Making Precision Lw VL nd Lw Rsisnc MsuRMns A g r e A t e r m e A s u r e o f c o n f i d e n c e
2
Ak a Qt
Low Voltage Measurements
IntroductionLow voltage and low resistance measurements areoten made on devices and materials with low sourceimpedance. This e-handbook discusses several potentialsources o error in low voltage measurements and howto minimize their impact on measurement accuracy,as well as potential error sources or low resistancemeasurements, including lead resistance, thermoelectricEMFs, non-ohmic contacts, device heating, dry circuittesting, and measuring inductive devices.
Signiicant errors may be introduced into low voltagemeasurements by oset voltages and noise sources thatcan normally be ignored when measuring higher voltagelevels. These actors can have a signicant eect on lowvoltage measurement accuracy.
Oset VoltagesIdeally, when a voltmeter is connected to a relatively lowimpedance circuit in which no voltages are present, itshould read zero. However, a number o error sourcesin the circuit may be seen as a non-zero voltage oset.These sources include thermoelectric EMFs, osetsgenerated by rectiication o RFI (radio requencyintererence), and osets in the voltmeter input circuit.
Figure 1: Eects o Oset Voltages on Voltage Measurement Accuracy
As shown in Figure 1, any oset voltage (VOFFSET) will add
to or subtract rom the source voltage (VS) so that thevoltage measured by the meter becomes:
VM = VS ± VOFFSET
The relative polarities o the two voltages will determinewhether the oset voltage adds to or subtracts rom thesource voltage. Steady osets can generally be nulledout by shorting the ends o the test leads together,then enabling the instrument’s zero (relative) eature.Note, however, that cancellation o oset drit mayrequire requent rezeroing, particularly in the case othermoelectric EMFs.
ThermoelecTric emFThermoelectric voltages (thermoelectric EMFs) arethe most common source o errors in low voltage
measurements. These voltages are generated whendierent parts o a circuit are at dierent temperaturesand when conductors made o dissimilar materials arejoined together, as shown in Figure 2. The Seebeckcoeicients (QAB) o various materials with respect tocopper are summarized in Table 1.
Figure 2: Thermoelectric EMFs
FeATured resource
n TroubleshootingLow VoltageMeasurementProblems
n Accurate Low-ResistanMeasurements Start wIdentiying Sources o
Making Precision Lw VL nd Lw Rsisnc MsuRMns A g r e A t e r m e A s u r e o f c o n f i d e n c e
3
Ak a Qt
Low Voltage Measurements
Table 1: Seebeck Coefcients
Pa mata* sbk cft, QAB
Cu - Cu ≤0.2 μV/°CCu - Ag 0.3 μV/°C
Cu - Au 0.3 μV/°C
Cu - Pb/Sn 1–3 μV/°C
Cu - Si 400 μV/°C
Cu - Kovar ~40–75 μV/°C
Cu - CuO ~1000 μV/°C
* Ag = silver Au = gold Cu = copper CuO = copper oxidePb = lead Si = silicon Sn = tin
Constructing circuits using the same material or all
conductors minimizes thermoelectric EMF generation.For example, crimping copper sleeves or lugs onto copperwires results in copper-to-copper junctions, which generateminimal thermoelectric EMFs. Also, connections mustbe kept clean and ree o oxides. Crimped copper-to-copper connections, called “cold welded,” do not allowoxygen penetration and may have a Seebeck coeciento ≤0.2μV/°C, while Cu-CuO connections may have acoecient as high as 1mV/°C.
Minimizing temperature gradients within the circuit alsoreduces thermoelectric EMFs. A technique or minimizingsuch gradients is to place corresponding pairs o junctions
in close proximity to one another and to provide goodthermal coupling to a common, massive heat sink. Electricalinsulators having high thermal conductivity must be used,but, since most electrical insulators don’t conduct heatwell, special insulators such as hard anodized aluminum,beryllium oxide, specially lled epoxy resins, sapphire, ordiamond must be used to couple junctions to the heat sink.
Allowing test equipment to warm up and reach thermalequilibrium in a constant ambient temperature alsominimizes thermoelectric EMF eects. The instrument zeroeature can compensate or any remaining thermoelectricEMF, provided it is relatively constant. To keep ambienttemperatures constant, equipment should be kept awayrom direct sunlight, exhaust ans, and similar sources oheat fow or moving air. Wrapping connections in insulatingoam (e.g., polyurethane) also minimizes ambienttemperature fuctuations caused by air movement.
connecTions To Avoid ThermoelecTric emF
Connections in a simple low voltage circuit, as shown inFigure 3, will usually include dissimilar materials at dierenttemperatures. This produces a number o thermoelectricEMF sources, all connected in series with the voltage sourceand the meter. The meter reading will be the algebraicsum o all these sources. Thereore, it is important that theconnection between the signal source and the measuring
instrument doesn’t interere with the reading.
Figure 3: Connections rom Voltage Source to Voltmeter
I all the connections can be made o one metal,the amount o thermoelectric EMF added to themeasurement will be negligible. However, this may not
always be possible. Test xtures oten use spring contacts,which may be made o phosphor-bronze, beryllium-copper, or other materials with high Seebeck coecients.In these cases, a small temperature dierence maygenerate a large enough thermoelectric voltage to aectthe accuracy o the measurement.
I dissimilar metals cannot be avoided, an eort shouldbe made to reduce the temperature gradients throughoutthe test circuit by use o a heat sink or by shielding thecircuit rom the source o heat.
Making Precision Lw VL nd Lw Rsisnc MsuRMns A g r e A t e r m e A s u r e o f c o n f i d e n c e
4
Ak a Qt
Low Voltage Measurements
Measurements o sources at cryogenic temperaturespose special problems since the connections betweenthe sample in the cryostat and the voltmeter are oten
made o metals with lower thermal conductivity thancopper, such as iron, which introduces dissimilarmetals into the circuit. In addition, since the sourcemay be near zero Kelvin while the meter is at 300K,there is a very large temperature gradient. Matchingthe composition o the wires between the cryostatand the voltmeter and keeping all dissimilar metaljunction pairs at the same temperature allows makingvery low voltage measurements with good accuracy.
reversinG sources To cAncelThermoelecTric emFs When measuring a small voltage, such as the dierencebetween two standard cells or the dierence betweentwo thermocouples connected back-to-back, theerror caused by stray thermoelectric EMFs can becanceled by taking one measurement, then careullyreversing the two sources and taking a secondmeasurement. The average o the dierence betweenthese two readings is the desired voltage dierence.
In Figure 4, the voltage sources, Va and Vb, representtwo standard cells (or two thermocouples). The voltagemeasured in Figure 4a is:
V1 = Vemf + Va – Vb
The two cells are reversed in Figure 4b and the measured
voltage is: V2 = Vemf + Vb – Va
The average o the dierence between these twomeasurements is:
V1 – V2 = Vemf + Va – Vb – Vemf – Vb + Va or Va – Vb 2 2
Figure 4: Reversing Sources to Cancel Thermoelectric EMFs
Notice that this measurement technique eectivelycancels out the thermoelectric EMF term (Vem), whichrepresents the algebraic sum o all thermoelectric EMFsin the circuit except those in the connections between V a and Vb. I the measured voltage is the result o a currentfowing through an unknown resistance, then either thecurrent-reversal method or the o set-compensated ohmsmethod may be used to cancel the thermoelectric EMFs.
rFi/emiRFI (Radio Frequency Intererence) and EMI (Electro-magnetic Intererence) are general terms used todescribe electromagnetic intererence over a wide rangeo requencies across the spectrum. RFI or EMI can becaused by sources such as TV or radio broadcast signalsor it can be caused by impulse sources, as in the case
o high voltage arcing. In either case, the eects on themeasurement can be considerable i enough o theunwanted signal is present.
RFI/EMI intererence may maniest itsel as a steadyreading oset or it may result in noisy or erratic readings.A reading oset may be caused by input ampliieroverload or DC rectication at the input.
RFI and EMI can be minimized by taking severalprecautions when making sensitive measurements. Themost obvious precaution is to keep all instruments, cables,and DUTs as ar rom the intererence source as possible.Shielding the test leads and the DUT (Figure 5) will otenreduce intererence eects to an acceptable level. Noiseshields should be connected to input LO. In extreme cases,a specially constructed screen room may be necessary to
attenuate the troublesome signal suciently.I all else ails to prevent RF intererence rom beingintroduced into the input, external ltering o the deviceinput paths may be required, as shown in Figure 6. Inmany cases, a simple one-pole lter may be sucient;in more dicult cases, multiple-pole notch or band-stoplters may be required. In particular, multiple capacitorso dierent values may be connected in parallel toprovide low impedance over a wide requency range.Keep in mind, however, that such ltering may have otherdetrimental eects, such as increased response time onthe measurement.
Making Precision Lw VL nd Lw Rsisnc MsuRMns A g r e A t e r m e A s u r e o f c o n f i d e n c e
5
Ak a Qt
Low Voltage Measurements
Figure 5: Shielding to Attenuate RFI/ EMI Intererence
Figure 6: Shielded Connections to Reduce Inducted RFI/EMI
inTernAl oFFseTsNanovoltmeters will rarely indicate zero when no voltageis applied to the input, since there are unavoidable voltage
osets present in the input o the instrument. A shortcircuit can be connected across the input terminals andthe output can then be set to zero, either by ront panelzero controls or by computer control. I the short circuit hasa very low thermoelectric EMF, this can be used to veriyinput noise and zero drit with time. Clean, pure copperwire will usually be suitable. However, the zero establishedin this manner is useul only or verication purposes andis o no value in the end application o the instrument.
I the instrument is being used to measure a smallvoltage drop resulting rom the fow o current througha resistor, the ollowing procedure will result in a properzero. First, the instrument should be allowed to warm upor the specied time, usually one to two hours. Duringthis time, the connections should be made betweenthe device under test and the instrument. No currentshould be supplied to the device under test to allow thetemperature gradients to settle to a minimum, stablelevel. Next, the zero adjustment should be made. In someinstruments, this is done by pressing REL (or Relative) orZERO button. The instrument will now read zero. Whenthe test current is applied, the instrument will indicate theresulting voltage drop. In some applications, the voltageto be measured is always present and the precedingprocedure cannot be used. For example, the voltagedierence between two standard cells is best observed
by reversing the instrument connections to the cellsand averaging the two readings. This same technique
is used to cancel osets when measuring the output odierential thermocouples. This is the same method used
to cancel thermoelectric EMFs.
Zero driFTZero drit is a change in the meter reading with no inputsignal (measured with the input shorted) over a periodo time. The zero drit o an instrument is almost entirelydetermined by the input stage. Most nanovoltmeters usesome orm o chopping or modulation o t he input signalto minimize the drit.
The zero reading may also vary as the ambient temperaturechanges. This eect is usually reerred to as the temperaturecoecient o the voltage oset. In addition, an instrumentmay display a transient temperature eect. Ater a step
change in the ambient temperature, the voltage oset maychange by a relatively large amount, possibly exceeding thepublished specications.
The oset will then gradually decrease and eventuallysettle to a value close to the original value. This is theresult o dissimilar metal junctions in the instrument withdierent thermal time constants. While one junction willadjust to the new ambient temperature quickly, anotherchanges slowly, resulting in a temporary change involtage oset.
To minimize voltage osets due to ambient temperaturechanges in junctions, make measurements in a
temperature controlled environment and/or slow downtemperature changes by thermally shielding the circuit.
Making Precision Lw VL nd Lw Rsisnc MsuRMns A g r e A t e r m e A s u r e o f c o n f i d e n c e
6
Ak a Qt
Low Voltage Measurements
NoiseSigniicant errors can be generated by noise sources,which include Johnson noise, magnetic elds, and ground
loops. An understanding o these noise sources and themethods available to minimize them is crucial to makingmeaningul low voltage measurements.
Johnson noiseThe ultimate limit o resolution in an electrical measure-ment is dened by Johnson or thermal noise. This noiseis the voltage associated with the motion o electrons dueto their thermal energy at temperatures above absolutezero. All voltage sources have internal resistance, so allvoltage sources develop Johnson noise. The noise voltagedeveloped by a metallic resistance can be calculated romthe ollowing equation:
where: V = rms noise voltage developed in source resistance
k = Boltzmann’s constant, 1.38 × 10 –23 joule/K
T = absolute temperature of the source in Kelvin
B = noise bandwidth in hertz
R = resistance of the source in ohms
For example, at room temperature (290K), a sourceresistance o 10kΩ with a measurement bandwidth o5kHz will have almost 1μV rms o noise.
Johnson noise may be reduced by lowering thetemperature o the source resistance and by decreasingthe bandwidth o the measurement. Cooling the sample
rom room temperature (290K) to liquid nitrogentemperature (77K) decreases the voltage noise byapproximately a actor o two.
I the voltmeter has adjustable ltering and integration,the bandwidth can be reduced by increasing the amounto ltering and/or by integrating over multiple power linecycles. Decreasing the bandwidth o the measurementis equivalent to increasing the response time o theinstrument, and as a result, the measurement time ismuch longer. However, i the measurement responsetime is long, the thermoelectric EMFs associated withthe temperature gradients in the circuit become moreimportant. Sensitive measurements may not be achievedi the thermal time constants o the measurement circuitare o the same order as the response time. I this occurs,distinguishing between a change in signal voltage and achange in thermoelectric EMFs becomes impossible.
mAGneTic FieldsMagnetic ields generate error voltages in two circum-stances: 1) i the eld is changing with time, and 2) i thereis relative motion between the circuit and the eld. Voltagesin conductors can be generated rom the motion o aconductor in a magnetic eld, rom local AC currents causedby components in the test system, or rom the deliberateramping o the magnetic ield, such as or magneto-
resistance measurements. Even the earth’s relativelyweak magnetic eld can generate nanovolts in danglingleads, so leads must be kept short and rigidly tied down.
Basic physics shows that the amount o voltage a magneticield induces in a circuit is proportional to the area thecircuit leads enclose and the rate o change in magnetic
fux density, as shown in Figure 7. The induced voltage isproportional both to the magnitude o A andB, as well asto the rate o change in A andB, so there are two ways tominimize the induced voltage:
n Keep both A andB to a minimum by reducing looparea and avoiding magnetic elds, i possible; and
n Keep both A andB constant by minimizing vibrationand movement, and by keeping circuits away romAC and RF elds.
To minimize induced magnetic voltages, leads must berun close together and magnetically shielded and theyshould be tied down to minimize movement. Mu-metal,a special alloy with high permeability at low magnetic fuxdensities and at low requencies, is a commonly usedmagnetic shielding material.
Figure 7: Low Voltages Generated by Magnetic Fields
Making Precision Lw VL nd Lw Rsisnc MsuRMns A g r e A t e r m e A s u r e o f c o n f i d e n c e
7
Ak a Qt
Low Voltage Measurements
Figure 8 shows two ways o locating the leads romthe source to the voltmeter. In Figure 8a, a large area isenclosed; thus, a large voltage is developed. In Figure 8b,
a much smaller area is enclosed because the leads aretwisted together, and the voltage induced is considerablyreduced. Twisted pair also cancels magnetically inducedvoltages because each adjacent twist couples a smallbut alternating polarity (equal) voltage. Conductors thatcarry large currents should also be shielded or run astwisted pairs to avoid generating magnetic elds that canaect nearby circuits. In addition to these techniques, ACsignals rom magnetic elds can be ltered at the inputo the instrument. I possible, the signal source and theinstrument should be physically relocated urther awayrom the interering magnetic eld.
Figure 8: Minimizing Intererence rom Magnetic Fields
Ground looPsNoise and error voltages also arise rom ground loops. When there are two connections to earth, such as when
the source and measuring instruments are both connectedto a common ground bus, a loop is ormed as shownin Figure 9a. A voltage (VG) between the source andinstrument grounds will cause a current (I) to fow aroundthe loop. This current will create an unwanted voltage inseries with the source voltage. From Ohm’s Law:
VG = IRwhere VG = ground loop interering voltage, R = theresistance in the signal path through which the groundloop current fows, and I = the ground loop current. Atypical example o a ground loop can be seen when anumber o instruments are plugged into power strips on
dierent instrument racks. Frequently, there is a smalldierence in potential between the ground points. Thispotential dierence can cause large currents to circulateand create unexpected voltage drops.
Figure 9b: Reduced Ground Loops
The cure or such ground loops is to ground all equipmentat a single point. The easiest way o accomplishing this isto use isolated power sources and instruments, then nd
a single, good earth-ground point or the entire system.Avoid connecting sensitive instruments to the sameground system used by other instruments, machinery,or other high power equipment. As shown in Figure 9b,ground loops can also be reduced by using a voltmeterwith high common mode impedance (ZCM), also knownas common mode isolation.
Making Precision Lw VL nd Lw Rsisnc MsuRMns A g r e A t e r m e A s u r e o f c o n f i d e n c e
8
Ak a Qt
Low Voltage Measurements
Excessive common-mode current can signicantly aectlow-level voltage measurements. Although common-mode currents are most oten associated with noise
problems, they can result in large DC osets in somecases. In the ollowing paragraphs, we will briefy discussthe basic principles behind errors generated by common-mode currents and ways to avoid lead reversal errors.
common-mode currenTCommon-mode current is the current that fows betweenthe instrument’s LO terminal and chassis or earth ground.As shown in Figure 10, common-mode current (ICM) iscaused by capacitive coupling (CCOUPLING) rom the powerline through the power transormer. The amplitude o thecommon-mode current is dened as:
ICM = 2πfCCOUPLING (V2 ± V1)
where is the power line requency.
Note that the common-mode current lows through theimpedance (ZCM), which is present between input LO andchassis ground. As a result, the amplitude o voltage (VCM)depends on the magnitude o ZCM as well as the value o ICM.
Figure 10: Common Mode Current Generation byPower Line Coupling
common-mode reversAl errorsReversing leads can result in errors caused by common-mode currents. As shown in Figure 11, many low voltage
sources have internal resistive dividers, which attenuatean internal voltage source to the desired level. Forexample, the output voltage rom the source is dened as:
With the correct connection scheme shown in Figure 11a,the low or chassis side o the voltage source is connectedto input LO o the measuring instrument. Any common-mode current (ICM) that may be present lows romthe voltmeter input LO to instrument chassis common,through earth ground to voltage source ground. Notethat no common-mode current fows through either o
the two divider resistors o the voltage source when thisconnection scheme is used.
I the input leads o the voltmeter are reversed, we havethe situation shown in Figure 11b. Now, the common-mode current (ICM) fows through R2, developing a voltagedrop, which is added to the voltage to be measured.This added voltage is mainly power line requency andits eect on the voltmeter reading will depend uponthe normal-mode rejection capability o the meter. Thereading may become noisy or it may have a constantoset. In some cases, the sensitivity o the meter maybe reduced, because the input stages are overloaded.To minimize common-mode reversal errors, choosean instrument with the lowest possible common-modecurrent. I possible, the voltage source being measuredshould be isolated rom ground.
Figure 11: Eects o Reversing Leads on Common Mode Errors
Making Precision Lw VL nd Lw Rsisnc MsuRMns A g r e A t e r m e A s u r e o f c o n f i d e n c e
9
Ak a Qt
Low Resistance Measurements
Lead Resistance and Four-Wire MethodResistance measurements are oten made using thetwo-wire method shown in Figure 12. The test current
is orced through the test leads and the resistance (R)being measured. The meter then measures the voltageacross the resistance through the same set o test leadsand computes the resistance value accordingly. The mainproblem with the two-wire method as applied to lowresistance measurements is that the total lead resistance(RLEAD) is added to the measurement. Because the testcurrent (I) causes a small but signicant voltage dropacross the lead resistances, the voltage (VM) measured bythe meter won’t be exactly the same as the voltage (VR)directly across the test resistance (R), and considerableerror can result. Typical lead resistances lie in the rangeo 1mΩ to 10mΩ, so it’s very dicult to obtain accurate
two-wire resistance measurements when the resistanceunder test is lower than 10Ω to 100Ω (depending onlead resistance).
Figure 12: Two-Wire Resistance Measurement
Figure 13: Four-Wire Resistance Measurement
Due to the limitations o the two-wire method, the our-wire (Kelvin) connection method shown in Figure 13
is generally preerred or low resistance measurements.These measurements can be made using a DMM, micro-ohmmeter, or a separate current source and voltmeter. With this coniguration, the test current (I) is orcedthrough the test resistance (R) through one set o testleads, while the voltage (VM) across the DUT is measuredthrough a second set o leads called sense leads.Although some small current may fow through the senseleads, it is usually negligible and can generally be ignoredor all practical purposes. The voltage drop across thesense leads is negligible, so the voltage measured by themeter (VM) is essentially the same as the voltage (VR)across the resistance (R). Consequently, the resistancevalue can be determined much more accurately thanwith the two-wire method. Note that the voltage-sensingleads should be connected as close to the resistor under
test as possible to avoid including the resistance o thetest leads in the measurement.
Thermoelectric EMFs and OsetCompensation MethodsThermoelectric voltages can seriously aect lowresistance measurement accuracy. The current-reversalmethod, the delta method, and the oset-compensatedohms method are three common ways to overcomethese unwanted osets.
currenT-reversAl meThodThermoelectric EMFs can be canceled by making twomeasurements with currents o opposite polarity, asshown in Figure 14. In this diagram, a voltmeter with aseparate bipolar current source is used. With the positivecurrent applied as in Figure 14a, the measured voltage is:
VM+ = VEMF + IRReversing the current polarity as shown in Figure 14b yields the ollowing voltage measurement:
VM– = VEMF – IRThe two measurements can be combined to cancelthermoelectric EMFs:
VM = = = IRVM+ – VM– (VEMF + IR) – (VEMF – IR)
2 2
The measured resistance is computed in the usual manner:
Making Precision Lw VL nd Lw Rsisnc MsuRMns A g r e A t e r m e A s u r e o f c o n f i d e n c e
10
Ak a Qt
Low Resistance Measurements
Note that the thermoelectric voltage (V EMF) is completelycanceled out by this method o resistance calculation.
Figure 14: Canceling Thermoelectric EMFs with Current Reversal
For the current-reversal method to be eective, it’s im-portant to use a low noise voltmeter with a response
speed that is ast compared with the thermal timeconstant o the circuit under test. I the response speedis too slow, any changes in the circuit temperatureduring the measurement cycle will cause changes in thethermoelectric EMFs that won’t be completely canceled,and some error will result.
delTA meThod When the thermoelectric voltages are constant with respectto the measurement cycle, the current-reversal methodwill successully compensate or these osets. However,i changing thermoelectric voltages are causing inaccurateresults, then the delta method should be used. The deltamethod is similar to the current-reversal method in terms
o alternating the current source polarity, but it diers inthat it uses three voltage measurements to make each
resistance calculation. This method can best be e xplainedthrough an illustration and mathematical computations.
Figure 15 shows the voltage drop o a DUT as a unction otime with an alternating polarity current applied. A voltagemeasurement (VM1, VM2, VM3, etc.) is taken each time thepolarity is changed. Each voltage measurement includesa constant thermal voltage oset (VEMF) and a linearlychanging voltage oset (δ V). The thermal voltage drit maybe approximated as a linear unction over short periods,so the rate o change o voltage as a unction o time ( δ V)can also be treated as a constant. The rst three voltagemeasurements include the ollowing voltages:
VM1 = V1 + VEMF
VM2 = V2 + VEMF + δV
VM3 = V3 + VEMF+ 2δV
where: VM1, VM2, and VM3 are voltage measurements
VM1 is presumed to be taken at time = 0
V1, V2, and V3 are the voltage drop o the DUT dueto the applied current
VEMF is the constant thermoelectric voltage oset atthe time the VM1 measurement is taken
δ V is the thermoelectric voltage change
Cancellation o both the thermoelectric voltage oset(VEMF) term and the thermoelectric voltage change (δ V)
term is possible through mathematical computationusing three voltage measurements. First, take one-hal
the dierence o the rst two voltage measurements andcall this term VA:
Then, take one-hal the dierence o the second (VM2) andthird (VM3) voltage measurements and call this term VB:
Figure 15: Canceling Thermoelectric EMFs with Delta Method
Both VA and VB are aected by the drit in thethermoelectric EMF, but the eect on VA and VB is equaland opposite. The nal voltage reading is the average o VA and VB and is calculated as:
VFINAL = =VA – VB (V1 + V3 – 2V2)
2 4Notice that both the VEMF and δ V terms are canceled outo the nal voltage calculation.
Making Precision Lw VL nd Lw Rsisnc MsuRMns A g r e A t e r m e A s u r e o f c o n f i d e n c e
11
Ak a Qt
Low Resistance Measurements
In the delta method, each data point is the movingaverage o three voltage readings. This additionalaveraging o the voltage measurements means that the
data resulting rom the delta method has lower noisethan the data derived when the current-reversal methodis used to calculate it, even when both sets o data aretaken over the same time period.
The success o the delta method depends on the linearapproximation o the thermal drit, which must beviewed over a short period. Compensating successullyor changing thermoelectric voltages dictates that themeasurement cycle time must be aster than the thermaltime constant o the DUT. Thereore, an appropriately astcurrent source and voltmeter must be used or the deltamethod to be successul.
oFFseT-comPensATed ohms meThodAnother oset-canceling method used by micro-ohmmeters and many DMMs is the oset-compensatedohms method. This method is similar to the current-reversal method except that the measurements arealternated between a ixed source current and zerocurrent. As shown in Figure 16a, the source current isapplied to the resistance being measured during onlypart o the cycle. When the source current is on, thetotal voltage measured by the instrument (Figure 16b)includes the voltage drop across the resistor as well asany thermoelectric EMFs, and it is dened as:
VM1 = VEMF + IR
During the second hal o the measurement cycle, thesource current is turned o and the only voltage measuredby the meter (Figure 16c) is any thermoelectric EMF
present in the circuit:
VM2 = VEMF
Given that VEMF is accurately measured during the secondhal o the cycle, it can be subtracted rom the voltagemeasurement made during the rst hal o the cycle, sothe oset-compensated voltage measurement becomes:
VM = VM1 – VM2
VM = (VEMF + IR) – VEMF
VM = IR
and,
R =VM
1
Again, note that the measurement process cancels thethermoelectric EMF term (VEMF).
Making Precision Lw VL nd Lw Rsisnc MsuRMns A g r e A t e r m e A s u r e o f c o n f i d e n c e
12
Ak a Qt
Low Resistance Measurements
Non-Ohmic ContactsNon-ohmic contacts are evident when the potentialdierence across the contact isn’t linearly proportional to
the current fowing through it. Non-ohmic contacts mayoccur in a low voltage circuit as a result o oxide lms orother non-linear connections. A non-ohmic connection islikely to rectiy any radio requency energy (RFI) present,causing an oset voltage to appear in the circuit. Thereare several ways to check or non-ohmic contacts andmethods to reduce them.
I using a micro-ohmmeter or DMM to make low resistancemeasurements, change the range to check or non-ohmiccontacts. Changing the measurement range usuallychanges the test current as well. A normal conditionwould indicate the same reading but with higher or lowerresolution, depending on whether the instrument was upor down ranged. I the reading is signicantly dierent, thismay indicate a non-ohmic condition.
I using a separate current source and voltmeter to makelow resistance measurements, each instrument must bechecked or non-ohmic contacts. I the current sourcecontacts are non-ohmic, there may be a signiicantdierence in the compliance voltage when the sourcepolarity is reversed. I the voltmeter contacts are non-ohmic, they may rectiy any AC pickup present andcause a DC oset error. I this is the case, the osetcompensated ohms method is preerred to the current-reversal method or canceling osets.
To prevent non-ohmic contacts, choose an appropriatecontact material, such as indium or gold. Make sure the
compliance voltage is high enough to avoid problemsdue to source contact non-linearity. To reduce errordue to voltmeter non-ohmic contacts, use shielding andappropriate grounding to reduce AC pickup.
Device Heating Device heating can be a consideration when makingresistance measurements on temperature-sensitivedevices such as thermistors. The test currents used or lowresistance measurements are o ten much higher than thecurrents used or high resistance measurements, so powerdissipation in the device can be a consideration i it is highenough to cause the device’s resistance value to change.
Recall that the power dissipation in a resistor is given bythis ormula:
P = I2RFrom this relationship, we see that the power dissipatedin the device increases by a actor o our each time thecurrent doubles. Thus, one way to minimize the eectso device heating is to use the lowest current possiblewhile still maintaining the desired voltage across thedevice being tested. I the current cannot be reduced, usea narrow current pulse and a ast responding voltmeter.
Most micro-ohmmeters and DMMs don’t have provisionsor setting the test current. It is generally determined by
the range. In those cases, alternate means must be oundto minimize device heating. One simple but eective way
to do so is to use the instrument’s one-shot trigger modeduring measurements. While in this mode, the instrument
will apply only a single, brie current pulse to the DUTduring the measurement cycle, thereby minimizing errorscaused by device heating.
Dry Circuit TestingMany low resistance measurements are made on devicessuch as switches, connectors, and relay contacts. I thesedevices are to be used under “dry circuit” conditions,that is, with an open-circuit voltage less than 20mV and ashort-circuit current less than 100mA, the devices shouldbe tested in a manner that won’t puncture any oxidelm that may have built up on the contacts. I the lm ispunctured, the measured contact resistance will be lowerthan i the lm remains intact, compromising the validityo the test results.
To avoid oxidation puncture, such measurements areusually made using dry circuit testing, which typicallylimits the voltage across the DUT to 20mV or less. Somemicro-ohmmeters and DMMs have this capability builtin, as shown in Figure 17. In this micro-ohmmeter, aprecision shunt resistor (RSH) is connected across thesource terminals to clamp or limit the voltage across theDUT to <20mV. The remaining aspects o the circuit arevery similar to the conventional our-wire measurementmethod: V and RREF make up the current source, whichorces current through the unknown resistance (R). This
current should be no more than 100mA. The value othe unknown resistance is computed rom the sense
Making Precision Lw VL nd Lw Rsisnc MsuRMns A g r e A t e r m e A s u r e o f c o n f i d e n c e
13
Ak a Qt
Low Resistance Measurements
voltage (VM), the voltage across clamping resistor (V SH),the known value o RSH, and the source current.
Figure 17: Dry Circuit Testing
I dry circuit testing is to be done with a separate currentsource and voltmeter, the compliance voltage on the
current source must be limited to 20mV or less. I it isn’tpossible to limit the compliance voltage to this level, acompliance limiting resistor must be used, as shown inFigure 18. In this circuit, RC is the resistor used to limitthe voltage to 20mV and R is the unknown resistance.
The value o RC must be chosen to limit the voltage at agiven test current. For example, i the voltage limit is 20mVand the test current is 200μA, RC can be calculated as:
RC = 20mV/200μA = 100ΩI the unknown resistance (R) is 250mΩ, then RC willcause a 0.25% error in the measured resistance.
The exact value o the unknown resistance (R) can thenbe calculated by the ollowing equation:
where RMEASURED is the calculated resistance measurementrom the measured voltage (VM) and the source current (I).
Testing Inductive DevicesInductive devices usually have a small resistance inaddition to the inductance. This small resistance is normallymeasured with a DMM or a micro-ohmmeter. However,the measurements are oten diicult because o theinteraction between the inductance and the measuringinstrument. This is particularly true with high L/ R ratios.
Some o the problems that may result include oscillations,negative readings, and generally unstable readings. Anoscilloscope picture o an unstable measurement o a200H inductor is shown in Figure 19.
Figure 19: An Unstable Measurement o a 200H Inductor, Acquired with an Oscilloscope
When problems occur, try to take measurements onmore than one range and check i the values correspond.
I possible, do not use oset compensation (pulsedcurrent) because inductive reaction to the current pulsemay cause unstable measurements or make autorangingdicult. Try using a higher resistance range when possible.
Check or oscillations by connecting an oscilloscope inparallel with the device and the meter. Sometimes, adiode across the inductor may settle down the oscillationsby reducing the inductive kick.
Figure 18: Dry Circuit Testing Using Current Source and Voltmeter
R =(RMEASURED × RC)
(RC – RMEASURED)
FeATured resourcen How to Avoid Sel-Hea
on Nanoscale Devices
n Achieving Accurateand ReliableResistance Measure-ments in Low Powerand Low Voltage Applications
n AutomaticResistanceMeasurementson HighTemperatureSuperconductors
AddiTionAl resourn Techniques or Reduci
Measurement UncertaCurrent Reversals vs. Compensation
n Tips or Electrical Chao Carbon NanotubesPower Nanoscale Dev
Making Precision Lw VL nd Lw Rsisnc MsuRMns A g r e A t e r m e A s u r e o f c o n f i d e n c e
14
Ak a Qt
Hall eect measurements have been valuable tools ormaterial characterization since Edwin Hall discoveredthe phenomenon in 1879. Essentially, the Hall eect can
be observed when the combination o a magnetic eldthrough a sample and a current along the length o thesample creates an electrical current perpendicular to boththe magnetic eld and the current, which in turn createsa transverse voltage that is perpendicular to both themagnetic eld and the current (Figure 20).
Figure 20: Illustration o Hall eect
Today, Hall eect measurements are used in many phaseso the electronics industry, rom basic materials researchand device development to device manuacturing. A Halleect measurement system is useul or determininga variety o material parameters, but the primary one
is the Hall voltage (VH). Other important parameterssuch as carrier mobility, carrier concentration (n), Hall
coecient (RH), resistivity, magnetoresistance (RB), andthe conductivity type (N or P) are all derived rom theHall voltage measurement.
Hall eect measurements are useul or characterizingvirtually every material used in producing semiconductors,such as silicon (Si) and germanium (Ge), as well as mostcompound semiconductor materials, including silicon-germanium (SiGe), silicon-carbide (SiC), gallium arsenide(GaAs), aluminum gallium arsenide (AlGaAs), indiumarsenide (InAs), indium gallium arsenide (InGaAs),indium phosphide (InP), cadmium telluride (CdTe), andmercury cadmium telluride (HgCdTe). They’re oten usedin characterizing thin lms o these materials or solarcells/photovoltaics, as well as organic semiconductorsand nano-materials like graphene. They are equallyuseul or characterizing both low resistance materials
(metals, transparent oxides, highly doped semiconductormaterials, high temperature superconductors, dilutemagnetic semiconductors, and GMR/TMR materialsused in disk drives) and high resistance semiconductormaterials, including semi-insulating GaAs, gallium nitride(GaN), and cadmium telluride (CdTe).
Hall eect measurements were irst routinely used inthe semiconductor industry more than two decadesago, when scientists and researchers needed tools orcharacterizing bulk silicon materials. However, once thebulk mobility o silicon was well understood, Hall eectmeasurements were no longer considered critical. Buttoday’s semiconductor materials are not just silicon—manuacturers oten add germanium to silicon in the
strain lattice to get higher mobility. Moreover, modernsemiconductor materials are no longer bulk materials—they’re oten in the orm o thin lms, such as those used
in copper indium gallium diselenide (CIGS) and CdTesolar cells. As a result, IC manuacturers now have togo back to determining carrier concentration and carriermobility independently, applications or which Hall eectmeasurements are ideal.
meAsurinG moBiliTy usinGhAll eFFecT TechniQuesThe irst step in determining carrier mobility is tomeasure the Hall voltage (VH) by orcing both a magneticeld perpendicular to the sample and a current throughthe sample. The combination o the current fow (I) andthe magnetic eld (B) causes a transverse current. Theresulting potential (VH) is measured across the device.
Accurate measurements o both the sample thickness(t) and its resistivity (ρ) are also required. The resistivitycan be determined using either a our-point probe or vander Pauw measurement technique. With just these veparameters (B, I, VH, t, and ρ), the Hall mobility can becalculated using this ormula:
Because Hall voltages are typically quite small (millivoltsor less), as is the measured van der Pauw resistivity, theright measurement and averaging techniques are critical toobtaining accurate mobility results when using this ormula.
Making Precision Lw VL nd Lw Rsisnc MsuRMns A g r e A t e r m e A s u r e o f c o n f i d e n c e
15
Ak a Qt
Lo Voltage pplato: Hall Effect Measurements
Figure 21. Hall eect voltage vs. van der Pauw resistancemeasurement confgurations
Figure 21 illustrates the measurement conigurationsor both the Hall eect voltage and the van der Pauwresistivity measurement. Although these measurementcongurations are very similar in as much as both use ourcontacts and both measurements involve orcing a currentand measuring a voltage, in the Hall e ect measurement,the current is orced on opposite nodes o the sampleand then the voltage is measured on the other opposite
nodes so the orce and the measure contact points areinterlaced and the voltage or semiconductors is typicallyaround KT/q, which is about 25 millivolts. It can also bemuch lower. In contrast, or van der Pauw resistivitymeasurements, the current is orced on adjacent nodesand then the voltage is measured on opposing adjacentnodes so everything that is being orced and measuredis on nearest pins; in that case, the voltages can be wellabove 20 millivolts. The voltages can be anywhere rommillivolts or low resistivity materials to 100 volts orvery high resistivity insulating materials. The other majordierentiator is that there is no magnetic eld being appliedin the van der Pauw measurement, whereas or the Halleect measurement, a transverse magnetic eld is applied.
To obtain results with high condence, the recommendedtechnique involves a combination o reversing source current
polarity, sourcing on additional terminals, and reversing thedirection o the magnetic eld. Eight Hall eect (Figure22)and eight van der Pauw (Figure 23) measurements are
perormed. I the voltage readings within each measurementdier substantially, it’s advisable to recheck the test setup tolook or potential sources o error.
Figure 22: Compute the Hall voltage with both positive and negativepolarity current and with the magnetic feld both up and down, and
with the t wo confgurations shown. Then average all voltages.
Figure 23: Computing average resistivity (ñ) with multiple van derPauw measurements. Four additional resistance measurements
are made with the source current polarity reversed in each o theconfgurations shown. I R A = RB, then R simplifes to pRA/ln(2).
A basic Hall eect measurement conguration will likelyinclude the ollowing components and optional extras:
n A tat-t a agt tat’
pt t ap’ ta. For lowresistivity material samples, the source must be ableto output rom milliamps to amps o current. Forsamples such as semi-insulating GaAs, which may havea resistivity in the neighborhood o 107 ohm·cm, asourcing range as low as 1nA will be needed. For highresistivity samples (such as intrinsic semiconductors),the constant current source may have to be able togo as low as 1nA, but a source capable o producingcurrent rom 10 microamps to 100 milliamps will suce.
n A g pt pa tt. Depending onthe level o material resistivity under test, the voltmeterused must be able to make accurate measurements
anywhere rom 1 microvolt to 100V. High resistivitymaterials may require ultra-high input Z or dierentialmeasurements.
n A pat agt a tagt. Theseare typically available with ranges rom 500 to 5000gauss. An electromagnet will also require a powersupply to drive it.
n A ap .
n opta qpt. A switch matrix is generallyincluded to eliminate the need or manual connections/disconnections between probe contacts; it may alsomake it possible to test multiple samples at once. A
switch matrix is denitely required i the sample is beingheld in a liquid nitrogen dewar or temperature studies.Reerence http://www.nist.gov/eeel/semiconductor/hall_algorithm.cm.
Making Precision Lw VL nd Lw Rsisnc MsuRMns A g r e A t e r m e A s u r e o f c o n f i d e n c e
16
Ak a Qt
Lo Voltage pplato: Hall Effect Measurements
Hall mobility is highly dependent on sample temperature,so it’s oten desirable to monitor this temperature,particularly i the application involves repeating
measurements each time the s ample’s temperature isadjusted. Many test congurations include a temperature-measuring probe; or high accuracy work, the probe’sresolution should be about 0.1° Celsius. A proberchuck that can either heat or cool the sample and atemperature controller are generally necessary or on-waermeasurements when doing temperature studies. A cryostatis necessary to hold the sample in the liquid nitrogen bathor low temperature studies.
For making on-waer measurements o numerous devices,a prober with a manipulator and probe tips will likely beessential.
The most appropriate Hall eect measurement
conguration or a particular application is based in largepart on the sample’s total resistance as measured by theelectrical test equipment. This total resistance is the sumo the sample resistance and the contact resistance, thatis, the resistance between the sample and the electricalcontacts to it. The sample resistance depends on thesample’s intrinsic resistivity, which is expressed in units oohm-centimeters (ohm·cm), and its thickness.
Each resistance range has dierent measurementrequirements and the type and number o componentso the systems needed to test them can vary signicantly.To illustrate the coniguration process, here’s anexample o a conguration appropriate or the widestrange o sample resistances, rom 1 micro-ohm to 1terra-ohm. A conguration o this type would be most
appropriate or those characterizing thin-lm photovoltaicmaterials, those who are studying the eects o dopingconcentration on an intrinsic semiconductor, or those
studying the eects o doping polymers with increasingamounts o carbon nanotubes.
The system conguration illustrated in Figure 24 is ap-propriate or the widest range o sample resistances,rom 1 micro-ohm to 1 terra-ohm. It employs Keithley’sspecial matrix switching card optimized or Hall eectmeasurements, the Model 7065, housed in a Model 7001Switch Mainrame. This card buers test signals romthe sample to the measurement instrumentation andswitches current rom the current source to the sample.The Model 7065 card oers the advantage o unity gainbuers that can be switched in and out to allow themeasurement o high resistances by buering the sample
resistance rom the meter.
Figure 24: Example test confguration or characterizing materials with wide range o sample resistances (1 micro-ohm to 1 tera-ohm)
The test setup also includes the Model 6485 Picoammeter,the Model 6220 DC Current Source, and the Model2182A Nanovoltmeter. The Model 6485 Picoammeter is
included to measure leakage currents so they can eitherbe subtracted out or monitored to make sure they aren’timpacting the high resistance measurement. The Model6220 and the Model 2182A are designed to work togetherseamlessly, using a delta mode technique to synchronizetheir operation and optimize their perormance. Essentially,the delta mode automatically triggers the current source toalternate the signal polarity, then triggers a nanovoltmeterreading at each polarity, cancelling out both constant anddriting thermoelectric osets, and ensuring the resultsrefect the true value o the voltage. Once the Model 6220and the Model 2182A are connected properly, all it takes tostart a test is pressing the current source’s Delta button andthen the Trigger button. The Model 2182A also provides asecond channel o voltage measurement capability, whichis useul or monitoring the temperature o the sample.Although the Model 6220 serves as the constant currentsource in the conguration shown, substituting the Model6221 AC+DC Current Source, which has a built-in arbitrarywaveorm generator, has the advantage o allowing usersto make AC Hall eect measurements. For applicationsor which it is acceptable to trade o the low resistancecapability o the system shown to reduce the system cost(i.e., to provide just mid-range to high resistance capability),a Model 2000 Digital Multimeter can be substituted or theModel 2182A Nanovoltmeter.
FeATured resourcen Hall Eect
Measurementsin MaterialsCharacterization
n Hall Eect MeasuremFundamentals webina
AddiTionAl resourn Precision, Low Curren
Device Testing and Ch
n Low-Level Pulsed ElecCharacterization with6221/2182A Combina
Making Precision Lw VL nd Lw Rsisnc MsuRMns A g r e A t e r m e A s u r e o f c o n f i d e n c e
17
Ak a Qt
At extremely low temperatures, some metals and alloyslose their resistance to electrical current and becomesuperconductive. A superconductor’s transition temperature
and critical current density are two commonly measuredparameters. The superconducting transition temperatureis the point at which a material’s resistance changesrom a nite value to zero. The critical current density isthe maximum current density a material can carry underspecic temperature and magnetic eld conditions beoreit becomes resistive. The higher these two parameters are,the better the superconductor is. Determining these twoparameters requires measuring very small resistances, soa nanovoltmeter and a programmable current source areessential or precision measurements.
Figure 25 shows a basic superconductor resistancemeasurement test system using the combination o a
Model 2182A Nanovoltmeter and a Model 6220 CurrentSource or measuring the resistance. The voltage leadsshould be made o a material with a low Seebeckcoecient with respect to the sample. The sensitivity othe Model 2182A Nanovoltmeter is crucial to obtainingprecision measurements because the applicationdemands the ability to measure extremely low voltages.
For transition temperature measurements, the currentsource must be kept below the critical current o thesample. I the current becomes too high, the powerdissipated may damage the sample and the cryostat.For critical current measurements, however, the currentsource must be able to exceed the critical current o the
sample. I that means that more than 100mA is needed(the current the Model 6220 Current Source can provide),
a Model 2440 5A Current Source may be an appropriatesolution. The current source should have programmablepolarity, so the test can be perormed using the current-
reversal method.The resistance is measured using the same techniquesemployed in low voltage and low res is tancemeasurements. It is essential to use a our-wiremeasurement technique because it eliminates leadresistance by orcing a current through the samplewith one pair o leads while measuring the voltagedrop with a second pair o leads. In addition, theDelta method is essential to eliminate the eects ochanging thermoelectric EMFs, which may interere withmeasurement accuracy.
The Delta method consists o measuring the voltagedrop across the material with the current in one direction,
then reversing the polarity o the current source andtaking a second voltage measurement. Three voltagemeasurements are used to calculate each resistance value.In cases where hysteresis, non-linearity, or asymmetryis apparent, the current can be varied rom one value toanother o the same polarity. This will provide the averageresistance between these two currents.
The Model 2182A Nanovoltmeter and Model 6220 CurrentSource work together to implement the Delta methodautomatically. In this mode, the Model 6220 automaticallyalternates the polarity, then triggers the nanovoltmeterto take a reading at each polarity. Then, the Model 6220displays the “compensated” resistance value. As shown inFigure 26, the resistance can be plotted vs. temperature asthe sample temperature is changing.
Figure 25: Superconductor Resistance Test System
Figure 26: Resistance vs. Temperature o Superconductor
For determining the critical current, the Model 2182A
and Model 6220 Current Source can be used together toproduce a precision I-V curve over a range o currents.
Making Precision Lw VL nd Lw Rsisnc MsuRMns A g r e A t e r m e A s u r e o f c o n f i d e n c e
19
Ak a Qt
Abt Aa. A measure o the closeness o agreement o an instrument readingcompared to that o a primary standard having absolute traceability to a standardsanctioned by a recognized standards organization. Accuracy is oten separated into gainand oset terms. See also Relative Accuracy.
A/d (Aag-t-dgta) ct. A circuit used to convert an analog input signalinto digital inormation. All digital meters use an A/D converter to convert the input signalinto digital inormation.
Aag otpt. An output that is directly proportional to the input signal.
Ab. A molecular manuacturing device that can be used to guide chemicalreactions by positioning molecules. An assembler can be programmed to build virtuallyany molecular structure or device rom simpler chemical building blocks.
At-ragg. The ability o an instrument to automatically switch among ranges todetermine the range oering the highest resolution. The ranges are usually in decade steps.
At-ragg T. For instruments with auto-ranging capability, the time intervalbetween application o a step input signal and its display, including the time ordetermining and changing to the correct range.
Bawt. The range o requencies that can be conducted or amplied within certainlimits. Bandwidth is usually specied by the –3dB (hal-power) points.
Ba vtag. A voltage applied to a circuit or device to establish a reerence level oroperating point o the device during testing.
capata. In a capacitor or system o conductors and dielectrics, that property whichpermits the storage o electrically separated charges when potential dierences existbetween the conductors. Capacitance is related to the charge and voltage as ollows: C= Q/V, where C is the capacitance in arads, Q is the charge in coulombs, and V is thevoltage in volts.
cab natb. A tube-shaped nanodevice ormed rom a sheet o single-layer carbonatoms that has novel electrical and tensile properties. These bers may exhibit electricalconductivity as high as copper, thermal conductivity as high as diamond, strength 100times greater than steel at one-sixth o steel’s weight, and high s train to ailure. They canbe superconducting, insulating, semiconducting, or conducting (metallic). Non-carbonnanotubes, oten called nanowires, are oten created rom boron nitride or silicon.
ca (wtg). One o several signal paths on a switching card. For scanner ormultiplexer cards, the channel is used as a switched input in measuring circuits, or as a
switched output in sourcing circuits. For switch cards, each channel’s signals paths areindependent o other channels. For matrix cards, a channel is established by the ac tuationo a relay at a row and column crosspoint.
caxa cab. A cable ormed rom two or more coaxial cylindrical conductors insulatedrom each other. The outermost conductor is oten earth grounded.
c-m rjt rat (cmrr). The ability o an instrument to rejectintererence rom a common voltage at its input terminals with respect to ground. Usually
expressed in decibels at a given requency.
c-m ct. The current that fows between the input low terminal andchassis ground o an instrument.
c-m vtag. A voltage between input low and earth ground o aninstrument.
ctat rta. The resistance in ohms between the contacts o a relay orconnector when the contacts are closed or in contact.
ctaat. Generally used to describe the unwanted material that adversely aec tsthe physical, chemical, or electrical properties o a semiconductor or insulator.
d/A (dgta-t-Aag) ct. A circuit used to convert digital inormation intoan analog signal. D/A converters are used in many instruments to provide an isolatedanalog output.
dt Abpt. The eect o residual charge storage ater a previously chargedcapacitor has been discharged momentarily.
dgta mtt (dmm). An electronic instrument that measures voltage, current,resistance, or other electrical parameters by converting the analog signal to digitalinormation and display. The typical ve-unction DMM measures DC volts, DC amps, ACvolts, AC amps, and resistance.
dt. A gradual change o a reading with no change in input signal or operatingconditions.
d ct Ttg. The process o measuring a device while keeping the voltage acrossthe device below a certain level (e.g., <20mV) in order to prevent disturbance o oxidationor other degradation o the device being measured.
eta et. A phenomenon whereby currents are generated by galvanicbattery action caused by contamination and humidity.
ett. A highly rened DC multimeter. In comparison with a digital multimeter,an electrometer is characterized by higher input resistance and greater current sensitivity.It can also have unctions not generally available on DMMs (e.g., measuring electric
charge, sourcing voltage).
emF. Electromotive orce or voltage. EMF is generally used in condierence caused by electromagnetic, electrochemical, or thermal e
ettat cpg. A phenomenon whereby a current is generatemoving voltage source near a conductor.
e. The deviation (dierence or ratio) o a measurement rom its values are by their nature indeterminate. See also Random Error and S
Fa T. The time required or a signal to change rom a large pe90%) to a small percentage (usually 10%) o it s peak-to-peak value. See
Faaa cp. A Faraday cup (sometimes called a Faraday cage or icepaimade o sheet metal or mesh. It consists o two electrodes, one inside the oan insulator. While the inner electrode is connected to the electrometer, this connected to ground. When a charged object is placed inside the innercharge will fow into the measurement instrument. The electric eld insidconductor is zero, so the cup shields the object placed inside it rom any atmelectric elds. This allows measuring the charge on the object accurately.
Fbak Pat. A sensitive ammeter that uses an opereedback conguration to convert an input current into voltage or meas
Fatg. The condition where a common-mode voltage exists betweenand the instrument or circuit o interest. (Circuit low is not tied to earth
F-Pt Pb. The our-point collinear probe resistivity measureinvolves bringing our equally spaced probes in contact with the matresistance. The array is placed in the center o the material. A known through the two outside probes and the voltage is sensed at the two inresistivity is calculated as ollows:
π Vρ = —— × — × t × k
ln2 I
where: V = the measured voltage in volts, I = the source current in amthickness in centimeters, k = a correction actor based on the ratio o thdiameter and on the ratio o waer thickness to probe separation.
F-Ta rta mat. A measurement where twto supply a current to the unknown, and two dierent leads are used to drop across the resistance. The our-terminal conguration provides mwhen measuring low resistances.
loar
inRducin | Lw VLg M suRMns | Lw Rsisnc MsuRMns | Lw V Licin: HLL ffc MsuRMns | Lw R Licin: suRcnducR Rsisnc MsuRMns | sLcR guid | gLs
Making Precision Lw VL nd Lw Rsisnc MsuRMns A g r e A t e r m e A s u r e o f c o n f i d e n c e
20
Ak a Qt
F. Reers to C60, an approximately spherical, hollow, carbon molecule containing60 carbon atoms arranged in interlocking hexagons and pentagons, reminiscent othe geodesic dome created by architect R. Buckminster Fuller. Sometimes called“buckminsterullerene” or “buckyball.”
G lp. A situation resulting when two or more instruments are connected todierent points on the ground bus and to earth or power line ground. Ground loops candevelop undesired oset voltages or noise.
Gag. A technique that reduces leakage errors and decreases response time. Guardingconsists o a conductor driven by a low impedance source surrounding the lead o a highimpedance signal. The guard voltage is kept at or near the potential o the signal voltage.
ha et. The measurement o the transverse voltage across a conductor whenplaced in a magnetic eld. With this measurement, it is possible to determine the type,concentration, and mobility o carriers in silicon.
hg ipa Ta. A terminal where the source resistance times the expectedstray current (or example, 1μA) exceeds the required voltage measurement sensitivity.
ipt Ba ct. The current that fows at the instrument input due to internalinstrument circuitry and bias voltage.
ipt ipa. The shunt resistance and capacitance (or inductance) as measured
at the input terminals, not including eects o input bias or oset currents.ipt ot ct. The dierence between the two currents that must be supplied tothe input measuring terminals o a dierential instrument to reduce the output indicationto zero (with zero input voltage and oset voltage). Sometimes inormally used to reerto input bias current.
ipt ot vtag. The voltage that must be applied directly between the inputmeasuring terminals, with bias current supplied by a resistance path, to reduce the outputindication to zero.
ipt rta. The resistive component o input impedance.
iat rta. The ohmic resistance o insulation. Insulation resist ancedegrades quickly as humidity increases.
J n. The noise in a resistor caused by the thermal motion o c harge carriers.It has a white noise spectrum and is determined by the temperature, bandwidth, andresistance value.
lakag ct. Error current that fows (leaks) through insulation resistance when avoltage is applied. Even high resistance paths between low current conductors and nearbyvoltage sources can generate signicant leakage currents.
lg-T Aa. The limit that errors will not exceed during a 90-day or longertime period. It is expressed as a percentage o reading (or sourced value) plus a numbero counts over a specied temperature range.
max Awab ipt. The maximum DC plus peak AC value (voltage or current)
that can be applied between the high and low input measuring terminals withoutdamaging the instrument.
mems. mtaa t. Describes systems that can re spond toa stimulus or create physical orces (sensors and actuators) and that have dimensionson the micrometer scale. They are typically manuactured using the same lithographictechniques used to make silicon-based ICs.
m-t. An ohmmeter that is optimized or low resistance meas urements.The typical micro-ohmmeter uses the our-terminal measurement method and has specialeatures or optimum low level measurement accuracy.
ma et. Any system with atomically precise electronic devices o nanometerdimensions, especially i made o discrete molecular parts, rather than the continuousmaterials ound in today’s semiconductor devices.
ma mapat. A device combining a proximal-probe mechanism oratomically precise positioning with a molecule binding site on the tip; can serve as thebasis or building complex structures by positional synthesis.
ma maatg. Manuacturing using molecular machinery, giving molecule-by-molecule control o products and by-products via positional chemical synthesis.
ma natg. Thorough, inexpensive control o the structure o matterbased on molecule-by-molecule control o products and by-products; the products andprocesses o molecular manuacturing, including molecular machinery.
mosFeT. A metal oxide eld eect transistor. A unipolar device characterized byextremely high input resistance.
na-. A prex meaning one billionth (1/1,000,000,000).
nat. Electronics on a nanometer scale. Includes both molecularelectronics and nanoscale devices that resemble current semiconductor devices.
natg. Fabrication o devices with atomic or molecular scale precision.Devices with minimum eature sizes less than 100 nanometers (nm) are consideredproducts o nanotechnology. A nanometer [one-billionth o a meter (10–9m)] is the unit
o length generally most appropriate or describing the size o single molecules.
natt. A voltmeter optimized to provide nanovolt sensitivity (gthermoelectric EMF connectors, oset compensation, etc.).
n. Any unwanted signal imposed on a desired signal.
na-m rjt rat (nmrr). The ability o an instrintererence across its input terminals. Usually expressed in decibels at a ssuch as that o the AC power line.
na-m vtag. A voltage applied between the high and low ian instrument.
ot ct. A current generated by a circuit even though no sigOset currents are generated by triboelectric, piezoelectric, or electropresent in the circuit.
oa Ptt. A circuit that protects the instrument rom excvoltage at the input terminals.
Pat. An ammeter optimized or the precise measurement oGenerally, a eedback ammeter.
Pzt et . A term used to describe currents generated wstress is applied to certain types o insulators.
P. Reers to the reedom o uncertainty in the measurement. It i
the context o repeatability or reproducibility and should not be used in See also Uncertainty.
Qat dt. A nanoscale object (usually a s emiconductor island) tsingle electron (or a ew) and in which the electrons occupy discrete enethey would in an atom. Quantum dots have been called “articial atoms
ra e. The mean o a large number o measurements infueerror matches the true value. See also Systematic Error.
rag. A continuous band o signal values that can be measured or s oinstruments, range includes positive and negative values.
rag. The displayed number that represents the characteristic o the
rag rat. The rate at which the reading number is updated. The rreciprocal o the time between readings.
rat Aa. The accuracy o a measuring instrument in reerenstandard. See also Absolute Accuracy.
loar
inRducin | Lw VLg M suRMns | Lw Rsisnc MsuRMns | Lw V Licin: HLL ffc MsuRMns | Lw R Licin: suRcnducR Rsisnc MsuRMns | sLcR guid | gLs
Making Precision Lw VL nd Lw Rsisnc MsuRMns A g r e A t e r m e A s u r e o f c o n f i d e n c e
21
Ak a Qt
rpatabt. The closeness o agreement between successive measurements carriedout under the same conditions.
rpbt. The closeness o agreement between measurements o the samequantity carried out with a stated change in conditions.
rt. The smallest portion o t he input (or output) signal that can be measured(or sourced) and displayed.
rp T. For a measuring instrument, the time between application o a step inputsignal and the indication o its magnitude within a rated accuracy. For a sourcing instrument,the time between a programmed change and the availability o the value at its outputterminals. Also known as Settling Time.
r T. The time required or a signal to change rom a small percentage (usually 10%)to a large percentage (usually 90%) o its peak-to-peak amplitude. See also Fall Time.
stt. The smallest quantity that can be measured and displayed.
sttg T. For a measuring instrument, the time between application o a stepinput signal and the indication o its magnitude within a rated accuracy. For a sourcinginstrument, the time between a programmed change and the availability o the value atits output terminals. Also known as Response Time.
sg. A metal enclosure around the circuit being measured, or a metal sleeve
surrounding the wire conductors (coax or triax cable) to lessen intererence, interaction,or leakage. The shield is usually grounded or connected to input LO.
st At. A type o ammeter that measures current by converting the inputcurrent into a voltage by means o shunt resistance. Shunt ammeters have higher voltageburden and lower sensitivity than do eedback ammeters.
st capata lag. The eect on a measurement o the capacitance acrossthe input terminals, such as rom cables or xtures. Shunt capacitance increases both risetime and settling time.
st-T Aa. The limit that errors will not exceed during a short, speciedtime period (such as 24 hours) o continuous operation. Unless specied, no zeroing oradjustment o any kind are permitted. It is expressed as percentage o reading (or sourcedvalue) plus a number o counts over a specied temperature range.
sg et Tat. A switching device that uses controlled electron tunnelingto ampliy current. An SET is made rom two tunnel junctions that share a commonelectrode. A tunnel junction consists o two pieces o metal separated by a very thin(~1nm) insulator. The only way or electrons in one o the metal electrodes to travel to
the other electrode is to tunnel through the insulator. Tunneling is a discrete process,so the electric charge that fows through the tunnel junction fows in multiples o e, thecharge o a single electron.
s ipa. The combination o resistance and capacitive or inductive reactancethe source presents to the input terminals o a measuring instrument.
s-ma ut (smu). An electronic instrument that sources and measures DCvoltage and current. Generally, SMUs have two modes o operation: source voltage andmeasure current, or source current and measure voltage. Also known as source-monitorunit or stimulus-measurement unit.
smt. A SourceMeter instrument is very similar to the source- measure unit inmany ways, including its ability to source and measure both current and voltage and toperorm sweeps. In addition, a SourceMeter instrument can display the measurementsdirectly in resistance, as well as voltage and current. It is designed or general-purpose,high speed production test applications. It can also be used as a source or moderate tolow level measurements and or research applications.
s rta. The resistive component o source impedance. See also TheveninEquivalent Circuit.
spt. Electronics that take advantage o the spin o an electron in some way,rather than just its charge.
staa c. An electrochemical cell used as a voltage reerence in laboratories.
spt. A conductor that has zero resistance. Such materials usually becomesuperconducting only at very low temperatures.
swt ca. A type o card with independent and isolated relays or switching inputsand outputs on each channel.
swtg maa. A switching instrument that connects signals among sourcingand measuring instruments and devices under test. A mainrame is also reerred to as ascanner, multiplexer, matrix, or programmable switch.
stat e. The mean o a large number o measurements infuenced bysystematic error deviates rom the true value. See also Random Error.
Tpat cft. A measure o the change in reading (owith a change in temperature. It is expressed as a percentage o readvalue), plus a number o counts per degree change in temperature.
Tpat cft rta. The change o resistance
device per degree o temperature change, usually expressed in ppm/ °CTt emF. Voltages resulting rom temperature diemeasuring circuit or when conductors o dissimilar materials are joined t
T eqat ct. A circuit used to simpliy analysis oterminal linear networks. The Thevenin equivalent voltage is the open-cthe Thevenin equivalent resistance equals the open-circuit voltage dividcircuit current.
Ta Aa. A comparison o two nearly equal measurementemperature range and time period. It is expressed in ppm. See also ReShort-Term Accuracy.
Tbt et. A phenomenon whereby currents are generated byby riction between a conductor and an insulator.
Tgg. An external stimulus that initiates one or more instrument ustimuli include: an input signal, the ront panel, an external trigger pulsbus X, talk, and GET triggers.
Tw-Ta rta mat. A measurement where the sosense voltage are applied through the same set o test leads.
utat. An estimate o the possible error in a measurement; in estimated possible deviation rom its actual value.
a Paw mat. A measurement technique used resistivity o arbitrarily shaped samples.
vtag B. The voltage drop across the input terminals o an am
vtag ct. The change in resistance value with applied expressed in percent/V or in ppm/V.
Wa-p T. The time required ater power is applied to an instrurated accuracy at reerence conditions.
Z ot. The reading that occurs when the input terminals o a measare shorted (voltmeter) or open-circuited (ammeter).
Specifications are subject to change without notice.
All Keithley trademarks and trade names are the property of Keithley Instruments, Inc.
All other trademarks and trade names are the property of their respective companies.
KE ITHLE Y INSTRUME NTS, INC. n 28775 AURORA RD . n CLEVELAND, OH 44139-1891 n 440-248-0400 n Fax: 440-248-6168 n 1 -888-KE ITHLEY n www.ke i t h l e y .com
