commands are found on the Toolsmenu in the Hardware Setup window; the output is displayed in the Status panel near thebottom of the window.

Menu Commands

Each of the menu items available in the Hardware Setup Window (quickstart) is describedin online help.

Macro Files

GPIB analyzer requests can be combined and placed in a macro file that can be executedat any time.

NoteIC-CAP macros for IC-CAP Models are different from macro files interpreted by the GPIB analyzer.

The GPIB analyzer's ability to interpret a file containing a series of requests is valuablefor:

Prototyping an instrument driver for testing a series of commands and checkinginstrument responses. (The GPIB analyzer macro facility includes some capabilitiesthat are not available in interactive use, such as serial polling until a particular bit isset in the response, or delaying for a fixed number of seconds.)Repeatedly manually executing a sequence of GPIB analyzer commands.

Macro File Example

This section provides an example of a acceptable GPIB analyzer macro file. The syntax ofeach line is very simple and the system can readily distinguish comments fromcommands. Note that expressions, accepted in IC-CAP macros, are not accepted here;most arguments are treated literally. These commands are typed in a text file using anytext editor, for example, vi.

The Macros submenu (from the Tools menu) provides 2 macro commands: choose Specifyto provide the name of the file to be read and executed; then, choose Execute. If changesare made to the file, it is necessary only to save the changes and again select Execute.

$c This is a small GPIB Analyzer macro file; this line is a comment

$a 17 active address = 17

$c send request for instrument ID string:

ID\n

$r read answer back

$p print it to the Status window

$c now send a string to reset the instrument:

RST\n

$w 2 wait 2 seconds after sending RST to instrument

$c The following '$m' command opens a dialog window, and asks:

$c 'Will now call othermacrofile; want to continue?'

$c At that point, the user can use the mouse to

$c cancel the execution of this macro or continue.

$m Will now call othermacrofile

$i /users/icuser/othermacrofile call another macro, like a subroutine

RST\n

$w 2 wait 2 seconds after sending RST to instrument

Macro Commands

The macro file contains 2 kinds of statements:

Literal strings to send to the instruments, such as in the Send String commandCommands and directives, such as set the active address or do a serial poll.

Commands and directives start with a dollar sign ( $). Descriptions of the availablecommands and directives are shown in the following table.

Commands and Directives

Command/Directive

Description

$c Indicates that the current line is purely a comment. Do not attempt to substitute anexclamation mark (!) to indicate a comment.

$r Read data into the GPIB analyzer's read buffer, as in the Receive String command. The resultis also copied onto the top-level IC-CAP system variable named HPIB_READ_STRING, if thisvariable has been defined by the user.

$a 2 Sets the active address to a literal integer value (2 in this case)

$w 3 Specifies the wait time, in this example, 3 seconds; if the optional argument is absent, adefault of 2 is used.

$p Prints the GPIB analyzer's read buffer, as in the Display String command

$m Displays a message panel for the user to indicate whether to stop or continue. The systemappends the phrase want to continue? to the characters that follow $m on the command line.Refer to Macro File Example.

$n Prints status to the status panel of the Instrument Setup window as the macro is executed, forexample: $n the macro has reset the instrument, and is about to download set points

$s Performs a serial poll of the active address. If an integer parameter is present, then it isconsidered a serial poll mask, and the program loops until ( <poll result> AND <integermask> ) is non-zero, that is, a desired bit is set. If the mask is negative, the loopingcontinues until a mask-specified bit is clear . For example, to loop until the serial poll responseat the active address has a 1 in bit 6, do this: $s 64To loop until the serial poll response at the active address has a 0 in bit 6, do this: $s -64

$i Calls or includes another file and execute the macros in it. This is like calling a subroutine; forexample, $i /users/icuser/macrofile

Macro File Syntax Rules

The following rules apply when writing GPIB analyzer macro files.

Macro command files are read by the GPIB bus analyzer and lines in the files elicitGPIB bus analyzer actions. Use only 1 action per line.Blank lines, or lines with only white space are ignored. In any line, leading whitespace is ignored.Some lines are sent to the instrument, others are commands or directives.If the dollar sign ( $) appears after optional leading white space, a line is considered


96

a directive or command. Otherwise, the first non-white and all subsequent charactersin the line are sent to the active address.Characters preceded by backslashes will first be converted to control characters orother characters. Use this for sending carriage-return, linefeed, or other terminators.Conversions are listed in the following table. ( \<any other single character> is reallya no-op; it causes the <any other single character> to be sent. If it is necessary tosend a line that starts with the dollar sign, it can be sent by preceding it with abackslash as shown in the following table.)

Control Characters in the GPIB analyzer

String in Macro File Character Sent to Instrument

\b backspace

\r CR

\n newline (linefeed)

\0 null

\f formfeed

\t tab

\v vertical tab

\ backslash

\$ dollar sign

\<any other single character> <any other single character>

Directives and commands have the dollar sign ($) character, a single commandcharacter that is not case-sensitive, and optional trailing arguments. White spacebetween the command character and the first argument is optional. Recall that anyother characters appearing on a line, after the directive and its arguments, areignored. This allows comments alongside directives if desired. For example, $a 17this sets the active address to 17 To stay consistent with other facilities in IC-CAP, and to keep GPIB analyzer macro files more readable, you may wish to adoptthe following style for end-of-line comments:

$a 17 ! this sets the active address to 17

However, do not use the exclamation mark (!) to associate an end-of-linecomment to a string sent to an instrument. The ! character and the rest of thecomment will be sent to the instrument:

RST\n ! OUCH. Not only RST<LF>, but all these other characters go out

also!


97

icedil FunctionsThe 'icedil' functions are used to control a raw GPIB device file from userc functions. Thesefunctions provide a thin layer on top of the HP DIL library to hide GPIB specific details,thus making it easier to use GPIB I/O as well as to port userc functions to non-HPmachines. The source files are written in plain C and are provided with a Makefile.

There are two icedil files in $ICCAP_ROOT/src:icedil.h Contains low level I/O call prototypesicedil.c Includes actual codes for each ice_hpib_xxxx call

Most icedil functions are named after corresponding DIL functions and return the samevalues. However there are some functions that do not correspond to any DIL functions.

DIL-related Functions

The following functions perform the same task as their corresponding DIL functions. Theyare provided to facilitate porting to non-HP machines. For descriptions of these functions,refer to your GPIB documentation.

ice_hpib_abortice_hpib_bus_statusice_hpib_eoi_ctlice_hpib_ren_ctlice_hpib_send_cmndice_hpib_spollice_hpib_status_wait

The following functions also perform the same task as their corresponding DIL functions,however the corresponding DIL functions are named as io_xxxx instead of hpib_xxxx.

ice_hpib_get_term_reasonice_hpib_eol_ctlice_hpib_lockice_hpib_unlockice_hpib_timeout

Other Functions

The following functions are translated either to system calls or to a set of low levelfunctions to make GPIB programming easier. The controller address on a GPIB is kept in astatic array hidden in icedil.o so that users of this library do not need to remember it.However users can read this address at any time using ice_hpib_get_address().

ice_hpib_bus_init This function takes an eid (entity id) and initializes the raw GPIB associated with thiseid as follows: EOI Enable EOL No EOL character Timeout 2 sec Remote Line Asserted

ice_hpib_open This function takes a device file name (full path name) and a flag (usually O_RDWR) tobe passed to open() system call. It opens this device file and initializes it by callingice_hpib_bus_init(). This function must be used prior to using a raw GPIB device file.

ice_hpib_read This takes an eid, a pointer to a character buffer, the maximum number of bytes toread, and an addrs. It sets myself a listener and the addressed device a talker, thenreceives the response from that device.

ice_hpib_write This function takes an eid, a pointer to a character buffer, the number of bytes to send,and addrs. It sets myself a talker and the addressed device a listener, then sends thespecified characters to that device.

ice_hpib_close This function takes an eid and closes the device file associated with this eid. Thisfunction is used to terminate communication with this device file.

ice_hpib_clear This function takes an eid and a GPIB address. It sends a Selected Device Clear to theGPIB device specified by the addrs.

ice_hpib_get_address This function takes an eid and returns my GPIB address on the device file associatedwith the eid. Usually this is 21. Note that currently no function is provided to changethis controller address.

ice_hpib_listen This function takes an eid and a GPIB address. It prepares the GPIB so that the devicespecified by addrs is a talker and myself is a listener. This function is called fromice_hpib_read().

ice_hpib_talk This function takes an eid and a GPIB address. It prepares the GPIB so that the devicespecified by addrs is a listener and myself is a talker. This function is called fromice_hpib_write().

ice_hpib_wait This takes one float and waits for the float second. This is implemented as a busy waitwith the gettimeofday() system call.

ice_hpib_strpos This takes two strings (string1, string2) and returns the position of the string2 instring1. This is similar to POS() function in HP BASIC.

ice_hpib_raw_read This takes an eid, a pointer to a character buffer, and a length. This reads up to thelength number of bytes from the device file specified by eid until it sees either an EOIor a terminal character. This function does not set up a talker/listener pair on the bus.

ice_hpib_raw_write This takes an eid, a pointer to a character buffer, and a length. This writes out thelength number of bytes from the buffer without setting up a talker/listener pair on thebus.

ice_hpib_check_eid This is a "static" function visible only inside of icedil.c file. This function returns 0 if thegiven eid is out of the valid range and non-zero for a valid eid.


98

Configured SystemsConfigured Systems represent multiple instruments pre-configured by Agilent to worktogether with IC-CAP as a single system to meet your modeling measurment needs.

Using IC-CAP with an Agilent 85122A Precision Modeling System (measurement)Using IC-CAP with an Agilent 85123A Device Modeling System (measurement)


99

Using IC-CAP with an Agilent 85122A PrecisionModeling SystemThis section documents procedures for setup, calibration, and control of the Agilent85122A precision modeling system. Some of these are hardware procedures independentof IC-CAP, and some are hardware control procedures performed inside IC-CAP.These procedures are common for the various high-frequency IC-CAP models. This is not asequential set of procedures, but a set of separate hardware-related procedures that areused in conjunction with any of the device modeling procedures documented in thechapters. The procedures in this appendix give detailed instructions for the following:

Configuring the system hardwarePerforming hardware setup in IC-CAPCalibrating the network analyzerDefining the instrument states in IC-CAP (for DC, capacitance, and S-parametermeasurements)

In some cases, you may not need to perform all these procedures each time you model adevice. The system setup procedures are performed rarely. On the other hand, thenetwork analyzer calibration must be performed regularly. Setting the instrument states isdone for each measurement setup in each model, and the settings can be stored.

Description of the System

The Agilent 85122A microwave parameter extraction system is specially designed for usewith Agilent 85190-series high-frequency IC-CAP software (and a compatible controller) tomeasure the DC and high-frequency performance of active devices. The software is thenused to extract the device model parameters. The standard system uses an Agilent 8510Cnetwork analyzer subsystem for S-parameter device characterization from 45 MHz to 20GHz, and an Agilent 4142B DC source/monitor to provide precision DC characterization aswell as bias for the S-parameter measurements.

The synthesized sweeper in the Agilent 8510 subsystem supplies RF signals. The Agilent4142B DC source/monitor provides DC force (supply) and sense (measure) capability fromits plug-in SMUs (source/monitor units). The DC signals are routed via cable feedthroughpanels to the Agilent 11612 option Kxx external bias networks. The RF signals are alsoconnected to the bias networks, and thus RF and DC signals are applied together to thedevice under test. The bias networks have 3.5 mm (female) connectors for interface to atest fixture or probe station.

The system can also be custom-configured to meet individual needs, to provide differentoperating frequency ranges, bias power levels, or cabling configurations, for example. Orit may include instrumentation for different types of measurements, such as power ornoise figure measurements. The system is factory-installed in a rack. A rack-mountedwork surface is included, for maximum convenience in making on-wafer, in-fixture, orcoaxial measurements. The work surface is coated with antistatic material and isconnected to chassis ground, therefore a static mat is not required.

The following figure shows the general configuration of a typical high-frequency modelingsystem, although the Agilent 85122A system is installed in a system rack.

General Configuration of a Typical System

Configuring the System Hardware

The Agilent 85122A system is shipped from the factory with all rack components andinstruments assembled and cabled in the system rack cabinet, and installed at your siteby the Agilent customer engineer.System hardware configuration need only be done in the following circumstances:

At initial system setup.If changes are made to the system hardware.To change the bias connections for a different device type.

Before starting a parameter extraction procedure, make sure the GPIB and Agilent 8510system cables are connected as shown in Figure: GPIB and System Interconnections andGPIB Addresses, with the GPIB addresses set correctly. Note that the synthesizer and theDC source/monitor both use address 19: this is not a conflict, because the synthesizer ison the Agilent 8510 system bus and the DC source/monitor is on the GPIB. Make sure thesynthesizer language is set correctly: the first three switch positions from the left on thesynthesizer's GPIB/language switch must be set to 001 for network analyzer language.

GPIB and System Interconnections and GPIB Addresses


100

Make sure the system RF cables are connected as shown in following figure:

RF Cabling Interconnections

If you are using a special custom system, refer to its installation and user's guide forcustom cabling diagrams. The standard and special option Agilent 85122A installation andthe IC-CAP user manual also provide any additional information you may need aboutsetting up your system components. Connect the RF cables from the test set ports to theRF IN ports of the bias networks.The following provides details of the DC bias connections:

For additional hardware information, refer also to the installation and user's guide foryour Agilent 85122A microwave parameter extraction test system, and to themanuals for the individual instruments in the system.For information on fixtures or probe stations, refer to the manufacturer'sdocumentation.

DC Bias Safety Considerations

Bias current and voltage are supplied to a device under test from the plug-insource/monitor units (SMUs) in the Agilent 4142B DC source/monitor, connected to thedevice through the bias networks and probes or fixture.

Interlock Connection

To prevent electric shock from DC voltages in excess of +/−42V, do not close the INTLK(interlock) connection of the Agilent 4142B DC source/monitor. The high-power SMUoutput can be as high as +/−200V, and the medium-power SMU output can be as high as+/−100V. As long as the INTLK connection is open, the DC voltage is clamped at +/−42VDC.

Floating Ground

IC-CAP measurements are normally performed with the device in a floating-groundconfiguration, to prevent ground loop noise or, in the case of a BJT, possible damage to adevice. A floating-ground configuration is accomplished by insulating the fixture or probestation from power-line ground, for example with an insulator between the wafer undertest and the chuck of a probe station. If you are measuring in a floating-groundconfiguration, make sure the shorting bar of the Agilent 4142B is connected between theCIRCUIT COMMON and CHASSIS GROUND terminals on the front of the GNDU plug-in.If you cannot implement a floating-ground configuration, it may be necessary to open theAgilent 4142B shorting bar and connect the DUT ground to the circuit common grounds atthe DUT ends of the SMU and GNDU cables. The circuit common grounds are notconnected through the bias networks: they are available at the input jacks.

A potential shock hazard exists when the shorting bar is disconnected for floating measurements. Do nottouch any of the Agilent 4142B front panel connectors while a floating-ground measurement is inprogress.

Configuring the SMUs and Bias Networks


101

Bias current and voltage are supplied to a device under test from plug-in source/monitorunits (SMUs) in the Agilent 4142 DC source/monitor, connected to the device through thebias networks and probes or fixture. The standard Agilent 85122 system includes onemedium-power SMU and one high-power SMU. The quadraxial cables from the SMUs arerouted through a cable feed-through panel in the system rack, around the back of thesynthesizer and test set, and out to the front of the rack through another feed-throughpanel.

Bias Connections

The following figure illustrates the connections from the SMUs to the bias networks.Connect the FORCE and SENSE connectors on the quadraxial cables to the FORCE andSENSE connectors on the bias networks. Connect the triaxial cable from the GNDU SMU tothe ground connector on one of the bias networks (usually port 2). Leave the groundconnector of the other bias network unconnected.

Connections from DC Source/Monitor to Bias Networks

NoteFor in-fixture measurements, the measurement procedures in this manual apply to a FET mounted in thefixture in a common-source configuration, or a BJT mounted in a common-emitter configuration.

Connect the DC/RF signals from the outputs of the bias networks via the semi-rigid cablesto the corresponding terminals of the fixture or probe station. Set up the angle of thecables and bias modules to minimize stress and torque, and ensure that the cables areproperly supported. Iterate the positioning of the cables and bias networks until theconnectors line up correctly. Use a torque wrench to tighten the connections.The following figure illustrates the connections for a BJT. The connections for a FET arecomparable, with the port 1 DC/RF signal connected to the gate, the port 2 DC/RF signalconnected to the drain, and the source grounded.

Connections from Bias Networks to In-Fixture DUT

Switching on the System

It is important to apply power to the system in the correct sequence, so that the networkanalyzer recognizes the test set and synthesizer, and the computer recognizes thenetwork analyzer subsystem and the DC source/monitor. Switch on power to the systeminstruments in this sequence (from bottom to top of the system rack):

DC source/monitor1.Synthesizer2.Test set3.Network analyzer4.Computer5.

Preparing for Calibration

Let the system warm up for at least 1 hour before calibrating the network analyzer. TheDC measurements can be performed during this time. The calibration standards will needto be at ambient roomtemperature when the measurement calibration is performed. If you are using a fixture,open the fixture calibration kit and place all the devices on top of the foam so they willreach room temperature by the time the system is warmed up. If you are using a probestation, place the probed impedance standard substrate (ISS) on the probe chuck andallow it to reach ambient temperature before use.The calibration procedure is described in Calibrating the Network Analyzer.

Terminating the Input Ports to Prevent Oscillation

Leave the network analyzer disconnected from the bias networks while you perform theDC measurements. Connect terminations with >20 dB return loss (such as 10 dB pads) tothe RF IN connectors of the bias networks, to prevent possible bias oscillation. Thenetwork analyzer is calibrated later in the modeling procedures, immediately before thecapacitance and S-parameter measurements. Alternatively, isolate the device from RFpower by manually setting 90 dB of attenuation at the network analyzer test ports. Alsoput the network analyzer in hold sweep mode by pressing STIMULUS > MENU > MORE> HOLD.

Installing the Device


102

The best time to install the device is immediately before making measurements. Installthe device carefully on the probe chuck or in the fixture. Be sure to handle the device aslittle as possible, to avoid damaging it. Once the device is in place, avoid bumping the teststation.

CautionGround yourself with an antistatic wrist strap to reduce the chance of electrostatic discharge. A groundterminal is provided on the lower left corner of the test set front panel for your convenience. (Note thatthe system work surface is coated with antistatic material and is connected to chassis ground, therefore astatic mat is not required.)

Install the device carefully on the probe chuck or in the fixture. Be sure to handle thedevice as little as possible, to avoid damaging it. Once the device is in place, avoidbumping the teststation.

NoteEliminate direct light sources (such as microscope light) during measurement, especially for GaAs devices,as light may influence the measurement and cause inaccurate results.

Performing Hardware Setup in IC-CAP

The measurement system hardware is controlled by the IC-CAP software. This procedureconfigures IC-CAP to recognize the system instruments on the GPIB, and the individualsource/monitor units (SMUs) in the DC source/monitor. The complete procedure need onlybe done after initial system installation, or any time the system hardware is changed, toestablish a system handshake. The second part of the procedure (renaming the SMUs)also needs to be done any time you change device types, such as from FET to BJT or fromBJT to FET. For help with launching IC-CAP and opening the model file youneed, refer to Chapter A, "Agilent 85190A IC-CAP," then continue with these steps:

In the IC-CAP Main menu bar select Tools > Hardware Setup. The hardware window1.is displayed. It will vary depending on the system hardware used, but will be similarto the following figure.

Under the Instrument List select Rebuild. This command causes IC-CAP to poll all2.available GPIB addresses and add to the Instrument List any newly connectedinstruments that are powered up. (It does not affect instruments already on the list,whether they are powered up or not.)

Renaming the SMUs

Use this procedure any time you change device types, such as from FET to BJT or fromBJT to FET, so that the measurement setups can properly communicate with the biassupplies. When you run the IC-CAP software, it initially identifies the plug-in SMUsaccording to the numbers of the Agilent 4142 slots in which they are installed. A medium-power SMU occupies one slot. A high-power SMU occupies two slots, and is identified bythe higher slot number of the two. An Agilent 85122 system can include differentcombinations of SMUs: the standard system used in the example procedures includes twoSMUs, one Agilent 41421B medium-power SMU and one Agilent 41420A high-power SMU.The medium-power SMU is factory-installed in slot 1 and is initially identified in the IC-CAPsoftware asMPSMU1. The high-power SMU is installed in slots 2 and 3, initially identified as HPSMU3.The ground unit is identified as GNDU. It is generally more convenient to assign unitnames to the SMUs that identify their purpose in a device measurement, depending on thebias connections at the device terminals and the type of device you are modeling. Inconfiguring IC-CAP to recognize the system hardware, you need to set the SMU namesused in software to identify the SMU connections at the measurement terminals. In a FETmeasurement, the SMUs are identified as VG (gate supply) and VD (drain supply). In aBJT measurement, the SMUs are identified as SMU1 and SMU2, corresponding to thenetwork analyzer test ports: this helps to facilitate measurement of a BJT with only twoSMUs, since the port 2 SMU will be used for measurements at both the collector andemitter terminals.The following figure illustrates the SMU unit names for a FET measurement and thecorrespondence to the device terminal connections.

SMU Unit Names for a FET Measurement

The following figure illustrates the SMU unit names for a BJT measurement and thecorrespondence to the device terminal connections.

SMU Unit Names for a BJT Measurement


103

In the Instrument List click on HP4142 to highlight it. Select Configure and the1.Configuration of HP4142 window is displayed.The Unit Table lists the SMUs. Also listed for each is a unit ID (identification) name2.that can be edited. In a FET measurement, the medium-power MPSMU1 is renamedVG, because it is connected to the gate terminal; and the high-power HPSMU3 isrenamed VD, because it is connected to the drain terminal. In a BJT measurement,the medium-power MPSMU1 is renamed SMU1 because it is the port 1 SMU; and thehigh-power HPSMU3 is renamed SMU2 because it is the port 2 SMU.To make this change, click left with the mouse and move the mouse pointer over the3.name assigned to MPSMU1 to highlight it. Then type in the new name, and pressReturn. Similarly, change MPSMU3's assigned unit name. Then select OK to close theconfiguration window.You can close the Unit Table tile by clicking middle again.4.To save this hardware configuration in a file for future use, select Hardware, then5.Write to File. A file filter is displayed, showing the pathname of the currentdirectory. Type in anappropriate name for the .hdw file you are creating, and select OK.Close the hardware window.6.After IC-CAP is configured to recognize the system hardware, you can proceed to oneof the modeling chapters.

Calibrating the Network Analyzer

The network analyzer must be calibrated before any S-parameter measurements areperformed. The system must be allowed to warm up for at least 1 hour before calibration,and the calibration standards must be at ambient room temperature. It is convenient toswitch on the analyzer and expose the calibration devices to air before you make the DCmeasurements. Good calibration of the network analyzer system is critical to a good high-frequency measurement and extraction. A good calibration is dependent on the quality ofthe calibration kit standard devices, the care with which they are maintained, and thecorrectness and repeatability of the device connections. A stable ambient temperature(±1°C) is required to maintain a good calibration. For the high-frequency IC-CAPmeasurement procedures, the network analyzer is calibrated over a broadband frequencyrange, to perform the S-parameter measurements for parasitic and AC extractions. Inaddition, one or two subset calibrations may be needed at CW frequencies for specificmeasurement setups. This procedure explains how to set up both the broadbandcalibration and one of the CW calibrations with one set of standards measurements,making the CW frequency cal a subset of the broadband frequency cal. If a thirdcalibration is done, its frequency may not be known until some of the S-parametermeasurements have been made, therefore it cannot be performed now. However, whenthe calibration frequency is determined later, the subset cal can be performed usingexactly the method defined under CW Frequency Calibration Subset without another set ofstandards measurements. The instructions given below are very abbreviated: moredetailed information is available in the Agilent 8510 Operatingand Programming Manual.

NoteIt is important that values entered in software correspond with actual instrument settings. Be sure to writedown the parameters of your calibrations, so that they are readily available when you need to enter themin IC-CAP.

To connect the network analyzer:

Disconnect the terminations from the bias networks. Connect the RF cables from thenetwork analyzer test ports to the RF inputs of the bias networks.

Swept Broadband Calibration

On the network analyzer, press PRESET.1.To define the frequency range of the calibration and measurement, on the network2.analyzer press STIMULUS > START, and use the numerical keypad to set the startfrequency, ending with one of the terminator keys (such as G/n) at the right of thekeypad. Similarly, press STOP and set the stop frequency. Set a frequency range atleast as wide as the operating range of the device.Press STIMULUS > MENU > NUMBER of POINTS, and set the number of data3.points you wish to measure. A reasonable number might be 51 or 101 points.Press PRIOR MENU > FREQUENCY > STEP to set the stepped frequency mode,4.which provides greatest measurement accuracy by phase-locking the synthesizedsweeper at each frequency point.Press PRIOR MENU > MORE > CONTINUAL to set a continual stimulus sweep.5.Press PRIOR MENU > POWER MENU, and set the desired power level. In setting6.the source power and attenuation, take care that the power level will not beexcessive at the device input. Also consider the gain of the device, and set a powerlevel that will not saturate the input port samplers of the analyzer. If the power levelat the sampler goes above −8 dBm, an IF OVERLOAD error message is displayed andyou will need to reduce the source output power. The default network analyzer powerlevel is 0 dBm.For a device with power dropoff at higher frequencies, you may wish to set a power7.slope using the stimulus menus. An appropriate power slope would be in the regionof 2-3 dB/GHz.Press RESPONSE > MENU > AVERAGING ON/restart, and enter an averaging8.factor high enough to reduce trace noise and increase dynamic range as appropriatefor your device measurements. The default averaging factor is 256, but as little as 16may be adequate.If the cal kit constants for your calibration kit are not loaded into the network9.analyzer, load them from disk now.If you wish to modify one of the internal calibration kit definitions (see below), do so10.now.Perform a full two-port calibration using short-open-load-thru standards. Omit11.isolation cal. (A TRL or LRM calibration would also be an appropriate alternative.)At the end of the calibration sequence, the cal set numbers are listed in the softkey12.menu. Press CAL SET #1 to store this calibration in cal set #1. (Or use anotheravailable cal set.)

CW Frequency Calibration Subset

In some of the high-frequency models, extraction of certain parameters is done at a singleCW frequency. The CW frequency must be equal to the start frequency plus an integermultiple of the step size: the model chapter you use gives more information on decidingan appropriate frequency. Once you have determined the frequency of the calibratedmeasurement point, use the following procedure to create a CW frequency subset of theswept broadband calibration.

Press CAL > CORRECTION ON > CAL SET 1 (this turns on the swept cal you just


104

1.completed), then MORE > MODIFY CAL SET > FREQUENCY SUBSET.Use the SUBSET: START and SUBSET: STOP softkeys in the SUBSET menu (NOT2.the front panel keys) to set both start and stop to the frequency you selected.Press CREATE & SAVE, and store this calibration in cal set #2 (or another available3.cal set).

After the network analyzer is calibrated, return to the modeling section you are using, forthe S-parameter measurement procedures.

Modifying a Cal Kit Definition

The network analyzer's internal calibration kit models are accurate for most applications,however, you may want to modify the internal cal kit definitions in certain situations, forexample:

If you are using a probe station and you know the probe capacitance, you can enter1.them into the cal kit definition so that their effects can be removed by themeasurement calibration.The impedance of an on-wafer load may not be exactly 50Ω. You can characterize the2.DC impedance of a load exactly with a precision ohmmeter measurement, and enterthe impedance value into the cal kit definition.

The procedure to modify an internal cal kit definition is complex, and not to be undertakenwithout a solid understanding of error correction and the system error model. It isexplained in Product Note 8510-5, Specifying Calibration Standards for the Agilent 8510Network Analyzer (Agilent literature number 5954-1559), and in the Agilent 8510COperating and Programming Manual.

More on Power Levels and Capacitance Measurements

The primary factor to consider in understanding the available range of power is the S-parameter test set, and whether it is coupler-based or power-splitter-based. The test setgenerally used in a 26.5 GHz system is the Agilent 8515B, which uses a power splitter tocouple signals to the detector. The "economy" version test set is the Agilent 8514B (20GHz), which uses a directional coupler. The directional coupler in the Agilent 8514B has a−20 dB/decade power rolloff below 500 MHz, with some signal loss above that. The resultis that in the frequency range 45 MHz to 2 GHz the minimum detectable signal of theAgilent 8514B is only −66 dBm. However, over the same frequency range, the minimumdetectable signal of the Agilent 8515B is −101 dBm. (Above 2 GHz, both test sets candetect at least −95 dBm signals.)

The model of synthesizer used does not affect these differences. Any potentialimprovement due to the synthesizer is obscured by the effects of the test set.

Capacitance Junction Measurements

In measuring capacitance junctions, the signal level at the device is generally kept below−30 dBm to keep nonlinearity-induced errors below 5 percent. Capacitance measurementsare typically made in the 50 to 100 fF region. Why is it necessary to keep the signal levelbelow −30 dBm for a capacitance measurement? In fact, it is not necessary! The device isbiased so as not to have any gain, therefore the input samplers and the test set are notsaturated, nor is the output of the device. Additionally, the power level set on thesynthesizer is not the same as the power level arriving at the device.

Depending on the frequency range and the test set, there can be anywhere from 15 to 30dB of loss between the source and the device. One method is to solve for the minimumresolvable capacitance,given an input power and a dynamic range. This should be done for both S21 (Cbc) andS11 (Cbe), because their respective resolutions are different.

Device input power −10 dBm.

Noise Floor −90 dBm.

Dynamic range >80 dB. (80 dB = 1E−04 linear)

S21 Related to Cbc

S11 Related to Cbe

This analysis verifies the observation that Cbc measurements are cleaner than Cbemeasurements. With an Agilent 8515B test set the same capacitance would put the signal40 dB above the noise floor, which would allow 20 dB of leeway even as low as 100 MHz.This verifies the initial statement that it is the S-parameter test set that determines theavailable range of power, and not the synthesized sweeper.

Defining the DC Source/Monitor Instrument State


105

NoteFor details on IC-CAP instrument options (instrument states) for specific instruments, refer to SupportedInstruments (measurement).

IC-CAP provides remote control of the DC source/monitor and the network analyzer tomeasure the device under test. This procedure configures IC-CAP with the instrumentstate settings (instrument options) for the DC source/monitor. Use this procedure if youhave been referred here from one of the main device modeling procedures. You will needto set the DC source/monitor instrument state separately for each measurement setup ina modeling procedure. However, use the same settings for all the measurement setups inone model (except that Integ Time can differ among setups, as explained below). Unlessyou use the default settings documented here, it is helpful to write down the settings toduplicate in the other setups.

NoteDC signal levels and other input parameters that differ among measurement setups are not set in thisprocedure, but in the individual setup procedures in each model chapter.

This procedure uses the forward gummel setup in the BJT measurement as an example.However, the settings used are defaults and the explanations are general.

In the DUT/Setup panel select a setup, for example fgummel.1.Select the Instrument Options tab, and a window will be displayed showing the2.instrument states for the DC source/monitor, as illustrated in the following figure.Set each of the DC source/monitor measurement parameters according to the actual3.Agilent 4142 configuration and the device to be measured, using the guidelines in thefollowing steps.

Example Agilent 4142 DC Source/Monitor Instrument State

Use User Sweep is normally set to No, to use the instrument's internalsweep, which is faster than a user-defined sweep.Hold Time is the delay in seconds before the sweep, to allow for DCsettling. Generally, no hold time is necessary.Delay Time is the time in seconds the instrument waits beforemeasuring at each step of an internal or user sweep. It can generallybe set to 0.000.For Integ Time, M (medium) is a good default choice. Inmeasurements where a long integration time is needed for noisereduction, you would use L (long), and you will be instructed to do soin certain procedures. An example is the forward gummelmeasurement, which measures very low current values. You canspeed up a measurement by using S (short), but this is notrecommended because it degrades the dynamic range of themeasurement.

Set Range to 0 to implement SMU auto-ranging.4.Power Compliance is used to set the maximum current/voltage combination for the5.DC source/monitor. However, in the models, SMU voltage and current compliancesare set individually in the individual setups. Therefore the value here can be set to0.000.SMU Filters ON is set to Yes to switch in low-pass filters on the SMU outputs. This6.protects the device from voltage spikes caused by DAC output changes.The IC-CAP models documented here are generally not configured for pulsed7.measurements, therefore Pulse Unit can be left blank. All other Pulse settings arethen irrelevant and can be ignored.Module Control is not used in these procedures: leave the field blank.8.Init Command sets the instrument to a mode not supported by other fields in this9.table. It is not used in these models. Leave the field blank.If you wish, refer to the DC source/monitor manual for more detail.10.If the measurement setup you are configuring calls for a DC measurement only, close11.the instrument options window. Then return to the modeling procedure.If the measurement setup also calls for a network analyzer measurement, continue12.to Defining the Instrument States for S-Parameter Measurements.

Defining the Instrument States for S-Parameter Measurements

IC-CAP provides remote control of the DC source/monitor and the network analyzer tomeasure the device under test. These procedures configure IC-CAP with both the DCsource/monitor and Agilent 8510 network analyzer instrument states for S-parametermeasurements. The first procedure sets the instrument states for a broadband swept-frequency measurement, and the second for a CW frequency measurement.

NoteFor details on IC-CAP instrument options (instrument states) for specific instruments, refer SupportedInstruments (measurement).

Instrument States for Swept Measurements

You will need to set the instrument states separately for each measurement setup in amodeling procedure. However, use the same settings for all the swept S-parameter


106

measurement setups in one model (except that Integ Time can differ among setups, asexplained under Defining the DC Source/Monitor Instrument StateUnless you use the example settings documented here, write down the settings you use,to duplicate among setups within one model. This procedure uses the sparm_bias setup inthe BJT measurement as an example, though the explanations here apply to all swept-frequency S-parameter measurements in the different device measurement procedures.

NoteSource parameters such as frequency range are not set in this procedure, but in the individual S-parameter setup procedures in each model chapter.

NoteDo not use these settings for a CW S-parameter measurement. The settings for a CW measurement aredifferent, and are documented later.

In the DUT/Setup panel select the S-parameter setup you need, for example1.sparm_0v.Select the Instrument Options tab, and a window will be displayed showing the2.instrument states. The following figure illustrates example instrument states.

Instrument States for Swept Measurements

Set the Agilent 4142 instrument state the same as for the DC measurements.3.Set the Agilent 8510 instrument state according to the actual hardware configuration,4.the calibration used, and the device to be measured, following these guidelines:Set Use User Sweep to No, to use the instrument's standard internal sweep. This is5.necessary for the network analyzer's calibration to be switched on.Hold Time is the time in seconds before the instrument starts a sweep, to allow for6.DC settling. Generally, no hold time is required.Delay Time is the delay in seconds the instrument waits before setting each7.frequency in user sweep mode. The default is 100.0m, but the value set will notaffect an internal sweep.Set Port 1 Atten and Port 2 Atten to the same levels in dB that you set in the network8.analyzer calibration. While the default is 0 dB, you may need to add attenuation formeasurements of amplifiers or high-power devices.Source Power is the RF synthesizer output power. Set it to the same value in dB that9.you set in the network analyzer calibration.Power Slope is used only if you use power slope in the network analyzer calibration:10.if so, set the corresponding value here. This is useful to view the response of a devicewith power dropoff at higher frequencies. The units for power slope are dB/GHz; thedefault value is 0.000. (A network analyzer message will caution you that thecorrection may be invalid, but this can be ignored.)Set Fast Sweep (RAMP) to No, because the network analyzer is in stepped-sweep11.mode, set earlier in the calibration procedure.Sweep Time applies only to ramp sweep mode, therefore the value set is irrelevant12.for a stepped-sweep measurement.Set Use Fast CW to Yes, to minimize repeated switching between the test set ports.13.Trim Sweep is set to 0. This feature is used only in ramp sweep mode.14.Set Avg Factor to the same averaging factor you set in the calibration. The default15.value is 256, but as little as 16 may beadequate.Set Cal Type (SHN) to H, for hardware.16.Set Cal Set No to the cal set number in the analyzer where you stored your swept17.calibration, so that IC-CAP can find the calibration.Soft Cal Sequence refers to the sequence of measurements of the cal standard18.devices: load-open-short-thru.Delay for Timeouts can generally be set to the default value of 0.000.19.Set Use Linear List to No because this is a standard stepped-frequency measurement.20.Init Command sets the instrument to a mode not supported by other fields in this21.table. It is not used in this model. Leave the field blank.Close the instrument options window. Then return to the modeling procedure.22.

Instrument States for CW Frequency Measurements

The instrument states for a CW S-parameter measurement are similar to those for aswept measurement, with a few important exceptions, listed below. You will need to setthe instrument states separately for each measurement setup in a modeling procedure.However, use the same settings for all the CW S-parameter measurements in onemodel (except that Integ Time can differ among setups.) Unless you use the defaultsettings documented here, write down the settings you use, to duplicate among setupswithin one model. This procedure uses the sparm_cje setup in the BJT measurement as anexample, though the explanations here apply to all CW S-parameter measurements in thedifferent device measurement procedures.

Frequencies are set in the setup Inputs, therefore they do not affect instrument states even if you chooseto use different CW frequencies fordifferent measurements, except for the Cal Set No where they are stored.

In the DUT-Setup panel select the CW S-parameter setup you need, for example1.sparm_cje.Select the Instrument Options tab, and a window will be displayed showing the2.


107

instrument states. The following figure illustrates example instrument states.


Set the Agilent 4142 instrument state the same as for the DC measurements.3.Set Use User Sweep to Yes, because CW is a nonstandard sweep.4.(Be sure to set Port 1 Atten and Port 2 Atten and Source Power to the same values as5.you used in the network analyzer calibration.)Set Use Fast CW to No.6.Set Cal Set No to the cal set number in the analyzer where you stored your CW7.subset calibration. This must be different from the cal set used by the sweptbroadband calibration, and from any CW calibration you make at a differentfrequency for any different measurement.Close the instrument options window. Then return to the modeling procedure.8.


108

Using IC-CAP with an Agilent 85123A DeviceModeling SystemThis section documents procedures for setup, calibration, and control of the Agilent85123A RF parameter extraction test system. Some of these are hardware proceduresindependent of IC-CAP, and some are hardware control procedures performed inside IC-CAP. These procedures are common among the high-frequency IC-CAP models. This is nota sequential set of procedures, but a set of separate hardware-related procedures used inconjunction with any of the device modeling procedures documented in the modelingchapters. Use the information in this appendix when you are directed here, at theappropriate times in the modeling procedures.

The procedures give detailed instructions for the following:

Configuring the system hardwarePerforming hardware setup in IC-CAPDefining the instrument states in IC-CAP (for DC, capacitance, and S-parametermeasurements)Calibrating the network analyzer

The instructions explain when the hardware-related procedures are needed and when theyare not. In some cases, you may not need to perform all these procedures each time youmodel a device. The system setup procedures are performed rarely. Setting theinstrument states is done for each measurement setup in each model, and the settingscan be stored for repeated use. The network analyzer calibration must be performedregularly for accuracy.

NoteThe instructions provided in this appendix are for an Agilent 8753 network analyzer. The procedures for anAgilent 8720 are very similar. For information specific to a different network analyzer, refer to itsoperating manual.

Description of the System

The Agilent 85123A RF parameter extraction test system is specially designed for use withAgilent 85190-series high-frequency IC-CAP software (and a compatible controller) tomeasure the DC and RF performance of active devices. The software is then used toextract the device model parameters. The standard system uses an Agilent 8753Dnetwork analyzer subsystem for S-parameter device characterization, and an Agilent4142B DC source/monitor to provide precision DC characterization as well as bias for theS-parameter measurements. The Agilent 8753D subsystem consists of the networkanalyzer with its integrated synthesizer and S-parameter test set, the RF cables, and thecalibration kit and verification kit if used.

The integrated synthesizer in a standard Agilent 8753D network analyzer supplies a sweptor CW RF signal in the range of 30 kHz to 3 GHz. The integrated test set separates the RFsignal into reference and test signals, and provides RF connections via cables and adaptersto the external bias networks. The Agilent 4142B source/monitor provides DC force(supply) and sense (measure) capability from its plug-in SMUs (source/monitor units). TheDC signals are routed via feedthrough panels to the bias networks, and thus RF and DCsignals are applied together to the device under test. The bias networks have 3.5 mm(female) connectors for interface to a test fixture or probe station. The transmitted andreflected signals from the device are measured and displayed by the receiver.

The system can be custom-configured to meet individual needs, for example to provide anoptional 6 GHz frequency range, different bias arrangements, or different cablingconfigurations. The network analyzer can be replaced with an Agilent 8720 (20 GHz) or8719 (13.5 GHz). Or the system can include instrumentation for different types ofmeasurements, such as power or noise figure measurements.

The system is factory-installed in a rack. A rack-mounted work surface is included, formaximum flexibility and convenience in making on-wafer, in-fixture, or coaxialmeasurements. The work surface is coated with antistatic material and is connected tochassis ground, therefore a static mat is not required.

Configuring the System Hardware

The Agilent 85123A system is shipped from the factory with all rack components andinstruments assembled and cabled in the system rack cabinet.System hardware configuration need be done only in the following circumstances:

At initial system setup.If changes are made to the system hardware.To change the bias connections for a different device type.Before starting a parameter extraction procedure, make sure the GPIB cables areconnected with the GPIB addresses set correctly. If you are using a special customsystem, refer to its installation and user's guide for custom cabling diagrams. Thestandard and special option Agilent 85123A installation and user's guides also provideadditional information you may need about setting up your system components.Connect the RF cables from the test ports to the RF IN ports ofthe bias networks, using adapters as needed.

The following provides details of the DC bias connections.

GPIB Interconnections and GPIB Addresses


109

DC Bias Safety Considerations

Bias current and voltage are supplied to a device under test from the plug-insource/monitor units (SMUs) in the Agilent 4142B DC source/monitor, connected to thedevice through the bias networks and probes or fixture.

Interlock Connection

To prevent electric shock from DC voltages in excess of +/−42V, do not close the INTLK(interlock) connection of the Agilent 4142B DC source/monitor. The high-power SMUoutput can be as high as +/−200V, and the medium-power SMU output can be as high as+/−100V. As long as the INTLK connection is open, the DC voltage is clamped at +/−42VDC.

Floating Ground

IC-CAP measurements are normally performed with the device in a floating-groundconfiguration, to prevent ground loop noise or, in the case of a BJT, possible damage to adevice. A floating-ground configuration is accomplished by insulating the fixture or probestation from power-line ground, for example with an insulator between the wafer undertest and the chuck of a probe station. If you are measuring in a floating-groundconfiguration, make sure the shorting bar of the Agilent 4142B is connected between theCIRCUIT COMMON and CHASSIS GROUND terminals on the front of the GNDU plug-in. Ifyou cannot implement a floating-ground configuration, it maybe necessary to open the Agilent 4142B shorting bar and connect the DUT ground to thecircuit common grounds at the DUT ends of the SMU and GNDU cables. The circuitcommon grounds are not connected through the bias networks: they are available at theinput jacks.

A potential shock hazard exists when the shorting bar is disconnected for floating measurements. Do nottouch any of the Agilent 4142B front panel connectors while a floating-ground measurement is inprogress.

Configuring the SMUs and Bias Networks

Bias current and voltage are supplied to a device under test from plug-in source/monitorunits (SMUs) in the Agilent 4142 DC source/monitor, connected to the device through thebias networks and probes or fixture. The standard Agilent 85123A system includes onemedium-power SMU and one high-power SMU. The quadraxial cables from the SMUs arerouted through a cable feedthrough panel to the rear of the system rack, and out to thefront of the rack above the work surface through another feedthrough panel.

Connect the cables emerging from the upper feedthrough panel to the bias networks, asexplained in the next paragraphs.

Bias Connections

The following figure illustrates the connections from the SMUs to the bias networks.Connect the FORCE and SENSE connectors on the quadraxial cables to the FORCE andSENSE connectors on the bias networks. Connect the triaxial cable from the GNDU SMU tothe ground connector on one of the bias networks (usually port 2). Leave the groundconnector of the other bias network unconnected.

Connections from DC Source/Monitor to Bias Networks

Connections from Bias Networks to Device

NoteFor in-fixture measurements, the measurement procedures in this manual apply to a FET mounted in thefixture in a common-source configuration, ora BJT mounted in a common-emitter configuration.

Connect the DC/RF signals from the outputs of the bias networks via the semi-rigid cablesto the corresponding terminals of the fixture or probe station. Set up the angle of thecables and bias networks to minimize stress and torque, and ensure that the cables are


110

properly supported. Iterate the positioning of the cables and bias networks until theconnectors line up correctly. Use a torque wrench to tighten the connections.The following figure illustrates the connections for a FET. The connections for a BJT arecomparable, with the port 1 DC/RF signal connected to the base, the port 2 DC/RF signalconnected to the collector, and the emitter grounded.

Connections from Bias Networks to In-Fixture DUT

For additional hardware information, refer also to the installation and user's guide foryour Agilent 85123A RF parameter extraction test system, and to the manuals for theindividual instruments in the system.For information on fixtures or probe stations, refer to the manufacturer'sdocumentation.

Switching on Power to the System

Be sure to switch on the network analyzer and the DC source/monitor before thecomputer, so that the computer recognizes the network analyzer subsystem and the DCsource/monitor.

Preparing for Calibration

Let the system warm up for at least 1 hour before calibrating the network analyzer. TheDC measurements can be performed during this time. The calibration standards will needto be at ambient room temperature when the measurement calibration is performed. Ifyou are using a fixture, open the fixture calibration kit and place all the devices on top ofthe foam so they will reach room temperature by the time the system is warmed up. Ifyou are using a probe station, place the probed impedance standard substrate (ISS) onthe probe chuck and allow it to reach ambient temperature before use. The calibrationprocedures can be found in the section, [#SweptBroadband Calibration].

Terminating the Input Ports to Prevent Oscillation

Leave the network analyzer disconnected from the bias networks while you perform theDC measurements. Connect terminations with >20 dB return loss (such as 10 dB pads) tothe RF IN connectors of the bias networks, to prevent possible bias oscillation. Thenetwork analyzer is calibrated later in the modeling procedures, immediately before thecapacitance and S-parameter measurements. Alternatively, isolate the device from RFpower by manually setting 90 dB of attenuation at the network analyzer test ports. Alsoput the network analyzer in hold sweep mode by pressing MENU > TRIGGER MENU >HOLD.

Installing the Device

The best time to install the device is immediately before making measurements.

Ground yourself with an antistatic wrist strap to reduce the chance of electrostatic discharge. A groundterminal is provided on the lower left corner of the test set front panel for your convenience. (Note thatthe system work surface is coated with antistatic material and is connected to chassis ground, therefore astatic mat is not required.)

Install the device carefully on the probe chuck or in the fixture. Be sure to handle thedevice as little as possible, to avoid damaging it. Once the device is in place, avoidbumping the test station.

NoteEliminate direct light sources (such as microscope light) during measurement, especially for GaAs devices,as light may influence the measurement and cause inaccurate results.

Performing Hardware Setup in IC-CAP

The measurement system hardware is controlled by the IC-CAP software. This procedureconfigures IC-CAP to recognize the system instruments on the GPIB, and the individualsource/monitor units (SMUs) in the DC source/monitor. The complete procedure need onlybe done after initial system installation, or any time the system hardware is changed, toestablish a system handshake. The second part of the procedure (renaming the SMUs)also needs to be done any time you change device types, such as from FET to BJT or fromBJT to FET. For help with launching IC-CAP and opening the model file you need, refer toChapter A, "Agilent 85190A IC-CAP" thencontinue with these steps:

In the IC-CAP Main window, select Tools > Hardware Setup. The hardware window is1.displayed. It will vary depending on the system hardware used, but will be similar tothe following figure:

Hardware Window


111

Under the Instrument List select Rebuild. This command causes IC-CAP to poll all2.available GPIB addresses, and add to the Instrument List any newly connectedinstruments that are powered up. (It does not affect instruments already on the list,whether they are powered up or not.)

Renaming the SMUs

Use this procedure any time you change device types, such as from FET to BJT or fromBJT to FET, so that the measurement setups can properly communicate with the biassupplies. When you run the IC-CAP software, it initially identifies the plug-in SMUsaccording to the numbers of the Agilent 4142 slots in which they are installed. A medium-power SMU occupies one slot. A high-power SMU occupies two slots, and is identified bythe higher slot number of the two. An Agilent 85123 system can include differentcombinations of SMUs: the standard system includes two SMUs, one Agilent 41421Bmedium-power SMU and one Agilent 41420A high-power SMU. The medium-powerSMU is factory-installed in slot 1 and is initially identified in the IC-CAP software asMPSMU1. The high-power SMU is installed in slots 2 and 3, initially identified as HPSMU3.The ground unit is identified as GNDU. It is generally more convenient to assign unitnames to the SMUs that identify their purpose in a devicemeasurement, depending on the bias connections at the device terminals and the type ofdevice you are modeling.

In configuring IC-CAP to recognize the system hardware, you need to set the SMU namesused in software to identify the SMU connections at the measurement terminals. In a FETmeasurement, the SMUs are identified as VG (gate supply) and VD (drain supply). In aBJT measurement, the SMUs are identified as SMU1 and SMU2, corresponding to thenetwork analyzer test ports: this helps to facilitate measurement of a BJT with only twoSMUs, since the port 2 SMU is used for measurements at both the collector and emitterterminals. Some models use more than two SMUs: their SMU names are defined in theirrespective chapters. The following figure illustrates the SMU unit names for a FETmeasurement and the correspondence to the device terminal connections.

SMU Unit Names for a FET Measurement

The following figure illustrates the SMU unit names for a Gummel-Poon BJT measurementand the correspondence to the device terminal connections.

SMU Unit Names for a BJT Measurement

In the Instrument List click on HP4142 to highlight it. Select Configure and the1.Configuration of HP4142 window is displayed.The Unit Table lists the SMUs. Also listed for each is a unit ID (identification) name2.that can be edited.In a FET measurement, the medium-power MPSMU1 is renamed VG, because it isconnected to the gate terminal; and the high-power HPSMU3 is renamed VD, becauseit isconnected to the drain terminal.In a Gummel-Poon BJT measurement, the medium-power MPSMU1 is renamed SMU1because it is the port 1 SMU; and the high-power HPSMU3 is renamed SMU2 becauseit is the port 2 SMU.To make this change, click left with the mouse and move the mouse pointer over the3.name assigned to MPSMU1 to highlight it. Then type in the new name, and pressReturn. Similarly, change MPSMU3's assigned unit name. Then select OK to close theconfiguration window.To save this hardware configuration in a file for future use, select File > Save. A file4.filter is displayed, showing the pathname of the current directory. Type in anappropriate name for the .hdw file you are creating, and select OK.Close the hardware window.5.

After IC-CAP is configured to recognize the system hardware, you can proceed to one ofthe modeling chapters.

NoteFor details on IC-CAP instrument options (instrument states) for specific instruments, refer to SupportedInstruments (measurement).

IC-CAP provides remote control of the DC source/monitor and the network analyzer tomeasure the device under test. This procedure configures IC-CAP with the instrumentstate settings (instrument options) for the DC source/monitor. Use this procedure if youhave been referred here from one of the device modeling procedures.You will need to set the DC source/monitor instrument state separately for each DC or DC-biased measurement setup in a modeling procedure. However, use the same settings forall the measurement setups in one model (except that Integ Time can differ amongsetups). Unless you use the default settings documented here, it is helpful to write downthe settings to duplicate in the other setups.

NoteDC signal levels and other input parameters that differ among measurement setups are not set in thisprocedure, but in the individual setup procedures in each model chapter.

The settings used in this procedure are defaults and the explanations are general.


112

Select the DC measurement setup of your choice.1.Select the Instrument Options tab, and a window will be displayed showing the2.instrument states for the DC source/monitor. Figure 203 illustrates example DCsource/monitor instrument states.Set each of the DC source/monitor measurement parameters according to the actual3.Agilent 4142 configuration and the device to be measured, using the guidelines in thefollowing steps.

Example Agilent 4142 DC Source/Monitor Instrument State

Set Use User Sweep to No, as it is unnecessary for these procedures, and the4.source/monitor internal sweep is faster.Hold Time is the delay in seconds before starting a sweep, to allow for DC settling.5.Generally, no hold time is required.Delay Time is the time in seconds the instrument waits before taking a measurement6.at each step of a sweep. Generally no delay time is needed.For Integ Time, M (medium) is a good default choice. In measurements where a long7.integration time is needed for noise reduction, you would use L (long), and you willbe instructed to do so in certain procedures. You can speed up a measurement byusing S (short), but this is not recommended because it degrades the dynamic rangeof the measurement.Set Range to 0 to implement SMU autoranging.8.Power Compliance is used to set the maximum current/voltage combination for the9.DC source/monitor. However, in the models, SMU voltage and current compliancesare set individually in the individual setups. Therefore the value here can be set to0.000.Set SMU Filters ON to Yes. This switches in low-pass filters on the SMU outputs, to10.protect the device from voltage spikes caused by DAC output changes.High-frequency IC-CAP is not generally configured for pulsed measurements,11.therefore Pulse Unit can be left blank. All other Pulse settings are then irrelevant andcan be ignored.Module Control is not used in these procedures: leave the field blank.12.Init Command sets the instrument to a mode not supported by other fields in this13.table. It is not generally used in these models. Leave the field blank unless instructedotherwise.If you wish, refer to the DC source/monitor manual for more detail.14.If the measurement setup you are configuring calls for a DC measurement only, close15.the instrument options window. Then return to the modeling procedure.If the setup also calls for a network analyzer measurement, continue to Defining the16.Instrument States for S-Parameter Measurements

Defining the Instrument States for S-Parameter Measurements

NoteFor details on IC-CAP instrument options (instrument states) for specific instruments, refer SupportedInstruments (measurement).

IC-CAP provides remote control of the DC source/monitor and the network analyzer tomeasure the device under test. These procedures configure IC-CAP with both the DCsource/monitor and the Agilent 8753 network analyzer instrument states for S-parametermeasurements. The first procedure sets the instrument states for a broadband swept-frequency measurement, and the second for a CW measurement.

Instrument States for Swept Frequency Measurements

Use this procedure if you have been referred here from one of the device modelingchapters. You will need to set the instrument states separately for each measurementsetup in a modeling procedure. However, use the same settings for all the swept S-parameter measurement setups in one model (except that Integ Time can differ amongsetups). Unless you use the example settings documented here, write down the settingsyou use, to duplicate among setups within one model.

The settings used here are examples. Your settings may differ. Explanations here apply toall swept-frequency S-parameter measurements in the different device measurementprocedures.

NoteSource parameters such as frequency range are not set in this procedure, but in the individual S-parameter setup procedures in each model chapter.Do not use these settings for a CW S-parameter measurement. The settings for a CW measurement areslightly different, and are documented following this procedure.

Select the swept S-parameter setup of your choice.1.Select the Instrument Options tab, and a window will be displayed showing the2.instrument states. The following figure shows an example.

Instrument States for Swept Frequency Measurements with Agilent 8753D


113

Set the Agilent 4142 instrument state the same as for the DC measurements.3.Set the Agilent 8753 instrument state according to the actual hardware configuration,4.the calibration used, and the device to be measured, following these guidelines:Set Use User Sweep to No, because this is a standard internal sweep.5.Hold Time is the time in seconds before the instrument starts a sweep, to allow for6.DC settling. Generally, no hold time is required.Delay Time is the delay in seconds the instrument waits before setting each7.frequency. No delay is needed here, but the default is 100 ms.Port 1 Source Power and Port 2 Source Power are used with an analyzer that has an8.integrated test set. Set the power to the same level as in the network analyzercalibration. Take care that the power level will not be excessive at the device input orat the input port samplers of the analyzer. The default is -10 dBm.Port 1 Power Range and Port 2 Power Range are used with an analyzer that has an9.integrated test set. The synthesized source in the analyzer contains a programmablestep attenuator with eight power ranges. This lets you determine which range isused: the default is 0.Port 1 and Port 2 Auto Power Range are used with an analyzer that has an integrated10.test set. These settings enable an autoranging power level and attenuation capability.The LOCAL mode by pressing MENU > SWEEP TIME. If you set a reduced IFbandwidth for calibration, the sweep time may have been slowed down automatically.Set IF Bandwidth to the same value as in the calibration. The noise floor can be11.reduced by reducing the receiver input bandwidth.Set Use Fast CW to No, because fast CW is not compatible with the instrument12.calibration.Set Avg Factor to the same averaging factor you set in the calibration.13.Set Cal Type to H (hardware) so that IC-CAP will recognize calibration device14.measurements at the network analyzer front panel.Cal Set No must be set to the analyzer register number where you store your swept15.broadband calibration. Do not use register 6, which stores the active instrumentstate.Soft Cal Sequence refers to the sequence of measurements of the cal standard16.devices: load-open-short-thru. You can set a different sequence if you prefer.Delay for Timeouts increases the timeout and wait times. It can generally be set to17.the default value of 0.000.If you define a specific number of points for the measurement calibration, set Use18.Linear List to No, the normal setting for a swept measurement. The setting wouldbe Yes for a log or list sweep.Close the instrument options window. Then return to the modeling procedure.19.

Different Settings for Agilent 8753 with External Test Set

An Agilent 8753 with an external test set uses different settings for power level andattenuation. Also, these analyzers do not have the power autoranging or coupled portpower functions. These are the different settings:

Port 1 Atten and Port 2 Atten must be set to the same level of attenuation you set in1.the network analyzer calibration. The default is 20 dB.Source Power must be set to the same level as in the network analyzer calibration.2.The default is −10 dBm.


Use this procedure if you have been referred here from one of the device modelingchapters. You will need to set the instrument states separately for each measurementsetup in a modeling procedure. However, use the same settings for all the CW S-parameter measurement setups in one model (except that Integ Time can differ amongsetups).

Unless you use the settings documented here, write down the settings you use, toduplicate among setups within one model. The settings used here are examples. Yoursettings may differ. Explanations here apply to all CW S-parameter measurements in thedifferent device measurement procedures.

Select the CW S-parameter setup of your choice.1.Select the Instrument Options tab and a window will be displayed showing the2.instrument states. The following figure shows an example.

Instrument States for CW Frequency Measurements with Agilent 8753D

Set the Agilent 4142 instrument states the same as for the DC measurements.3.The network analyzer instrument state settings for a CW-frequency S-parametermeasurement are the same as for a swept S-parameter measurement, with thefollowing exceptions:Set Use User Sweep to Yes because CW is a nonstandard sweep.4.Be sure to set the source power and attenuation levels the same as in the network5.analyzer calibration.


114

Set Cal Set No to the analyzer register number where you stored your CW6.calibration. This must be different from the register used by the swept broadbandcalibration, and from any other CW calibration you make at a different frequency forany different measurement.You can set Use Linear List to either Yes or No. This setting is ignored by IC-CAP for a7.CW measurement.Close the instrument options window. Then return to the modeling procedure.8.

Calibrating the Network Analyzer

The network analyzer must be calibrated before any S-parameter measurements areperformed. The system must be allowed to warm up for at least 1 hour before calibration,and the calibration standards must be at ambient room temperature. It is a good practiceto switch on the analyzer and expose the calibration devices to air before you make theDC measurements.

Good calibration of the network analyzer system is critical to a good high-frequencymeasurement and extraction. A good calibration is dependent on the quality of thecalibration kit standard devices, the care with which they are maintained, and thecorrectness and repeatability of the device connections. A stable ambient temperature(±1°C) is required to maintain a good calibration.

For the high-frequency IC-CAP measurement procedures, the network analyzer iscalibrated over a broadband frequency range, to perform the S-parameter measurementsfor AC and parasitic extractions. In addition, one or two calibrations may be needed at CWfrequencies for specific measurement setups. These procedures explain how to set up abroadband calibration and a separate CW calibration. For the Agilent 8753 and 8720systems, IC-CAP does not support the subset calibration technique used for an Agilent8510 system. The instructions given here are very abbreviated: more detailed informationis available in the Agilent 8753 operating manual.

NoteIt is important that values entered in software correspond with actual instrument settings. Be sure to writedown the parameters of your calibrations, so that they are readily available when you need to enter themin IC-CAP.

To connect the network analyzer:Disconnect the terminations from the bias networks. Connect the RF cables from thenetwork analyzer test ports to the RF inputs of the bias networks.

Swept Broadband Calibration

In the IC-CAP model, select the S-parameter setup of your choice.1.In the measurement setup, set the instrument options appropriately for your2.measurement.You can set the network analyzer manually from its front panel, or use the IC-CAP3.Calibrate command in the setup menu to download the instrument options into thenetwork analyzer. In either case, you perform the cal standards measurements at theanalyzer front panel.The following steps explain how to perform a manual network analyzer front panel4.calibration.On the network analyzer, press LOCAL to gain front panel control. Press PRESET to5.return to a known standard state.If you are using a system with the 6 GHz receiver option and you wish to measure in6.the 3 MHz to 6 GHz range, press SYSTEM > FREQ RANGE 3GHz6GHz.To define the frequency range of the calibration and measurement, press START, and7.use the numerical keypad to set the start frequency, ending with one of theterminator keys (such as G/n) at the right of the keypad. Similarly, press STOP andset the stop frequency. Set a frequency range at least as wide as the operating rangeof the device.Press MENU > NUMBER OF POINTS, and enter the number of points to be8.measured across the range. Allow enough points to provide good measurementresolution: at least 51 points are recommended.Press MENU > TRIGGER MENU > CONTINUOUS to set a continuous stimulus9.sweep.Press MENU > POWER, and set the desired power level. In setting the source power10.and test set attenuation, take care that the power level will not be excessive at thedevice input. Also consider the gain of the device, and set a power level that will notsaturate the input port samplers of the analyzer. If a receiver input is overloaded(>+14 dBm), the analyzer automatically reduces the output power of the source to−85 dBm and displays the error message OVERLOAD ON INPUT (R, A, B) POWERREDUCED. In addition, the annotation P appears in the left margin of the display toindicate that the power trip function has been activated. When this occurs,reset the power to a lower level, then toggle the SOURCE PWR on/OFF softkey toswitch on the power again.For a device with power dropoff at higher frequencies, you may wish to set a power11.slope using the SLOPE ON/off softkey. An appropriate power slope would be in theregion of 2-3 dB/GHz.Press AVG > AVERAGING FACTOR, and enter an averaging factor high enough to12.reduce trace noise and increase dynamic range as appropriate for your devicemeasurements.A good default averaging factor is 256. To speed your measurements, you may find itconvenient to set an averaging factor as low as 16. Press AVERAGING ON.You can further reduce the noise floor by reducing the receiver input bandwidth.13.Press IF BW, and enter one of the following allowed values in Hz: 3000, 1000, 300,100, 30, or 10. A tenfold reduction in IF bandwidth lowers the measurement noisefloor by about 10 dB; however, the sweep time may be slower. For more informationon averaging and the different trace noise reduction techniques, refer to the Agilent8753 operating manual.Press CAL > CAL KIT, and select the appropriate default or user-defined cal kit for14.your calibration devices.If you wish to modify an internal calibration kit definition (see below), do so now.15.Press CAL > CALIBRATE MENU, and perform the calibration of your choice at the16.analyzer front panel, measuring each of the standard devices in turn and pressing thesoftkeys as each measurement is complete. A full two-port cal provides the greatestaccuracy. A TRL* or LRM* cal is an appropriate alternative for in-fixturemeasurements. Omit isolation cal.Press DONE, and save the cal in the register number specified in the instrument17.options table. This must be a different register than you use for any CW cal.Detailed procedures for measurements of calibration standards are provided in the18.Agilent 8753 operating manual.

CW Frequency Calibration


115

NoteA CW frequency calibration is similar to a broadband calibration with some exceptions. Repeat theprocedure used for a broadband calibration, withthese changes.

To set the CW frequency, press MENU > CW on the analyzer, and enter the1.frequency using the numerical keypad and one of the terminator keys.Make a separate set of standards measurements.2.When you save the calibration, use the cal register number specified in the3.instrument options table for this CW cal. This must be a different register than youused for the swept cal or any other CW cal.

After the network analyzer is calibrated, return to the modeling chapter you are using forthe S-parameter measurement procedures.

Using the Automatic IC-CAP Calibrate Function

Alternatively, you can use the IC-CAP Calibrate command for a broadband or CWcalibration. This downloads the instrument states from IC-CAP to the network analyzerbefore you perform a calibration with the analyzer. This helps to avoid errors in matchingvalues set in the analyzer and in IC-CAP.

In the IC-CAP model, select the S-parameter setup of your choice.1.In the measurement setup, set the instrument options appropriately for your2.measurement.From the setup menu select Calibrate. The instrument options are downloaded into3.the analyzer, together with the frequency stimulus values set in the Inputs.The message Calibrate HP8753.7.16 before measuring is displayed. Select OK and4.perform the cal standards measurements from the analyzer front panel. Save the calinthe register specified in the instrument options.

Modifying a Cal Kit Definition

The network analyzer's internal calibration kit models are accurate for use with thestandard coaxial calibration kits defined for use with the Agilent 8753. In certaincircumstances, however, you may need a user-defined kit for compatibility with yourcalibration standards. Some possible cases might be these:

If you are using a probe station and you know the probe capacitances, you can enterthem into the cal kit definition so that their effects can be removed by themeasurement calibration.The impedance of an on-wafer load may not be exactly 50Ω.You can characterize the DC impedance of a load exactly with a precision ohmmetermeasurement, and enter the impedance value into the cal kit definition. Theprocedure to modify an internal cal kit definition is complex, and not to beundertaken without a solid understanding of error correction and the system errormodel.It is explained in detail in Product Note 8510-5, SpecifyingCalibration Standards forthe Agilent 8510 Network Analyzer (Agilent literature number 5954-1559). Forinformation specific to the Agilent 8753 network analyzer, including an exampleprocedure, refer to the Agilent 8753 operating manual. For information on thecharacteristics of calibration standards used with a fixture or probe station, refer tothe manufacturer's device data sheets or other documentation.

TRL* Calibration Under IC-CAP Control

The Agilent 8753D network analyzer has the capability to perform a TRL* or LRM*calibration using internal firmware. TRL* (TRL-star) is an implementation of TRL thathas been adapted for three-sampler network analyzers for use in fixturedmeasurement environments. (Note that this requires that you modify a user cal kitdefinition: see above.)Earlier models of the Agilent 8753 do not have the internal TRL* calibrationcapability. The IC-CAP procedure described in the following sections allows you toperform a similar Agilent 8753C calibration using TRL cal standards compatible withyour fixture. IC-CAP calculates the error coefficients and downloads them into thenetwork analyzer. The procedure can be used for either a TRL* (thru-reflect-line) orTRM* (thru-reflect-match) calibration.

When the calibration is complete, the reference plane is defined at the middle of the thrustandard, or at the interface to the DUT when it is installed in the compatible carrier. Formore information about TRL* calibration, refer to Product Note 8720-2, In-fixturemicrostrip device measurements using TRL* calibration, Agilent literature number 5091-1943E.

NoteThe network analyzer GPIB address must be set to 16. It is recommended that you connect 10 dB pads atthe RF IN connectors of both bias networks to improve the effective source and load match of themeasurement setup.

The IC-CAP software for the TRL* calibration is in the model file TRLCAL.mdl. Find theTRL* calibration model under:/examples/model_files/misc/TRLCAL.mdlThe message WARNING: HP8753.7.16 Options saved in TRLCAL/cal/short (or anotherfilename) may be displayed. In this case, you will need to perform a Rebuild, as explainedunder Performing Hardware Setup in IC-CAP.

The TRL* Calibration DUTs and Setups

The following figure illustrates the DUTs and setups for TRLCAL.mdl. The SETUPs are usedto collect the measured raw S-parameters of various TRL (or TRM) standards, usingvariable values defined by you in the macros. IC-CAP computes the required errorcoefficients from this raw S-parameter data.

TRLCAL DUT/Setup Panel


116

The TRL* calibration uses separate setups for a swept frequency cal (cal_sw) and a CWfrequency cal (cal_cw), which are both required in some modeling procedures. For theAgilent 8753 and 8720 systems, IC-CAP does not support the subset calibration techniqueused for an Agilent 8510 system, therefore two sets of standards measurements arerequired. IC-CAP stores measured data in the setups for both the swept and CW cals.

The README macro provides an abbreviated set of instructions for the TRL* calibrationprocedure. The other macros provide interactive prompts for defining the instrumentstates and calibration frequencies, and for performing the calibration.

Definitions for Standards and Measurement Setups

The IC-CAP TRL* calibration requires measurements of three or more standard devices: ashort, a thru, and at least one line standard (for TRL*) or match standard (for TRM*). Theperformance characteristics of the standards must be specified over the frequency rangeof the calibration. Therefore it may be necessary to use as many as three line standards tocover the total frequency range for a swept frequency cal. If so, you will need to definethe transition frequencies between the line standards.

Some confusion can arise concerning the transition frequencies between the line standards, because thesetup names do not necessarily correspond with the actual line standard names. Read the sectionMeasurement Setups for further details.

Measurement Setups

The setups in the TRLCAL model measure the calibration standards, and store themeasured raw S-parameter data in files with the same names as the setups. Eachstandard is measured over the full frequency range of the calibration. However, if multipleline standards are used, IC-CAP processes only the data delimited by the transitionfrequencies to compute the error coefficients of each. For a CW cal, IC-CAP processes onlythe data measured at the CW frequency you specify. The setups are used as follows:Short: Measures the short standard and stores the measured data.

Thru: Measures the thru standard and stores the measured data.

LineA: Measures the first (lowest frequency) line standard and stores the measured data.In calculating the error coefficients for a swept frequency cal, IC-CAP uses only the datafrom the start frequency to the first transition frequency, or to the stop frequency if anabove-range or zero-value first transition is defined. (You will define the calibration rangeand the transition frequencies later in the procedure.) The lineA standard can berepresented either by a length of transmissionline for a TRL* cal, or by a matched load for a TRM* cal.

LineB: Measures the second line standard (if any) and stores the measured data. Incalculating the error coefficients for a swept frequency cal, IC-CAP uses only the data fromthe first transition frequency to the second transition frequency, or to the stop frequency ifan above-range or zero-value second transition is defined. Typically the standardmeasured is a length of transmission line different from either the thru or lineA standards.

LineC: Measures the third line standard (if any) and stores the measured data. Incalculating the error coefficients for a swept frequency cal, IC-CAP uses only the data fromthe second transition frequency to the stop frequency. Typically the standard measured isa length of transmission line different from either the thru, lineA, or lineB standards.

Transition Frequencies of Line Standards

Your calibration kit may have only one line (or match) standard, or it may have more thanone line standard, each specified over a portion of the total specified cal kit frequencyrange. Refer to the data sheet for your calibration kit to determine the specified frequencyranges of each of the devices. For the RF frequency range you wish to measure, you mayneed to measure all the line standards, or you may need to measure only one or two. Ifyou measure more than one line standard, the specified upper frequency limit for the firstline standard is the first transition frequency. Similarly, the specified upper frequency limitfor the second line standard is the second transition frequency. For example, if your totalmeasurement frequency range is 300 kHz to 6 GHz and your first line standard is specifiedfrom300 kHz to 1 GHz, then your first transition frequency is 1 GHz. If your second linestandard is specified from 1 GHz to 5 GHz, then your second transition frequency is 5GHz.

The TRLCAL.mdl file provides for measurements of up to three line standards for a sweptfrequency cal. If you use only one or two, you can define transition frequencies above therange of the calibration, and IC-CAP defaults to the start and/or stop frequencies. Forexample, if your total measurement frequency range is 300 kHz to 3 GHz and your secondline standard is specified from 1 GHz to 5 GHz, you can set the second transitionfrequency to 5 GHz (or 0), and IC-CAP will measure the second line from 1 GHz to 3 GHz.

Measurement Setup Names vs. Actual Standards Names

The measurement setup names may not correspond with the names of the standards inyour calibration kit. It is important to understand that the setup names are only filenamesfor the measured data. The actual fixtured standards you use for the line measurementswill depend on the frequency range of your calibration, not on the names of the setups.You must use the lineA setup for your lowest frequency line (or match) measurement,even if the cal standard you use is not line 1 in your calibration kit. This is because IC-CAP


117

processes measured data from the setup files in the sequence they are shown in themodel editor window. If a file is left empty (for example lineA), IC-CAP will stopprocessing data, even if there is measured data in lineB or lineC. For example, you mayhave a calibration kit with a frequency range of 300 kHz to 6 GHz, and you may becalibrating for device measurements over a range of only 5 GHz to 6 GHz, using a short, athru, and a single line (or match) standard. In this case, you will use the lineA setup tomeasure your line or match standard that is specified for performance over the 5 GHz to 6GHz frequency range, no matter how it is labeled.

If you are performing a CW cal alone, you must use the lineA setup for your line or matchmeasurement.

Defining the IC-CAP TRL* Cal Instrument States

IC-CAP controls the network analyzer in measuring the calibration standards. Theseprocedures configure IC-CAP with the required network analyzer instrument states tomeasure the TRL* calibration standards for the swept frequency and CW frequency calsfor both the Agilent 8753D and earlier-model Agilent 8753s. (Instrument states are similarfor Agilent 8720-family instruments, and are explained in Supported Instruments(measurement) for each cal and load it into all the corresponding measurement setups.

Select the Macros tab. Select the Instr_opts macro, and Execute.1.The instrument options window illustrated in the following figure is for the short setupfor a swept frequency cal using an Agilent 8753D.

TRL* Cal Instrument States for Swept Measurements with Agilent 8753D

In the dialog box, select OK as prompted, then make any instrument state changes2.required for your calibration and measurements, using the illustrations and thefollowing instructions for guidance. The illustrated values are typical defaults for aTRL* calibration used for measuring BJTs. The settings you use may be different,depending on the characteristics of the devices you intend to measure and model.Swept Measurements with Agilent 8753DFor swept measurements, set Use User Sweep to No because this is a standard3.internal sweep.Set the source power level in dBm to 0.000 for measurements below 3 GHz. For4.measurements above 3 GHz, 20.00 is appropriate. The power levels are tied to thefrequency range set in Init Command, below.1000 Hz is a good default value for IF bandwidth. Reducing the IF bandwidth lowers5.the noise floor, at the cost of slowing the sweep speed. Allowed IF bandwidth valuesin hertz are 3000, 1000, 300, 100, 30, and 10.Set Use Fast CW to Yes for a swept frequency measurement, to minimize repeated6.switching between the test set ports.Cal Type (SHN) refers to software/hardware/none. Set it to N for "none." This is7.because the measurements you will make in this cal procedure are uncalibratedmeasurements. The finished calibration set will be used in your later measurementand modeling procedures.Set Cal Set No to cal_reg, a variable used by the program.8.The setting you use for Soft Cal Sequence is irrelevant, since Cal Type is set to None.9.Set Use Linear List to No, otherwise the Agilent 8753 will be set to frequency list10.mode.Set Init Command to FREQRANG3GHZ if you are using a standard Agilent 875311.network analyzer. (If you are using an Agilent 8753 with the 6 GHz extendedfrequency option and measuring above 3 GHz, enter FREQRANG6GHZ.)

Swept Measurements with Earlier-Model Agilent 8753

The window illustrated in the following figure is for the short setup for a swept frequencycal using an earlier-model Agilent 8753.

TRL* Cal Instrument States for Swept Measurements with Earlier-Model Agilent 8753 (>3 GHz)


118

We recommend you use the same attenuator settings at port 1 and port 2, since theearlier Agilent 8753 analyzer uses the same attenuator for both ports. In setting powerlevels, you will need to set a source power sufficient to maintain phase lock, while takingcare not to saturate the device output. Settings of 10 to 20 dB attenuation at both portswith a source power level of 0 dBm are appropriate for many devices with gain.

CW Measurements

Some of the instrument state settings for the TRL* CW cal are the same as for the1.swept frequency cal. The following settings are different, as illustrated:

TRL* Cal Instrument States for CW Measurements with Agilent 8753D

For CW measurements, set Use User Sweep to Yes.2.You may not need to set attenuation for CW measurements, since gain may not be a3.factor. The CW measurements are usually used to determine the junction capacitanceof a device. For this purpose the device is biased in such a way that it has loss ratherthan gain.Set Use Fast CW to No.4.

TRL* Cal Instrument States for CW Measurements, Earlier Agilent 8753

When you have completed the settings, select the Instr_dupl macro, and Execute.5.This macro loads the short instrument options for the swept frequency cal into the


119

other cal_sw setups; and the short options for the CW cal into the other cal_cwsetups (thru, lineA, lineB, lineC).

NoteWhen you execute the Instr_dupl macro, the macro must write files to disk. If the instrument options donot load down, your working directoryprobably does not have write permission set. To correct this, from your working directory type, chmod 644<directory_name> and press Enter.

Defining the Calibration Frequencies

You define the values for frequency range, number of measurement points, and transitionfrequencies between line standards in the Cal_freqs macro, as follows:

Select the Cal_freqs macro, and Execute. A dialog box will be displayed.1.Enter the start frequency for your swept frequency calibration as prompted, and2.select OK.Enter the stop frequency and select OK. The stop frequency must be greater than the3.start frequency.At the prompt to enter the number of frequency points, use one of the following4.Agilent 8753 allowable values:If you are doing a CW cal along with the swept cal, enter the CW frequency. If not,5.enter 0. (If you are doing a CW cal alone, refer to the Note below.) Select OK.If you are using more than one line/match standard, enter the transition frequency6.for the first line (between lineA and lineB). Select OK.If you are using three line standards, enter the transition frequency for the second7.line (between lineB and lineC). Select OK.If you are using only one or two line standards, enter above-range or zero-value8.transition frequencies for the setups that do not apply. IC-CAP will measure only fromthe start to the stop frequency. Refer to the paragraphs under Definitions forStandards and Measurement Setups for explanations of transition frequencies andthe difference between the software setup names and the hardware standardsnames.The information there can prevent possible problems in the way IC-CAP processesthe data.When you have defined all the calibration frequencies, the UNIX window will list9.them, and will conclude with the message Completed.

NoteIf you are doing a CW cal alone (instead of with a swept cal), enter a value of 0 to the prompts for startfrequency, stop frequency, and number ofpoints. Remember to use the lineA setup for your line/match measurement. The Calibration macro willsubsequently provide only prompts that apply to a CW cal.

Performing the Calibration

The Calibration macro provides a series of prompts to guide you through the standardsmeasurements. In this part of the procedure, you follow the prompts to connect thedefined standards, and instruct IC-CAP to measure them. You then specify networkanalyzer registers to store the swept and CW calibrations, and IC-CAP calculates the errorcoefficients from the measured raw S-parameter data, then downloads them into theanalyzer registers.

Select the Calibration macro, and Execute.1.IC-CAP displays a dialog box, prompting you to perform the standards measurements2.you have defined. In each case, first connect the standard, then respond to theprompt by entering s for short, t for thru, A for the lineA setup, B for the lineB setup,and C for the lineC setup. Each time you enter a single letter, press Return. Amessage in the UNIX window tells you the standard is measured.

Be sure to connect the calibration devices that correspond with the frequencies you defined in theCal_freqs macro. Otherwise your calibration may be compromised. For more information, refer toDefinitions for Standards and Measurement Setups.

Following this prompt sequence, a dialog box asks you Done? [y/n]. If all the3.necessary cal standards have been measured, type y and select OK. (If you want toremeasure a standard, type n, reconnect the device, and type the appropriatecharacter for the device setup. IC-CAP will remeasure, and replace the old data withthe new.)The next dialog box asks you to enter the calibration storage register for the swept4.cal. Type in a number between 1 and 4, and select OK. (Do not use register 5,because IC-CAP reserves it for temporary storage.) IC-CAP calculates the errorcoefficients, downloads them into the network analyzer, and turns on the analyzercalibration.Next enter a different register number for the CW calibration, and select OK. IC-CAP5.performs the CW cal and downloads it.The instrument state and calibration are saved in the analyzer's short-term non-6.volatile memory (that is, even if the instrument is switched off for as long as 72hours).Whenever you use these swept and CW TRL* calibrations for a modeling procedure,7.enter the corresponding register numbers in the Cal Set No field of the IC-CAPinstrument options window. Also set the Cal Type (SHN) field to H for hardware.If you wish to verify the calibration, you can return the network analyzer to local8.operation and make a calibrated thru measurement.

Saving the Calibration Data in IC-CAP

When the calibration is complete, save the model file with the measured standards data init, so that you can always quickly retrieve the error coefficients and re-enter them into thenetwork analyzer:

Select File > Save As.1.In the dialog box that appears, type in an appropriate pathname and filename.2.Examples of possible file names might include the calibration date or frequency, suchas TRLCAL_101597.mdl or TRLCAL_6GHz.mdl. Select OK. First the CW cal is saved inone analyzer register using the variable cal_reg_cw; then the swept cal is saved inanother register using the variable cal_reg_sw.When you want to retrieve the error coefficients at some future time, read in the .mdl3.file you saved. Select the thru setup for either the swept or CW cal. Make sure theAgilent 8753 instrument state corresponds exactly to the instrument options in thethru setup. Select the CalHP8753 transform, and Execute. IC-CAP will calculate theerror coefficients, download them into the network analyzer, and turn on the analyzercalibration. Press LOCAL on the analyzer to return it to front panel control, then pressSAVE and a register softkey to save the cal in one of the analyzer's memoryregisters.

This ends the procedures for TRL* calibration under IC-CAP control.

Date post:	05-Jun-2020
Category:	Documents
Upload:	others
View:	19 times
Download:	0 times