41
INSTRUCTION MANUAL TC 170/TC 171 FET SPECTROSCOPY’PREAMP’~
INSTRUCTION MANUAL
TC 170/TC 171
FET SPECTROSCOPY’PREAMP’~
1. Wiring of the power connectordirectly compatible with the following
of this preamplifier ismain amplifiers:
TENNELEC TC 240 SeriesTENNELUC TC 205A with serial numbers above 1999TENNELEC TC205A with serial numbers up to 1999 if the
connector labelled "OTHER" is used.TENNELEC TC 222 when the internal cable is plugged into the
"OTHER" connector .All standard Aptec, Canberra, EG&G Ortec and PGT
In addition to the power leads, this TENNELEC preamplifiercontains signal and test-pulse coaxial cables. These cables areused when the preamplifier is connected to TENNELEC TC 240 seriesamplifiers and to TENNELEC TC 205A amplifiers with serial numbersabove 1999, thereby avoiding the need to use separate ones. Withthe TENNELEC TC 222, TC 205A with serial numbers up to 1999, andall other amplifiers, signal and test-pulse cables separate fromthe power cable must be used.
Differentially-driven cables are used in TENNELEC preamplifiersfor ground-loop noise reduction. See Section 3.2.3 for details.
If there are any questions regarding the compatibility of thepower connector of this instrument, please contact the TENNELECMarketing Department for assistance.
* * * * * * * * * * * * * * WARNING ******x******** ***
Improper connection to the shaping amplifier preamplifier ** power connector may permanently damage the amplifier *
and/or preamplifier.* TENNELEC assumes no liability for *such instrument damage. *
* **********************************
2. The TC 170/TC 171 Charge-Sensitive Preamplifiers are shippedwith a network installed that protects the input field-effecttransistor (FET) from damage due to abnormal transients (seeSection 3.3.2 for a detailed discussion of this topic). Thisnetwork slightly degrades noise performance and risetime, withthe degradation becoming worse as detector capacitance increases.However, in systems where performance is controlled by thedetector rather than the preamplifiers (usually the case withroom temperature surface-barrier detectors and gas filledproportional counters), the degradation may be negligible. Whereultimate performance is required, the network can bedisconnected. See Section 3.3.1 for details.
3. The detector load in the TC 170 is comprised of threeresistors connected in series, and in the TC 171, two resistors.With room temperature surface-barrier detectors, detector leakagecurrent maybe high enough to cause excessivevoltage dropintheload network, requiring one or more of the resistors to bejumpered. See Section 3.3.3 for details.
TABLE OF CONTENTS
1.0 INTRODUCTION . . . . . . . . . . . .
2.0 SPECIFICATIONS . . . . . . . . . . .
2.1 PERFORMANCE . . . . . . . . . .
2.2 CONNECTORS . . . . . . . . . .
2.3 POWER REQUIREMENTS . . . . . .
2.4 OTHER INFORMATION . . . . . . .
3.0 INSTALLATION . . . . . . . . . . . .
3.1 POWER . . . . . . . . . . . .
3.2 CONNECTIONS . . . . . . . . . .
3.2.1 DETECTOR CONNECTION . .
3.2.2 BIAS SUPPLY CONNECTION .
3.2.3 ENERGY OUTPUT CONNECTION
3.2.4 TIMING OUTPUT CONNECTION
3.2.5 TEST INPUT CONNECTION .
3.3 GENERAL PRECAUTIONS . . . . .
3.3.1 FET PROTECTION . . . . .
3.3.2 APPLYING BIAS VOLTAGE .
3.3.3 DETECTOR BIAS NETWORK .
3.3.4 BIAS LEVEL . . . . . . .
4.0 ADJUSTMENTS . . . . . . . . . . . .
4.1 DC OFFSET VOLTAGE . . . . . . .
4.2 RISETIME . . . . . . . . . . .
. . . . . . . . . 1
. . . . . . . . . 2
. . . . . . . . . 2
. . . . . . . . . 6
. . . . . . . . . 7
. . . . . . . . . I
. . . . . . . . . 7
. . . . . . . . . 7
. . . . . . . . . 7
. . . . . . . . . 7
. . . . . . . . . 8
. . . . . . . . . 9
. . . . . . . . .12
. . . . . . . . .13
. . . . . . . . .14
. . . . . . . . .14
. . . . . . . . .14
. . . . . . . . .15
. . . . . . . . .16
. . . . . . . . .17
. . . . . . . . .17
. . . . . . . . .18
2.1
2.2
2.3
2.4
3.1
4.1
4.2
_ 5.1
. 5.2
5.3
5.4
5.5
5.6
6.1
6.2
6.3
TABLE OF FIGURES
Electronic Noise vs Equivalent DetectorCapacitance (Shaping Time of 2.0 usec.) . . . . . . 3
Electronic Noise vs Equivalent DetectorCapacitance (Shaping Time of 0.5 usec.) . . . . . . 3
Typical Risetime vs Equivalent DetectorCapacitance . . . . . . . . . . . . . . . . . . . . 4
Typical Transient Response, TC 170 . . . . . . . . 4
Balanced Signal Cable System . . . . . . . . . . . 11
TC 170 Adjustment Locations . . . . . . . . . . . . 17
TC 171 Adjustment Locations . . . . . . . . . . . . 18
Electronic Noise vs Equivalent DetectorCapacitance (TC 170, 2.0 usec Shaping Time) . . . . 21
Electronic Noise vs Equivalent DetectorCapacitance (TC 170, 1.0 usec. Shaping Time) . . . 22
Electronic Noise vs Equivalent DetectorCapacitance (TC 170, 0.5 usec Shaping Time) . . . . 23
Electronic Noise vs Equivalent DetectorCapacitance (TC 171, 2.0 usec Shaping Time) . . . . 24
Electronic Noise vs Equivalent DetectorCapacitance (TC 171, 1.0 usec Shaping Time) . . . . 25
Electronic Noise vs Equivalent DetectorCapacitance (TC 171, 0.5 usec Shaping Time) . . . . 26
The Shape of a Typical Spectral Line (Gaussian) . . 28
Pulse Shape at the Output of a Preamplifier . . . . 30
The Waveform at the Output of a Preamplifier withPulses Applied to the Input in Rapid Succession . . 30
TC 170 SCHEMATIC DIAGRAM (Sheet 1 of 1) . . . . . . . . . . 36
TC 171 SCHEMATIC DIAGRAM (Sheet 1 of 1) . . . . . . . . . . 37
1.0 INTRODUCTION
The TENNELEC TC 170 and TC 171 are low-noise, fastrisetime charge-sensitive preamplifiers designed toprovide optimum energy and timing performance from anycharged particle detector.
The TC 170 is ac coupled and intended for use withdetectors having an equivalent capacitance from 0 pF to100 pF, including room-temperature silicon detectorsand low-energy, low-voltage proportional counters.
The TC171is also ac coupled and intended for use withdetectors having an equivalent capacitance of 100 pF ormore, including room-temperature silicon detectors.
Each of the preamplifiers consists of a single, charge-sensitive feedback loop which affords minimal noisedegradation of the input field-effect transistor (FET),very fast risetime, and excellent power supply noiserejection. Additionally, both preamplifiers provide afast, transformer-coupled timing output (differentia-tion time constant of 100 nsec) which may be directlycoupled to most timing instruments. Signal polarity ofthe timing output is opposite to that of the energyoutput.
An SHV connector is provided for introducing up to 2kVof detector bias through a load-resistor network. Inthe TC 170, the network consists of two 100 megohmresistors and a 10 megohm resistor, all three inseries. In the TC 171, one each 100 megohm and 10 meg-ohm resistors are series connected. These resistorsmaybe shorted across allowing a trade-off between theconflicting requirements of high resistance for goodnoise performance and low resistance for detectors withhigh leakage current. See Section 3.3.3 for details.
Each preamplifier has a removable FET protectionnetwork. See Section 3.2 for details.
Also included in each preamplifier is a screwdriver-adjustable resistor for optimizing the risetime withvarious detectors. Access is through a hole in the topof the case (the hole is covered with a press-in plug).
2.0
2.1
-2-
SPECIFICATIONS
PERFORMANCE
NOISE
MODEL
TC 170
TC 171
(1)
(2)
(FWHM referred to a silicon detector with w =3.6 eV per electron-hole pair).
DETECTOR NOISE (keV FWHM)CAPAC. TYPICAL(l) MAX.(l)(PF) t s = 2.0 t, = 0.5 t, = 2.0
0 0.95 1.2 1.410 1.1 1.52050
ia: 1.93.1
100 2:8 5.1 3.3
TYPICAL INTERCEPT: 0.95 keVTYPICAL SLOPE: 18.0 eV/pF
10: 2.3 3.2 6.0 4.2 4.0200 4.5 8.2300 5.6 10.0500 7.0 15.0
1,000 13.0 26.0 15.0
TYPICAL INTERCEPT: 2.34 keVTYPICAL SLOPE: 10.9 eV/pF
RISETIME(2)hE& 05b-90%)
. MAX.
3.53.54.2
2::
7.07.57.99.0
12.016.0
II" t " refers to shaping time in usec.TEfiNELEC amplifiers such as the TC 205A,
WithTC 222
and earlier designs, the peaking time (measuredfrom the 1% level) is approximately twice theshaping time as indicated on the front panel. SeeFigures 2.1 and 2.2 for graphs of TC 170/TC 171,Noise versus Equivalent Detector Capacitance atshaping times of 2.0 usec and 0.5 usec.
Based on a 10 MeV equivalent input, risetimeadjustments optimized at each measurement, E and Toutputs terminated in 50 ohms, measurements madewith an external test-input capacitor. See Figure2.3 for a graph of the TC 170/TC 171, Risetime asa Function of Equivalent Detector Capacitance.The typical transient response of the TC 170 isshown in Figure 2.4.
5.0
10.0
10.0
20.0
Figure 2.1 Electronic Noise vs Equivalent DetectorCapacitance (Shaping Time of 2.0 usec.)
.
I* I J .* 10 10 IO 1.o $0 100 200 SW 400 100 ,DooEOUWALENT DETECTOR CAPICITAHCE ,PF,
Figure 2.2 Electronic Noise vs Equivalent DetectorCapacitance (Shaping Time of 0.5 usec.)
Figure 2.3 Typical Risetime vs Equivalent DetectorCapacitance
E-OUT
V = 100 mV/Div
H = 20 nsec/Div
T-OUT
Figure 2.4 Typical Transient Response, TC 170
SENSITIVITY(NOMINAL)TC 170 44 mV/MeV, SiTC 171 20 mV/MeV, Si
FEEDBACK CAPACITORTC 170 1 PFTC 171 2.2pF
-5-
DECAY TIME CONSTANT(NOMINAL)E OUT TC 170
TC 171
T OUT TC 170, TC 171
DETECTOR LOAD RESISTORTC 170TC 171
SIGNAL POLARITYINPUT
E OUTPUT
T OUTPUT
TEST CAPACITORTC 170TC 171
DETECTOR BIAS
NONLINEARITY, INTEGRAL(t = 2usec, E OUTungerminated)0 to +1ov
0 to &7V
NONLINEARITY, DIFFERENTIAL(t = 2usec, E OUTungerminated)
MAXIMUM ENERGYTC 170TC 171
COUNT RATE CAPABILITY(5% of pulses innon-linear range)
TC 170
TC 171
1 msec2.2 msec
100 nsec
1OOM + 1OOM + 10M1OOM + 10M
Either
Inverse of input
Same as input
1 PF2.2 pF
12,OOOV Maximum
10.05% max., 10.03% typ.
10.02% max., 10.005% typ.
10.03% typical, -9V to +7V
200 MeV @ 5 c/s Si400 MeV @ 5 c/s Si
1.6 x lo7 c/s @ 1MeV Si1.5 x lo5 c/s @ 1OMeV Si
3.6 x lo7 c/s @ 1MeV Si3.5 x 105 c/s @ 1OMeV Si
2.2
-6-
CONVERSION GAINTEMPERATURE STABILITY
TC 170
TC 171
DYNAMIC INPUT CAPACITANCETC 170
TC 171
CONNECTORSINPUT (DETECTOR)
H.V. IN (DETECTOR BIAS)
TEST IN
E OUT
T OUT
POWER
I k50 ppm/OC, 0-50°C
I+75 ppm/OC, 0-50°C
240,000 pF minimum280,000 pF typical
260,000 pF minimum2200,000 pF typical
BNC (UG-290/UA),accoupled
SHV (AMP 51494-2),+2kV maximum
BNC (UG-1094/U),Zin=SO ohms
BNC (UG-1094/U),dc coupled,20 = 50 ohms +l%, dc offsetapproximately -lOOmV
BNC (UG-1094/U), ac coupled,20 = 50 ohms +l%, decay constant= 100 nsec +lO%
g-pin male, Amphenol17-20090 or equivalent
m The power connector includes connections for acable containing 512V wires ,wire,
+24V wires, a power groundand three RG 174/U cables. One cable is the
signal cable which duplicates the function of the E-OUTconnector but is isolated from it by a separate SO-ohm+l% terminating resistor. Another cable is groundedthrough a 50 ohm +l% resistor and constitutes thesource of out-of-phase ground-loop noise signal whenthe preamplifier is used with a differential-input mainamplifier,signal.
and the third cable carries the test pulseThis last cable is in parallel with the TEST
IN connector and shares a common 50 ohm &l% terminatingresistor.
.
2.3
2.4 OTHER INFORMATION
POWER REQUIREDTC 170
TC 171
WEIGHT(SHIPPING)(NET)
DIMENSIONS
WARRANTY
-7-
+24V @ 35mA; +12v @ 12mA-24V @ 15m~; -12V @ 12mA
+24V @ 65mA; +12v @ 12mA-24V @ 15mA; -12V @ 12mA
3.0 lbs (1.4 kg)1.1 lb (0.5 kg)
(L x W x H) 4 x 3 x 1.5 inches;10.2 x 7.6 x 3.8 cm.
One year
3.0
3.1
INSTRUCTION MANUAL One provided with each instrumentordered.
ACCESSORIES One (1) TENNELEC NC-PAC-10, 10 ft.preamplifier signal and power cableprovided with each preamplifierordered; Amphenol 17-20090 toAmphenol 17-10090 connectors.
INSTALLATION
POWER
The TC 170/TC 171preamplifiers are not self-poweredand must be connected via the power cable to a mainamplifier with provisions for providing preamplifierpower or a separate preamplifier power supply. Referto the CAUTION at the beginning of the manual beforeconnecting the TC 170/TC 171 power cable to TENNELECmain amplifiers other than the TC 240 series.
3.2 CONNECTIONS
3.2.1 DETECTOR CONNECTION
To preserve the low-noise characteristics of thesystem, the capacitance to ground at the input of thepreamplifier should be kept minimal. If a cablebetween detector and preamplifier must be used, it
-a-
should be as short as possible and it must be shieldedwith one end of the shield connected to the detectorhousing and the other to the preamplifier housing.Doubly shielded cable (RG71/U) is preferable to singlyshielded cable (RG62/U).
To avoid microphonics, it is desirable that thegeometrical relationship between detector andpreamplifier be kept rigidly fixed.
The length of cable connecting the preamplifier and thedetector should be kept to a minimum for reasons ofstability in addition to noise considerations. Thecable connecting the preamplifier and detectorintroduces a phase shift into the preamplifier feedbackloop which adversely affects stability. The cableshould be kept as short as possible. A maximum lengthof cable cannot be assigned to the infinite number ofdetector and cable combinations, but a maximum lengthfor the TC 170 is typically 3 ft. and for the TC 171,2 ft.
The noise performance of the,preamplifier can beestimated from the sum of connecting cable anddetector capacitance. The noise as a function of thisinput capacitance is shown in Figures 2.1 and 2.2 forshaping times of 2 usec and 0.5 usec.
3.2.2 BIAS SUPPLY CONNECTION
In the TC 170/TC 171, the bias connection is routedthrough the preamplifier case.
If a battery pack is used for bias, no specialprecautions need be taken in cable routing to avoidnoise pickup provided the case of the battery pack isconnected to the preamplifier.
If a power line operated supply is used, or if abattery pack is used which is grounded to the mainamplifier frame, then it is desirable to take thefollowing precautions to avoid ground-loop pickup.
a. Locate the power supply physically close tothe main amplifier.
b. Ground the supply to the main amplifier withlarge-gauge wire or shield braid at leastl/4" (6mm) wide.
-9-
C . Cut the high voltage cable to approximatelythe same length as the preamplifier signalcable and twist or tape the two together.
d. Never plug the main amplifier and HV supplyinto different wall outlets. If necessary,use a local distribution box for allcomponents of a spectrometer system to avoidmaking the building part of a ground loop.
3.2.3 ENERGY OUTPUT CONNECTION
The energy output (E out) is intended to drive a 50ohm line (RG 58A/U) which may be connected directly tothe input of the main amplifier. A 50 ohm terminationis not required as the preamplifier is st,ableunterminated. The preamplifier will drive any lengthof cable, but for long cable lengths, cable losses mustbe considered.
To avoid ground loop pickup, the philosophy at TENNELECis to route signal and test pulse cables through thesame shielded wire bundle as the supply voltage wires,terminating all connections in a single multipinconnector (Amphenol 17-10090 or equivalent) whichattaches to the main amplifier. This feature can beused with the "OTHER" preamplifier power connectorson the TC 205A shaping amplifier having serial numbersgreater than 1999. The preamplifier signal only (notest input) is available through the preamplifiershielded wire bundle to the TC 240 and TC 241. Whenusing main amplifiers of other manufacture, it isrecommended that the preamplifier power connector onthe amplifier be rewired to accept the preamplifiersignal and provide the preamplifier test input. Thiswill minimize the ground loop problem by allowing thesignal and test pulse lead in the power cable to beused.
If independent signal, test pulse, and power cables areused, the following pattern of connections isrecommended:
a. Place the test pulse generator as close tothe main amplifier as possible.
- 10 -
b. Cut the signal and test pulse cables toapproximately the same length as the 2 12V,+24V wires. Twist or tape all cables into onebundle.
The purpose of instruction (a) is to prevent powersupply noise spikes (which nearly always exist betweenwidely spaced ground points in a NIM bin) fromappearing in series with the signal ground returns.The purpose of instruction (b) is to avoid local radiostation pickup which frequently occurs because of theloop-antenna effect in a network of spread-out cables.
When a long cable run is necessary in an electricallynoisy environment, a balanced-to-ground signal cablesystem may be needed to reduce noise pickup to anacceptable level. In this arrangement (Figure 3.1),two signal cables of matched length and close proximityare used to connect the preamplifier to a differentialinput stage on the main amplifier. This feature isdirectly compatible with TC 205A's of serial numbergreater than 1999. The "OTHER" power connector on TC205A's with serial numbers up to 1999 will requirerewiring to provide this feature. The preamplifierpower connector ontheTC 222 canbe modifiedtoacceptthis function if the preamplifier power connector ismodified to provide + 12V and + 24V (the TC 222preamplifier power connectors must be configured toprovide + 12V and + 24V for the TC 170 series ofpreamplifiers to operate). As shown in Figure 3.1 onlyone of the cables carries the desired signal, but bothcables carry the noise signal. At the balanced inputstage, the noise signals cancel, but the desired signalis unaffected.
In the TC 170 Series preamplifiers, provision forbalanced operation is built into the preamplifier andthe ten-foot power cable. To take advantage of thisfeature, the user must be sure that the main amplifiercontains a balanced input stage and that Pins 3 and 8of the cable connector (Figure 3.2) are connected tothe respective inputs. Pin 2 is the signal groundconnection. In a single-ended main amplifier, Pin 2should be grounded.
- 11 -
PREAMPLIFIER
Figure 3.1 Balanced Signal Cable System
f INVERT
DUYYV OUT- - - - - - - - - -- - - - - - - -m----m-m---mREAMP C”T
f DIRECT
Figure 3.2 Main Amplifier Preamplifier Connector andPolarity Selector
Signal cables are terminated at the sending end byresistors R33, R38, and R40 as shown on the TC 170schematic. In the TC 170 Series, these resistors arenominally 49.9 ohm each to match 50 ohm cables. Othercable impedances may be used by changing to resistorsof appropriate value.
For best possible noise performance, linearity,transient response, and for the least possible heatingin the output transistors, the receiving end (mainamplifier end) of the signal cables should operate intoan essentially open circuit--500 ohms or more.
If a signal cable external to the one in thepreamplifier's power cable is used to join the E OUT
- 12 -
3.2.4
connector on the preamplifier to the INPUT connector ofthe main amplifier, a wrinkle in the waveform may beobserved 3.2~ nanosec from the start of the pulse,where L is the length in feet of the power cable. Fora standard 10' cable, this wrinkle will occur 32 nsecfrom the time-zero reference, but will be of noconsequence in normal operation. The distortion iscaused by signal reflection in the revised coax, andthe effect can be eliminated by removing ~35 from theTC 170 (R38 from the TC 171). Removing the resistorisolates the unused signal cable.
TIMING OUTPUT CONNECTION
The timing output is intended to drive a 50 ohm systemand should be connected using coaxial cable such as RG '58A/U. The timing output may be directly connected toa fast amplifier, timing filter amplifier, or fastdiscriminator. The polarity of the timing output isthe inverse of the energy output. When not being used,the timing output should be terminated in 50 ohms forbest pulse response, although this is not necessary ifsome leading-edge pulse distortion can be tolerated.
Timing measurements usually require cleaner waveformsthan energy measurements. If the TC 170 or TC 171 isused exclusively for timing measurements, energy-signalcables should be disconnected to avoid reflections.The signal cable within the TC 170 preamplifier powercable can be disconnected by removing R35, (R38 for theTC 171).
If energy and timing measurements must be madesimultaneously, waveform purity can be maintained byterminating the energy-signal cable in 50 ohms at themain-amplifier end. This applies both to the signalcable included ip the power bundle OK to an externalsignal cable, if used. In the latter case, the cablein the power bundle should be disconnected by removingthe appropriate resistor (R35 for the TC 170 and R38for the TC 171).
If the signal cable is terminated, the linearity willbe degraded slightly when the product of count rate andsignal amplitude approaches the upper limit ofpreamplifier dynamic range.
For a discussion of where on the waveform signalreflections appear, see the last paragraph of 3.2.3.
- 13 -
3.2.5 TEST INPUT CONNECTION
Test pulses may be applied either at the rear of thepreamplifier or at the main amplifier end of the powercable. In either case, the test pulse connector isinternally terminated by a 50 ohm resistor.
The test input capacitor of the TC 170 is 1pF and thetest input capacitor of the TC 171 is 2.2 pF. A pulsegenerator such as a TC 812 can be connected to the testinput of the TC170/TC171and be used to verify systemoperation and calibration. The detector (with biasapplied) or an equivalent detector capacitance shouldbe connected to the TC 17O/TC 171 input when verifyingsystem performance or calibration.
The transient response of the TC 170/TC 171 can best beexamined by applying the test signal through anexternal charge coupling capacitor to the INPUTconnector. Because of stray capacitive coupling andreflections from the test cable in the power supplybundle, this method will result in a more accuraterepresentation of the transient response than using theTEST IN connector.
If a test signal must be applied other than through anexternal charge coupling capacitor, the accuracy of thetransient response can be improved by following certainpractices. If the preamplifier is used with a TENNELECmain amplifier with a TEST INPUT connector on the frontpanel and fully compatible as outlined at the beginningof this manual (CAUTION), the test input signal shouldbe applied at the main amplifier. When thepreamplifier is used with a main amplifier not havingthe above feature, the test input must be applied atthe TEST IN connector of the preamplifier. When usingthe TEST IN connector, the accuracy of the transientresponse will be improved if the wire connecting theTEST IN connector to pin 5 of the power connector viathe printed circuit board is disconnected.Disconnecting this wire will eliminate the reflectioncaused by the unterminated test cable in the powersupply bundle.
- 14 -
3.3
3.3.1
GENERAL PRECAUTIONS
FET PROTECTION
In the TC 170/TC 171, a transistor connected as a diodeis used to protect the input FET against damage fromaccidental short circuits. The protection networkdegrades the preamplifier noise slightly (refer toSection 5.1 and Figures 5.1 thru 5.6 for more detail).After the system is operable, the user may wish todisconnect the diode. This is easily done by pullingthe diode lead (J2) out of the miniature jack mountedon the input Teflon standoff post and inserting thejumper (Jl) across the series protection resistor. Seethe component placement drawing atthe end of thismanual for location of the jumper storage position andlocation of Jl and J2.
IMPORTANT: See Section 7.1 for instructions on how toproperly open or remove the case.
If no resolution improvement is observed once the diodeis disconnected, it is strongly recommended that thediode be reconnected. If left disconnected, the diodelead should be bent down and away from the inputterminal.
3.3.2 APPLYING BIAS VOLTAGE
In the following statements, it is essential that theuser recognize the distinction between rapid voltagechanges at the H.V. IN connector and the SIGNAL INPUTconnector.
The TENNELEC TC 170/TC 171 preamplifiers can safelywithstand the application of detector bias voltage in&5OOV steps spaced 10 set apart, with or without diodeprotection for the FET, if the voltage is appliedthrough the H.V. IN connector.
Without diode protection, short circuits at the SIGNALINPUT terminal may cause FET damage if the bias exceeds5OV. Connecting a preamplifier to a detector with biasvoltage applied, either through the preamplifier ordirectly to the detector, is nearly equivalent to ashort circuit.
- 15 -
\the SIGNALif the bias
qith diode protection, an occasional short circuit atINPUT terminal will not cause FET damageis 500V or less.
With or without diode protection, the preamplifier maybe disconnected from a charged detector, but notreconnected except within the limits stated above forshort-circuiting the SIGNAL INPU~T terminal. When apreamplifier is disconnected from a charged detector,the SIGNAL INPUT terminal should not be short-circuitedexcept within the limits given above. Additionally,the detector bias supply should not be disconnectedwithout first reducing the voltage to zero in 500Vsteps spaced 10 set apart and then waiting for anadditional minute to allow the preamplifier filternetwork to discharge. The reason for this lastprecaution is that without a return path through thepower supply, neither the filter capacitors nor theinput coupling capacitor will have a discharge path; itmay take an hour for discharge to occur through leakageresistance alone.
The user is reminded at this point that because of thevulnerability to accidental damage, the FETs are notcovered by warranty.
3.3.3 DETECTOR BIAS NETWORK
Noise due to the detector load resistor diminishes asthe resistance increases, with the contributionbecoming negligible above approximately 100 megohm.Unfortunately, because of leakage current in roomtemperature surface-barrier detectors and because thisleakage current is temperature dependent (doubles foreach 8'-10°C increase in temperature), there is anupper limit to the size of load resistor which can besafely used. Usually, a maximum IR drop of 1OV can betolerated. A leakage current of 1uA through a 10 megohmload will produce such a drop.
The TC 171 is shipped with two resistors totaling110 megohm as the detector load resistor. These canbe reduced to 100 megohm and 10 megohm by the user.(See Section 7.3 for modification details.) If thedetector load resistors supplied as standard areinappropriate, the user can install a different value.The resistor chosen should have low end-to-end
- 16 -
capacitance and low dielectric losses. This isimportant only when working with low capacitancedetectors.
3.3.4 BIAS LEVEL
If a system containing a room-temperature surface-barrier detector is assembled and turned on in theabsence of bias voltage , the electronic noise will behigh. As the bias voltage is increased, the noise(observed as "grass" on an oscilloscope connected tothe output of the main amplifier) should drop sharply.As the voltage is increased further, the noise shouldcontinue to drop up to the point where rated detectorvoltage is reached, then it should increase again.However, the appearance of this noise at this biaslevel will not be clean "grass" as observed earlier,but as a series of discontinuities on the baseline.This later appearance is characteristic of avalanchebreakdown in the detector. The correct operatingvoltage for the detector is about 10% below thisavalanche level. If the load resistor is too high forthe leakage current of the detector, the noise levelwillnotdrop as the bias voltage is increased until itexceeds the IR drop in the load resistor. In extremecases, this will not occur until the bias voltage asindicated on the power supply exceeds the detectormanufacturer's specified maximum.
QAQZLQm The user is encouraged to discuss thedetector breakdown characteristics with themanufacturer of the detector. TENNELEC cannot assumeresponsibility for damage to the detector caused byimproper load-resistor selection or by improperapplication of detector bias voltage. With somedetectors, permanent damage will result from over-voltage.
With time and radiation damage, the onset of avalanchenoise (also known as "flicker" noise) may drop to alevel below the ratings of the detector, requiring areduction of operating voltage for acceptable energyresolution (or background count rate).
4.0
4.1
- 17 -
ADJUSTMENTS
DC OFFSET VOLTAGE
The TC 170/TC 171 have adjustments for the DC offsetvoltage as measured at the energy output. The DCoffset adjustment of the TC 170 is accessible through ahole in the preamplifier case as shown in Figure 4.1.This hole is normally covered with a press fit plug toreduce electrical pickup. The DC offset potentiometeris R7 as shown on the TC 170 schematic at the end ofthis manual. This adjustment effects the drain currentof the input PET (Cl) and therefore the noiseperformance. To ensure the optimum noise performance,the DC offset should be adjusted for -100 mV of offsetvoltage as measured at the energy output connector.This measurement is made with the energy outputconnector unterminated. The timing output connectortermination does not effect this measurement. The DCoffset voltage of the TC 170 was adjusted to -1OOmV(+.5OmV) before leaving the factory and should notrequire further adjustment unless the input FET ischanged.
I T@iNwCT@iNwC
ODC OFFSET ADJ.ODC OFFSET ADJ.
0 RISETIME ADJ.0 RISETIME ADJ.
Figure 4.1 TC 170 Adjustment Locations
The TC 171 DC offset adjustment is accessible through ahole in the preamplifier case as shown in Figure 4.2.This hole is normally covered with a press fit plug toreduce electrical pickup. The DC offset potentiometeris R37 as shown on the TC 171 schematic at the end ofthis manual. This adjustment does not affect the FETdrain current or the noise performance. To reduce
r
- 18 -
heating effects of the output stage due to DC offsetvoltage, the DC offset should be adjusted to zero (0)~100 mV as measured at the energy output connectorunterminated. The timing output connector terminationdoes not effect this measurement. The DC offset voltagewas adjusted to zero before leaving the factory andshould not require further adjustment unless the inputFET's are changed.
0 RlSEflME ADJ.
DC OFFSET ADJ.0
4 . 2
Figure 4.2 TC 171 Adjustment Locations
RISETIME
The TC 170/TC 171 have risetime adjustments as shown inFigures 4.1 and 4.2. The risetime adjustmentcorresponds to Rll on the TC 170 schematic and R9 onthe TC 171 schematic at the end of this manual. Toobtain optimum transient response and timingperformance, the risetime adjustment must be set withthe detector (or an equivalent detector capacitance)and any connecting cable connected to the preamplifierinput, with bias applied (if an equivalent detectorcapacitance is used in place of the detector, no highvoltage bias is necessary). When minimum risetime isobtained, a short period overshoot of 10% to 20% willbe observed. This overshoot has no effect on noiseperformance or linearity within the normal dynamicrange. Additionally, the overshoot will have no adverseaffect on timing measurements. Further improvement intiming performance can be obtained by adjusting therisetime control for a shorter risetime and moreovershoot. The increased overshoot will be accompaniedby some ringing which limits the minimum risetime.
- 19 -
Some ringing can be tolerated if the discriminator hasan adjustable deadtime control.
NOTE: Detector inductance has a very important bearingon preamplifier signal risetime and circuit stability.It is essential when operating the preamplifier for thefirst time to match it to detector characteristics. Todo this, connect a fast oscilloscope to the output ofthe preamplifier and excite the detector with a weak!aiation source that generates only one spectral line.
PO is a good source for this purpose with roomtemperature silicon detectors. The oscilloscope shouldhave a bandwidth of 250 MHz or more, and should betriggered internally. A continuously changing ensembleof randomly occurring pulses will be seen,observation of the pulse difficult.
makingUnfortunately, not
much can be done about this.
The adjustment is made by turning the risetime controlto obtain the optimum balance between ringing andrisetime.
Depending on the equivalent circuit of the detector,some settings may cause oscillation. This is normal.A ~stubborn case of oscillation indicates excessivedetector inductance or connecting cable length.
If the oscillation is difficult to eliminate, orelimination results in slower than expected risetime,it is recommended that the connecting cable be reducedto as short as practical. If this does not eliminatethe oscillation, the built-in series resistor R9 can beadded to the circuit by removing jumper Jl located atthe input connector.is inadequate,
If the original value of 22 ohmsit may be increased at the expense of
risetime and noise performance. The TC 170 normallywill be stable with up to 3 feet of cable betweendetector and preamplifier , and up to 2 feet with theTC 171.
Typical risetime versus detector capacitance is givenin Section 2.1. Due to the fast risetimes of the TC170/TC 171, it is recommended that a pulser andoscilloscope with 1 nsec or less risetime be used tocheck and adjust the transient response when not usinga detector and source. The actual preamplifierrisetime (trpa) is
t rpa = (tra2 -trb2N2
5.0
5.1
- 20 -
is the risetime of the preamplifier, pulseroscope, and trb
and oscilloscope only.is the risetime of the pulser
above,By using the equation given
the risetime contributions of the pulser andoscilloscope can be removed. All risetimespecifications apply when the timing output isterminated in its characteristic impedance (50 ohms)and is independent of the energy output termination (noload or 50 ohms).
NOISE
NOISE PERFORMANCE
To convert from full width-at-half-maximum (FWHM) Si toa different reference,~~~use Table 1.
TABLE 1
CONVERSION OF FWHM Si TO OTHER REFERENCE VALUES
Reference 1 Multiply eV FWHM Si by:
FWHM Si 1.00 (W=3.6 eV/electron-hole pair)FWHM Ge 0.819 (WE2.95 eV/electron-hole pair)FWHM PlO 6.94 (W=25 eV/electron-hole pair)Ion pairs rms 0.144Coulombs rms 2.3 x lO-2o
The noise performance of the TC 170/TC 171 for ashaping time of 2.0 usec and 0.5 usec is given inSection 2.1. The noise performance of the TC 170 andTC 171 at various shaping times and detector loadresistance, with and without the protection network, isshown in Figures 5.1 thru 5.6. With the aid of thisdata, the noise performance of the preamplifier can bepredicted for almost any combination of detectorcapacitance and shaping time. It is stressed that theforegoing figures are noise levels and not spectralresolution. The final spectral resolution depends notonly on the preamplifier noise but also on the type ofdetector used, the count rate, and other factors. Anadditional consideration in evaluating the preamplifiernoise limitations is the detector leakage current. Allthe previous noise data is representative of thepreamplifier detector combination with zero detectorleakage current.
EQUIVALENT DETECTOR CAPACITANCE
9.0 I I I II I I I Ill1-._
7.0 I I I I1111 I I I I
5.0 I I I II111 I I I I
I I
. .
EQUIVALENT DETECTOR CAPACITANCE (pF)..,
Figure 5.1 Electronic Noise. vs Equivalent Detector Capacitance(TC 170, 2.0 usec Shaping Time)
.
-
-
-
-
- 22 -
-
-
--
-
--
-
L
(IS IrYHMd AW 3SION
.o -
.O
.o -
. o -
. o -
. o -
.o -
. o -
.o -
. o r
.o -
.9 -
.8 -
.7 -
.6 -
.s -
.4 -
.3 -
.2 -2
TC 170 NOISE (TYPICAL)“S
EQUIVALENT DETECTOR CAPACITANCESHAPlNO TIME - 0.5p**q TC 205A
I I I lllll I I I I
3 45 10 2 0 3 0 40 50
- - - - WIT” PROTECI
1---------
--N
1--------
--- WITHOUT PROTECTION
EOUIVALENT DETECTOR CAPACITANCE (pF)
Figure 5.3 Electronic Noise vs Equivalent Detector Capacitance
(TC 170, 0.5 usec Shaping Time)~
I.-9 I I I lllll I I I I lllll I I .A,-
8 I I II b5 I
EQUIVALENT DETECTOR CAPACITANCE (pF)
Figure 5.4 Electronic Noise vs Equivalent Detector Capacitance
UC 171, 2.0 usec Shaping Time)
. .. Y
!I,‘., ::: ) ‘8
- 25 -
--=I#--- II I
- II I
I I I
I I I
I I I
(!S WHMj A=l) 3SION
I TC 17 1 NOISE (TYPICAL) I ’ I ’ I ‘l’l’l’l’l’T-v-mEQUIVALENT DETECTOR CAPACITANCE
100 SHAPING TIME - 0.5”~~. TC ZOSA
90807060
EOUIVALENT DETECTOR CAPACITANCE (pF)
Figure 5.6 Electronic Noise vs Equivalent Detector Capacitance(TC 171, 0.5 usec Shaping Time)
.
5.2
- 27 -
NOISE MEASUREMENTS
To verify the proper operation of the preamplifier,noise measurements can be made by either of twomethods. These measurements can be compared with thevalues given in Section 2.1 and Figure 5.1thru 5.6.
One method requires a calibrated step-generator (e.g.TENNELEC TC 812), a shaping main-amplifier (TC 205A, TC222, TC 240, or TC 241), and a multichannel pulseheight analyzer. After the shaping time constant hasbeen chosen and the analyzer has been calibrated interms of energy per channel, pulses are fed through thetest capacitor; the line width recorded by the analyzeris measured. For this test, as.for any measurement ofabsolute noise of the preamplifier, the detector shouldbe replaced by a dummy capacitor of the samecapacitance. The full-width-at-half-maximum of theline should be close to the values given in Section 2.1for typical performance and the typical data given inFigures 5.1 thru 5.6. If the noise at 2.0 usec exceedsthe guaranteed values given in Section 2.1, verify thatthe protection network is D.& in and that the correctdetector load resistor is installed.
The second method requires the use of a calibratedpulse generator, a shaping main-amplifier, an average-type ac voltmeter (such as a Hewlett-Packard 400D,400H, or 400~) or a true rms voltmeter (such as aHewlett-Packard 3400A), and a calibrated oscilloscope.A step of known amplitude V. is applied to the inputthrough the test capacitor 8transfer to the input of Y;.
resulting in a chargecoulombs.
resulting main-amplifier puls& heig t Vx cx The
is recordedwith the oscilloscope. The pulse gener"ator is thenturned off, and the true rms noiselevelv is measuredat the output of the main amplifier. I? a true rmsvoltmeter is used, the reading is directly V,. If anaverage-type voltmeter is used, the reading V, shouldbe multiplied by the factor 1.135 to obtain Vn. Thelevel in keV FWHM referred to Si detectors is given by
Noise (FWHM) =Vn x Vi x CT.
VOx 5.298 x 1016
where 5.298 x 1016 is a factor that contains the chargeof an electron in coulombs, the energy necessary toproduce one election-hole pair in silicon, and the
6.0
6.1
- 20 -
conversion constant between rms and FWHM. Fordetectors other than silicon, choose the appropriatemultiplier from Table 1.
If a problem with excessive noise should occur, eitherof the two procedures described should be used toevaluate the noise performance. The preamplifier noiseperformance can be verified by replacing the detectorwith a suitable capacitor having the same capacitancevalue as the detector. If this noise is withinspecifications at 2.0 usec shaping time, or similiar todata given in section 5.1, the problem is associatedwith the detector. The total noise of the system isgiven by
Ntotal '[(Npreamplifier)2 +(Ndetector) 12 l/2
Using the above equation and the noise of thepreamplifier (as previously determined) the noise ofthe detector can be calculated and compared with themanufacturer's data.
COUNT RATE EFFECTS
RESOLUTION
The shape of a typical spectral line is Gaussian and isshown in Figure 6.1.
Figure 6.1. The shape of a typical spectral line(Gaussian).
The resolution, or ability of a nuclear spectrometer toseparate different radiatio~n energies, is usuallyexpressed in terms of the full width of the spectrallines measured at half their maximum height. This
- 29 -
quantity is denoted by the letters FWHM and is given inunits of energy. The FWHM is 2.35 Q where P is thestandard deviation. If noise alone controls theresolution, then d and the rms noise level aresynonymous. We have been using FWHM to characterizepreamplifier noise levels in the previous section.
The resolution obtained in any particular spectralmeasurement is the result of several factors:preamplifier noise, detector characteristics, countrate, radiation energy, overall system stability,proper interfacing between instruments within thesystem, etc. In an experimental situation in which thecount rate is low enough sothatthe pulse shape canbeadjusted for the best signal-to-noise ratio withoutbeing affected by pile-up or baseline shift but highenough so that effects due to long term drifts can beneglected, the resolution will be determined by threefactors: (a) detector resolution for the particularradiation energy being observed, (b) electronic noise,and (c) interfacing. Furthermore, if it is assumedthat the different components of the system areproperly matched, the line-width is a function of onlythe detector resolution and the electronic noise. Thethree magnitudes are, then, related in the followingway:
R2 = (Total Resolution)2 =+ (Electronic Noise)2
(Detector Resolution)'
We shall call R the intrinsic resolution of the system.In a counting situation in which the conditions are notideal, the measured resolution will be worse than theintrinsic resolution. Usually, the main factor inline-width broadening is count rate. Count rate canhave a deleterious effect in spectral resolutionthrough several mechanisms. The three most commonlyfound are pile-up of pulses, baseline shifts, andthermal effects in components. The last two canusually be neglected in properly designed systems; thefirst one is more difficult to contend with.
Usually, pile-up of shaped pulses in the main amplifierwill set the practical upper count rate limit.However, at very high energy (lowest gain settings ofthe main amplifier), the limitation may occur in thepreamplifier. A discussion of preamplifier pile-upfollows, plus the technique of computing the uppercount rate limit.
- 30 -
The pulse obtained at the output of the preamplifierappears as shown in Figure 6.2.
Figure 6.2 Pulse shape at the output of a preamplifier.
When pulses come in rapid succession, the wave form atthe output of the preamplifier appears as shown inFigure 6.3. The dotted line at the top of Figure 6.3indicates the limit of the linear range of thepreamplifier.
Figure 6.3 The Waveform at the Output of a Preamplifierwith Pulses Applied to the Input in Rapid Succession
- 31 -
If the count rate is high enough, some of the pulseswill rise beyond the linear range and therefore, theiramplitudes will be distorted. (The meaning of"linearity" is explained in Section 6.1.) If we assume1OV to be the limit of the linear range, thepreamplifier sensitivity to be 44 mV/MeV, the averageradiation energy to be 2.5 MeV, and the decay timeconstant to be 1 msec, we can compute the count ratethat will be necessary to make 5% of the pulses fallbeyond the linear range from the formula
TdL 2.5 EorGC1
;= count rate in cps.
d = decay time constant in sec.vm = linear range in volts.EU = radiation energy in MeV.GC = preamplifier sensitivity in V/MeV.
Replacing symbolsn = 2.~58 x 10k
y actual numberscps.
Since the TC 170/TC 171 are ac-coupled preamplifiers, acount-rate product cannot be assigned with any usefulunits as the number would apply only for one specificenergy. The maximum count rate at 1MeV and 1OMeV ofthe TC 170/TC 171 are given in Section 2.1.
6.2 NONLINEARITY
If a graph of output signal level Vo vs. input pulseheight Vi (dynamic characteristic) is drawn, aperfectly straight line passing through the originshould result. In practice, the dynamic characteristiccould have a slight curvature up to a certain signallevel, beyond which the curvature increasesdrastically. The onset of this drastic change isusually considered to be the upper limit of the normaldynamic range (rated output).
Integral nonlinearity is defined as the maximumdeviation of the measured preamplifier response fromthe ideal response, expressed as a percentage of therated output (as described in the preceding paragraph).This definition is useful only for isolatedpreamplifier pulses as shown in Figure 6.2. When
7.0
7.1
7.2
7.3
- 32 -
pileup occurs as the result of an ensemble of closelyspaced small pulses (Figure 6.3), we are interested notonly in the integral nonlinearity but also in thedeviation of height of individual steps (tiithin the"linear"height.
range of the preamplifier) from the expectedThis incremental deviation in AVo/AVi from
the value at zero volts on the dynamic characteristicis described as the differential nonlinearity. It isthis definition which is used in the table ofspecifications.
PREAMPLIFIER MODIFICATIONS
TENNELEC representatives will help users withinformation about preamplifier modification. Therepresentative will require details of the desiredmodification, serial number of the preamplifier, typeof detector with which it will be used, approximatedetector capacitance and operating voltage, and typeand energy of the radiation being measured.
UNLESS THE USER IS ADEPT AT MAKING MODIFICATIONS OFTHIS SORT, IT IS STRONGLY RECOMMENDED THAT THEMODIFICATIONS BE PERFORMED AT THE TENNELEC PLANT.
REMOVING THE CASE
Remove the two press fit plugs, remove the fourmounting screws on the bottom of the case, andcarefully remove the preamp from the case.
CHANGING PREAMPLIFIER SENSITIVITY
Reducing the preamplifier sensitivity will almostcertainly cause it to oscillate unless thestabilization networks are changed as well. Forthis change the user is requested to return theinstrument to TENNELEC for modification. Increasingthe sensitivity will not cause oscillation, but maydegrade the risetime and pulse shape. Again, the useris requested to return the instrument to TENNELEC formodification.
FRONT-END MODIFICATIONS
The TC 170 is supplied with two 100 megohm resistorsand one 10M ohm resistor in series to function as adetector load resistor for low leakage room temperaturesilicon detectors. When bias voltage is applied, a
- 33 -
voltage drop will be developed across these resistorsdue to the detector leakage current. If this voltagedrop becomes significant, it should be reduced byremoving one or both of the 100 megohm resistors fromthe bias circuit. Physically removing the resistors isnot recommended. However, these resistors can beelectrically removed by soldering a wire across theresistor from standoff to standoff.
The TC171is supplied with one100 megohm resistor andone 10 megohm resistor in series to function as adetector load resistor. The voltage drop across theload resistors must be taken into consideration as forthe TC 170. Note that reducing the load resistor willraise the preamplifier noise slightly at very lowcapacitance, however, this increase will tend to beminor due to the noise from the leakage current of thedetector. For larger detectors (50 pF and above), theincrease in noise will be further masked by thedominance of the series noise of the preamplifier.
NOTE: If the preamplifier will be used with more than1kV applied to the detector terminal, it is veryimportant that all solder joints in the high voltagechain be smooth-surfaced and with no sharp pointsprotruding. Furthermore, it is important that allcapacitors and high megohm resistors used in this partof the circuit be free of surface contamination.Components that are contaminated can cause increases inpreamplifier noise, leakage current, noise spikes fromarcing, etc. In particular, the feedback resistor andthe 100 megohm load resistors are very fragile andshould not be touched with a hot soldering iron, sharp-pointed tool or bare fingers (body oil may increase theleakage significantly across these resistors).Similiar precautions should be taken when working withthe feedback capacitor, the test capacitor, the inputcoupling capacitor, the FET heatsink, and the Teflonstandoffs. These components can be safely cleaned withmethanol using a camel hair brush or a clean cottonswab.
7.4
- 34 -
MODIFICATION OF FET DRAIN CURRENT
The PET drain current of the TC 170 can be adjusted bychanging R7. However, this will affect the dc offset,which should never be set to less than -50mV dc. Thenormal dc offset voltage is -lOOmV, which is close tothe value resulting from operation of the invut FET at
Best noise-performance is obtained at or near2eSsidss level.
8.0
9.0
10.0
The FET drain current of the TC 171 is set by R6 atapproximately 40ma. The drain current can be loweredwith a resultant increase in noise, if powerconsumption is a major consideration. If the draincurrent is changed, the dc offset voltage must bereadjusted to zero + 1OOmV using R37.
SHIPPING DAMAGE
Upon receipt of the instrument, examine it for shippingdamage. Damage claims should be filed with thecarrier. The claims agent should receive a fullreport; a copy of that report should be sent toTENNELEC, Inc., P.O. Box D, Oak Ridge, Tennessee37830. The model number and serial number of theinstrument must be included in the report. Anyremedial action taken by TENNELEC, Inc. will be basedon the information contained in this report.
SERVICING
In the event of a component failure, replacement may bedone in the field or the instrument may be returned toour plant for repair. There will be no charge forrepairs that fall within the warranty.
WARRANTY
In connection with TENNELEC's warranty (inside frontcover), TENNELEC suggests that if a fault develops, thecustomer should immediately notify the TENNELECCustomer Service Manager. The Customer Service Managermay be able to prescribe repairs and to sendreplacement parts which will enable you to get theinstrument operating sooner and at less expense than ifYOU returned it. Additionally, due to thesusceptibility of input FET's to damage when operatedwithout the protection network installed, these devicesare not covered under the warranty.
- 35 -
Should return prove necessary, the TENNELEC CustomerService Manager must be informed in WRITING, BY CABLEor TWX of the nature of the fault and the model numberand serial number of the instrument. Pack theinstrument well and ship PREPAID and INSURED toTENNELEC, Inc., 601 Oak Ridge Turnpike, Oak Ridge,Tennessee 37830. As stated in the warranty DAMAGE INTRANSIT WILL BE REPAIRED AT THE SENDER's EXPENSE aswill damage that obviously resulted from abuse ormisuse of the instrument.
Quotations for repair of such damage will be sent foryour approval before repair is undertaken.**************************** ** TENNELEC's Quality Assurance Program requires ** that each and every instrument be fully aged, ** vibrated, and electronically checked. ** ** Should the user require a copy of the Quality ** Control Procedure and Test Record, please call ** the Customer Service Department of TENNELEC. ** Both model number and serial numbers are ** required. ** ****************************
MANUAL REV. 2
l/88 - Engineering and component improvements may bemade after date of printing.
,
-
.
