GATE Instrumentation (IN): Measurements (errors and uncertainty, regression, Wheatstone, Kelvin and AC bridges, wattmeters and instrument transformers, digital voltmeters, oscilloscope probes, Lissajous figures and counters), previous-year questions solved
Every measurements question from the last five GATE Instrumentation Engineering papers, solved: measurement errors, uncertainty and regression, DC and AC bridges, galvanometers, wattmeters and instrument transformers, digital voltmeter and multimeter accuracy, oscilloscope probes, Lissajous figures and counters. Each answer matches the official key and was recomputed by code from a model of the instrument or circuit with the values printed in the paper.
Questions from official GATE IN papers (IIT Kharagpur 2022, IIT Kanpur 2023, IISc Bengaluru 2024, IIT Roorkee 2025, IIT Guwahati 2026); solutions by Skill-Lync. · Updated 19 Nov 2026 · 17 of 30 questions on this page
Weighing machine readings 80, 79, 81, 79 and 81 kg: sample standard deviation in kg?
Answer: Official key: 0.98 to 1.02
- Mean: .
- Deviations: 0, , , , kg. Sum of squares: .
- The sample standard deviation divides by , because the mean was estimated from the same five readings: .
- 1.00 kg.
Trap: dividing by gives the population value, kg, which is outside the official range 0.98 to 1.02. The word “sample” asks for . We computed both from the five readings and checked them against a standard statistics routine.
Repeatability studies on scales, gauges and test rigs report the sample standard deviation of repeated readings, and calibration certificates use it in the type A part of the measurement uncertainty. With few readings the n minus 1 divisor matters, which is why it is the default in spreadsheet functions such as STDEV.S.
20 kOhm resistor read six times: 19, 18, 23, 21, 17 kOhm and one more; mean equals true value: last reading?
Answer: Official key: 22 to 22
- The mean of six readings equals the true value: .
- The six readings must add up to . The first five add up to .
- .
- The six readings now average exactly the true value, but they still scatter: the sample standard deviation is . The set is accurate on average and not precise.
Trap: answering 20, as if the last reading itself had to equal the true value, or averaging only the five readings (). The condition is on the mean of all six. We bisected for the sixth reading that makes the mean of six equal to 20: 22.0000 kOhm.
Calibration labs repeat a reading several times and report the mean with its standard deviation, because averaging reduces random error but does nothing for a systematic one. Bench LCR meters and DMMs often have a built-in averaging mode for the same reason.
Flow instrument, true value 0.25, readings 0.35 to 0.39 in a histogram: bias and maximum error?
- (A)0.12 and 0.14
- (B)0.01 and 0.10
- (C)0.10 and 0.14
- (D)0.04 and 0.12
Answer: (A) Official key: A
- Counts from the histogram: 0.35 (1), 0.36 (4), 0.37 (7), 0.38 (5), 0.39 (2), a total of 19 readings.
- Sum of the readings: .
- Mean: flow units.
- Bias (systematic error), the mean minus the true value: . The most frequent reading, 0.37, gives the same 0.12.
- Maximum error, from the reading farthest from the true value: .
- Bias 0.12 and maximum error 0.14: option (A).
Trap: taking the bias from the smallest reading, , gives (C); the spread of the readings, , is a measure of precision, not bias, and leads to (D). We computed the weighted mean and the largest deviation from the counts read off the figure and tested every option pair.
A tight cluster of readings far from the reference is the classic picture of a precise but biased instrument. A bias like this is removed by calibration against a reference standard (a zero or span correction), while the scatter can only be reduced by averaging or a better instrument.
20-psig gauge readings 10.3, 10.5, 10.4, 10.5, 10.5 at x = 1 to 5, line y = 0.04x + 10.32: R squared?
| 1 | 2 | 3 | 4 | 5 | |
| (psig) | 10.3 | 10.5 | 10.4 | 10.5 | 10.5 |
Answer: Official key: 0.450 to 0.550
- Mean of : psig.
- Total sum of squares: .
- Fitted values from : 10.36, 10.40, 10.44, 10.48, 10.52. Residuals: , , , , .
- Residual sum of squares: .
- 0.500.
Trap: reporting the correlation coefficient instead of its square. We fitted the table by least squares in code (it reproduces the paper's slope 0.04 and intercept 10.32 exactly) and computed from the residuals: 0.5000.
Calibration software fits a straight line to reference-versus-reading data and reports the slope, intercept and R squared alongside the residuals. A low R squared on readings that barely change, as here, says the scatter is as large as the trend, so the residuals and the repeatability of the gauge are the more useful numbers.
Series RLC at 7500 rad/s, L = 20 +- 0.8 mH, R = 10 +- 0.3 Ohm: percentage uncertainty in quality factor?
Answer: Official key: 4.5 to 5.5
- . Relative uncertainties: is and is .
- With taken as the stated 7500 rad/s, both enter with power 1. Root-sum-square: 5.0%.
- A second reading gives the same number. If the capacitor is what stays fixed, and enters with power . The worst-case sum is then .
- The key range (4.5 to 5.5) holds 5.0%. It excludes the worst-case sum for , which is , and the RSS for fixed , which is .
Trap: the paper does not say whether the uncertainties are to be combined in quadrature or added, or which of and is held fixed. Two readings land on 5.0% and two do not, so check which rule matches the model you pick. We evaluated at the four corners of and . With held the spread is +7.22% to -6.80%; with held it is +5.13% to -4.87%, half-width 5.0%. RSS on the first model gave 5.00%.
Q-meters and LCR meters report the quality factor of coils, and calibration certificates state a combined uncertainty built by root-sum-square of the contributions, following the GUM. Worst-case (limiting) error sums are still used for quick guaranteed bounds from instrument accuracy classes.
P = VI with I = 100 mA +- 2.5% and V = 5 V +- 5%: which expressions give the consumed power?
- (A)
- (B)
- (C)
- (D)
Answer: (A) and (C) Official key: A;C
- .
- For a product, the limiting (worst-case) relative errors add: .
- of 500 mW is . So and say the same thing: (A) and (C).
Trap: subtracting the percentages () gives (B) and (D). Errors never cancel in a worst-case bound. We evaluated at the four corners: and , whose mean is the 7.5% first-order bound. The key uses this worst-case rule. A root-sum-square combination gives 5.59%, which matches no option.
Power measured from separate voltage and current readings, as in DC test benches and battery cyclers, carries both instruments' errors. Datasheet accuracy figures are limits, so quick checks add them, while calibration labs combine standard uncertainties by root-sum-square.
SI unit definition dependencies X -> Y (X needed to define Y): which order is valid today?
- (A)metre second kilogram kelvin
- (B)second metre kilogram kelvin
- (C)metre second kelvin kilogram
- (D)second kilogram kelvin metre
Answer: (B) Official key: B
- Since 2019 every SI base unit is defined by fixing a constant of nature. The second fixes the caesium-133 hyperfine frequency, , and needs no other unit.
- The metre fixes , so it needs the second.
- The kilogram fixes , so it needs the metre and the second.
- The kelvin fixes , where , so it needs the kilogram, metre and second.
- Every arrow in second metre kilogram kelvin is a real dependency: (B).
Trap: (A) and (C) start with metre second, but the second depends on no other unit. (D) ends with kelvin metre, but the metre does not use the kelvin. We wrote the dependency set of all seven base units from the 2019 definitions and tested every arrow: the failing arrows were A: metre->second; C: metre->second, kelvin->kilogram; D: kelvin->metre; B had none.
National metrology institutes realise the kilogram with Kibble balances and silicon-sphere methods tied to the Planck constant, which in turn rely on length and time standards. Calibration certificates trace back through this chain to the caesium clock that defines the second.
Copper wire 10 cm long, 1 mm diameter: measure its resistance by two-, three-, four-wire or ellipsometry?
- (A)Two wire method
- (B)Three wire method
- (C)Four wire method
- (D)Ellipsometry
Answer: (C) Official key: C
- Estimate the value first. With for copper (assumed, room temperature): .
- , so . This is a very low resistance.
- Two-wire: the same two leads carry the test current and sense the voltage, so the meter reads the wire plus both leads plus both contact resistances. Leads and clip contacts are each tens of milliohms, many times the wire itself.
- Three-wire: the RTD connection. A third lead lets a bridge subtract one lead resistance, but only if the leads match, and the contact resistances at the wire ends do not cancel.
- Four-wire (Kelvin): one pair of leads forces the current; a separate pair senses the voltage right at the wire. The voltmeter draws almost no current, so the sense leads and contacts drop almost no voltage. This is the right method: option (C).
- Ellipsometry is an optical method: it measures the change in polarisation of light reflected from a surface, to find film thickness and refractive index. It does not measure resistance.
Trap: choosing three-wire because it is the usual answer for RTD lead compensation. That cancels lead resistance in a bridge; it does not remove contact resistance, which matters at milliohm level. We modelled each method with assumed leads of 40 to 60 mΩ and contacts of 10 to 50 mΩ: the worst-case error was 10285% for two-wire, 2805% for three-wire and 22 parts per billion for four-wire with a 10 MΩ voltmeter.
Four-terminal (Kelvin) connections are standard for low resistances such as busbar joints, motor windings, shunts and cable conductors; micro-ohmmeters and Kelvin clips are built this way. Precision current-sense resistors on circuit boards also have separate sense pads for the same reason.
Which bridges have a frequency-independent balance: Maxwell, Wien, Schering, Wheatstone?
- (A)Maxwell bridge
- (B)Wien bridge
- (C)Schering bridge
- (D)Wheatstone bridge
Answer: (A) and (C) and (D) Official key: A, C, D
- A bridge balances when the products of opposite arms are equal, . Both the real and the imaginary parts must match. The balance is frequency-independent if cancels from both conditions.
- Maxwell (inductance-capacitance) bridge: , , , unknown coil . Balance gives and . No : (A) is correct.
- Wien bridge: a series arm and a parallel arm. The imaginary parts balance only when . The balance depends on frequency, which is why this bridge measures frequency: (B) is wrong.
- Schering bridge: , standard capacitor , , unknown . Balance gives and . No : (C) is correct.
- Wheatstone bridge: four resistors, . No reactance, so no frequency: (D) is correct.
- Answer: (A), (C) and (D).
Trap: assuming every AC bridge with a capacitor depends on frequency. Only bridges where survives the balance equations do (Wien, and Hay's bridge for high-Q coils). We balanced each bridge at 1 kHz by complex nodal analysis, fitted the unknown components and moved the source to 50 Hz, 300 Hz, 5 kHz and 20 kHz. The detector current stayed below A for Maxwell, Schering and Wheatstone; for the Wien bridge (balanced at its own 1592 Hz) it reached 0.19 mA per volt of supply, and at 1 kHz its fourth arm would have needed reactance.
Schering bridges are the classic way to measure the capacitance and dissipation factor of high-voltage insulation in cables, bushings and transformers. Wien bridges set the frequency of Wien-bridge oscillators, and the same balance idea sits inside modern automatic LCR meters, which null a bridge or vector ratio electronically.
Bridge balanced with R1 = 100, R2 = 210 ohm, C2 = 2.9 uF, R4 = 50 ohm; 2 kHz, 10 Vp-p: L3 in mH?
Answer: Official key: 14.40 to 14.60
- This is a Maxwell inductance-capacitance bridge. The detector joins the junction of and to the junction of and .
- At balance the two detector nodes are at the same voltage, which needs the products of opposite arms to be equal: .
- So and .
- .
- Imaginary parts: , so . Real parts: .
- cancels, so the 2 kHz and the 10 Vp-p do not enter the answer.
Trap: multiplying adjacent arms instead of opposite arms. That gives , which has a capacitive (negative) reactance, impossible for a coil. Another slip is treating and as being in series. We solved the bridge by nodal analysis with an assumed 1 kΩ detector and drove the detector current to zero with a complex secant search on : mH and . With these values the detector current stayed below A at 200 Hz and 20 kHz too.
Maxwell bridges measure medium-Q coils by comparing them with a standard capacitor, which is easier to make accurate and stable than a standard inductor. Modern LCR meters do the same comparison electronically, but bridge methods are still used in calibration labs and teaching labs.
Kelvin-type bridge, Ra = Rb = Rc = 10 milliohm, 3 mA in the lower-right arm: current through Ra at balance?
Answer: Official key: 0.99 to 1.01
- This is the layout of a Kelvin double bridge: outer ratio arms on top, inner arms and , and the link between the two lower arms.
- At balance the galvanometer carries no current. So and carry the same current and act as one path of , from one end of to the other.
- The 3 mA in the lower-right arm reaches the right end of and must split between (10 mΩ) and the , path (20 mΩ).
- Current divider: , so 1.00 mA.
Trap: splitting the current equally because all three resistors are 10 mΩ (1.5 mA), or sending all 3 mA through . is in series with , not in parallel with on its own. The paper does not give the outer arms, so we assumed lower arms of 133.33 Ω and 200.00 Ω (3 mA from the 1 V source) and 1 kΩ for the two upper arms together, solved the whole network by nodal analysis and bisected the upper arm split until the galvanometer current was zero: mA. A different lower-arm choice gave the same current.
The Kelvin double bridge measures resistances below about one ohm, such as shunts, busbars and cable samples, because its inner ratio arms cancel the resistance of the link and of the connecting leads. The same four-terminal idea is used by modern micro-ohmmeters.
AC bridge, 500 Ohm + 1 H arm, 1000 Ohm ratio arms, 5 sin(5000t) V: series R and C at balance?
- (A) and
- (B) and
- (C) and
- (D) and
Answer: (C) Official key: C
- At balance the detector nodes are at the same potential, so the products of opposite arms are equal: .
- Here , (500 with 1 H), and .
- So and , giving : option (C).
This is the Hay bridge arrangement (series R-C opposite the inductive arm). Its balance depends on frequency, so the source frequency is needed.
Trap: (B) and (D) are the right pair halved and doubled; scaling R and C together moves the real and imaginary parts of in opposite directions, so both fail. (A) swaps the two numbers. We solved the bridge by complex nodal analysis with an assumed 1 k detector, iterated the R-C arm until the detector current vanished (, ), then put each option back in: the detector current relative to the open-arm value is 0.28 (A), 0.26 (B), 2.8e-08 (C) and 0.14 (D).
Hay and Maxwell bridges are the textbook ways to measure inductance and Q against a known capacitor and resistors; Hay suits high-Q coils. In practice an LCR meter does the same job with an auto-balancing bridge and phase-sensitive detection, and it reports the result in series or parallel equivalent form at a chosen test frequency.
Resistor network, galvanometer between top and middle nodes, 90, 40 and 60 Ohm arms: Rx for zero deflection?
Answer: Official key: 58 to 62
- Look at the lower part first. From the left node the path splits into 40 then 60 through the middle node, and 40 then 60 through the bottom node. Both dividers have the same ratio, so the middle and bottom nodes sit at the same potential and the 60 between them carries no current.
- The middle node is therefore at the 40 : 60 tap of the supply: measured from the right node.
- For zero galvanometer current the top node must be at the same potential: .
- .
Trap: pairing the arms the wrong way round, , gives 135 . Match each arm of the top divider with the arm on the same side (left with left). We solved the five-node network by nodal analysis with an assumed 100 galvanometer and bisected for zero galvanometer current: 60.0000 ; the current in the middle 60 at that point was zero to machine precision.
A null-balance bridge reads resistance by adjusting a known arm until a galvanometer shows no current, so the result depends only on resistor ratios and not on the supply voltage or the meter's own calibration. Precision resistance bridges and many strain-gauge and RTD instruments still work on this principle, with an electronic null detector in place of the galvanometer.
18 V bridge with 3k/6k and 6k/3k arms as common-mode source plus two 3 V sources: R1, R2 and V_C?
- (A), , and
- (B), , and
- (C), , and
- (D), , and
Answer: (C) Official key: C
- Each half of the bridge is a divider across the 18 V supply. Open-circuit and .
- Figure (b) gives and . So , the common-mode voltage, and the 3 V sources carry half of the 6 V differential output.
- With the supply shorted, each mid-node sees its two arms in parallel: .
- Answer: (C).
Trap: taking one arm as the source resistance (A or B) instead of the parallel pair, or using 4.5 V for (B, D): that makes and , which neither divider gives. is the average of the two outputs. We reduced both figures to a two-port: open-circuit voltages and the 2 x 2 resistance matrix from test currents at and . Figure (a) gave , and 2.0 k at each node with no coupling, and only option (C) reproduced all of it.
A bridge output carries a large common-mode voltage, about half the excitation, on top of a small differential signal. This is why bridge outputs are read with a differential or instrumentation amplifier with high common-mode rejection, and why datasheets quote input common-mode range.
AC bridge, series R-L and parallel R-L arms, R = 300 Ohm, L = 30 mH, R1 = 1000, R2 = 500: balance frequency?
Answer: Official key: 1.55 to 1.65
- Arms: (top left), (top right), and below. Balance: .
- Multiply out: .
- Real parts: , so .
- Imaginary parts: , so . The given 1000 and 500 meet it.
- 1.59 kHz.
Trap: stopping at and reporting 10 kHz, which forgets the . We solved the bridge by complex nodal analysis with a 1 k detector (assumed), swept the frequency and refined the minimum: the detector current falls to zero at 1591.5 Hz. With the current never reaches zero at any frequency, so both balance conditions are needed.
Frequency-dependent bridges, such as the Wien bridge, are used to measure frequency and to set the frequency of oscillators: balance happens only at one frequency. LCR meters have replaced manual bridges for routine work, but they use the same null-or-ratio idea internally.
Bridge at 4 V with 1000/1010 and 1000/1000 arms, 100 Ohm galvanometer: galvanometer current in uA?
Answer: () Official key: 9 to 9
- Remove G. Left mid-node: . Right mid-node: . Thevenin voltage , left side higher.
- Thevenin resistance (supply shorted): on the left and on the right, so .
- , so , from left to right.
- Rounded: 9 A.
Trap: leaving out the galvanometer's 100 gives 9.93 A, which rounds to 10 and misses the key. We solved all node voltages with G in place: 9.025 A; with G replaced by a short the same solve gives 9.925 A. The open-circuit difference was 9.950 mV.
Deflection-mode bridges read a small unbalance through a meter or amplifier instead of nulling it, so the detector's own resistance enters the reading. Bridge-based scales and strain indicators therefore use a high-impedance amplifier in place of a galvanometer.
Kelvin double bridge: measures low or high resistances, built with three- or four-terminal resistors?
- (A)low ; three-terminal
- (B)low ; four-terminal
- (C)high ; three-terminal
- (D)high ; four-terminal
Answer: (B) Official key: B
- Below about , lead and contact resistances (milliohms) are comparable with the unknown. A Wheatstone bridge reads them as part of the unknown.
- The Kelvin double bridge uses four-terminal resistors: current terminals carry the test current, and separate potential terminals sense the voltage across the resistance itself.
- A second set of ratio arms (the inner arms, ) across the link between the unknown and the standard cancels the link resistance: balance gives exactly.
- Answer: low; four-terminal, option (B).
Trap: three-terminal (guarded) construction belongs to high-resistance measurement, where surface leakage is the problem. That is a different bridge. We solved a Kelvin double bridge node by node with 5 m contact and 1 m link resistances (assumed) and bisected the standard for balance. On a 1 m unknown the four-terminal error was 45 parts per million and the three-terminal error 500%. A two-terminal Wheatstone bridge erred by 1000% at 1 m and by 0.2 parts per million at 1 M, so the Kelvin bridge is needed only at the low end.
Busbar joints, motor windings, circuit-breaker contacts and current shunts are tested with micro-ohmmeters that use four-terminal (Kelvin) connections. Precision current-sense resistors are also made with four terminals for the same reason.