GATE Electronics (EC): Electromagnetics (Maxwell's equations, plane waves and polarisation, transmission lines and Smith chart, waveguides and antennas), previous-year questions solved, part 3
Part 3 of 4 of the Electromagnetics hub: 8 more official GATE EC questions on transmission lines, Smith chart and S-parameters, each solved step by step and checked against the official answer key. Part 1 covers electrostatics, currents and Maxwell's equations, plane waves and polarisation and waveguides, antennas and link budgets.
Questions from official GATE EC papers (IIT Kharagpur 2022, IIT Kanpur 2023, IISc Bengaluru 2024, IIT Roorkee 2025, IIT Guwahati 2026); solutions by Skill-Lync. · Updated 17 Nov 2026 · 8 of 22 questions on this page
S-parameters of a two-port [S] after adding lossless line sections of theta1 and theta2 to its ports
- (A)
- (B)
- (C)
- (D)
Answer: (A) Official key: A
- is the ratio of an outgoing wave to an incoming wave at the reference planes and . Moving each reference plane outward by a lossless line of length only delays a wave by each way (time factor , a wave travelling towards goes as ).
- : the incident wave at the outer end reaches the network as . The reflected wave comes back to the outer end as . So . In the same way .
- : , , and . So , and by the same argument .
- The matrix is : option (A).
Trap: the phases are all negative, because the extra length sits between the new ports and the network. The all-positive matrix of option (C) is what you get when you remove a line from the measurement (de-embedding), and (B) and (D) mix the two signs. The code turned six random complex matrices into ABCD form, multiplied by the ABCD matrix of each line, converted back to and compared with all four options: only (A) matched (largest difference below ), and the matrix of (C) equals the one found with lines of length and .
Vector network analysers do exactly this in reverse. Cables and fixtures between the instrument and the device add terms, and port extension or de-embedding multiplies by to move the reference plane back to the device. A common mistake in the lab is a wrong sign on the extension, which rotates the Smith-chart trace the wrong way.
SWR 2.0 on a 50 ohm line, minima 30 cm apart, first minimum 10 cm from the load: equivalent Rm and lm
- (A)
- (B)
- (C)
- (D)
Answer: (B) and (C) Official key: B,C
- Successive minima are apart, so and . Then , and is , an electrical length of .
- At a voltage minimum the line impedance is real and smallest: . At a voltage maximum it is real and largest: . Maxima lie from the minima.
- The load plane is and distance is measured towards the generator. Minima are at and . Maxima are at and .
- The equivalent circuit puts at the end of a piece of the same line of length beyond the load plane, so sits at . A resistance sits at a minimum: , so . A resistance sits at a maximum: , so .
- Check: gives . A line long ending in and a line long ending in both have this input impedance. Answer: (B) and (C).
Trap: the pairs with (A) and with (D) swap the resistance and the length. The code evaluated on a grid (minima at and , SWR ) and tested every option by its input impedance: (A) gives and (D) gives , neither equal to . Choosing only one of (B) and (C) loses the marks: both are correct, because the line repeats every .
This is the slotted-line way of finding an unknown load: measure the SWR, the minimum spacing and the first minimum position. The same data tell you where a series section or a quarter-wave transformer must go to match the load, which is how the matching sections on antenna feeds are placed.
Which lumped circuits (L, T, pi forms) represent an infinitesimal section of a transmission line?
- (A)
- (B)
- (C)
- (D)
Answer: (B) and (C) and (D) Official key: B,C,D
- A line of length has total series impedance and total shunt admittance . As a circuit is a correct model if it contains all of both, in any order: the order differs only by terms of higher order in .
- (B) has the full shunt , and then the full series , : an L-section. Valid.
- (C) splits the shunt branch into two halves around the full series branch: a section. Valid.
- (D) splits the series branch into two halves around the full shunt branch: a T section. Valid.
- (A) has only and as its shunt elements, and they are on different sides of the series branch. Half of the shunt admittance is missing, so it models a line with half the and half the . Not valid.
- Answer: (B), (C) and (D).
Trap: (A) looks symmetric and uses the right symbols, so it is easy to accept. The code cascaded copies of each circuit as ABCD matrices (, , , , , ) and compared with the exact lossy line. At the worst relative error is 4.7e-05 (B), 4.7e-05 (C) and 4.7e-05 (D), falling as grows, while (A) stays at 2.55. The same cascade of (A) matches a line with and to 2.4e-05.
Circuit simulators model a PCB trace or cable as a chain of such sections (RLGC) and use many of them so the section is short compared with the wavelength, usually under a tenth of it and often a twentieth. Using a T, pi or L form gives the same answer once the sections are short enough.
Shorted lossless line: normalised input impedances at lambda/8, lambda/4, 3lambda/8 and lambda/2 from the load
- (A), , ,
- (B), , ,
- (C), , ,
- (D), , ,
Answer: (A) Official key: A
- A short circuit has load impedance . For a lossless line, the input impedance at a distance from the load is , which for gives , with .
- Point A is from the load: , so .
- Point B is from the load: , , so . A quarter-wave line turns a short into an open circuit.
- Point C is from the load: , , so .
- Point D is from the load: , , so . The impedance repeats every half wavelength.
- That is , option (A). The printed is a slip: a normalised impedance has no unit.
Trap: option (C) is the answer for an open-circuited line, , which gives . Options (B) and (D) contain , which never appears: takes only , and at multiples of . The code built the line from lumped -sections per wavelength (series , shunt ), started from , , and read at the four points: A , B , C , D . Only option (A) matched. With an open end instead (, ) the same cascade gave A , B , C , so option (C) is the open-circuit pattern.
A shorted stub that is long looks like an open circuit at its design frequency, so on microstrip it is used to feed DC bias to a transistor without loading the RF path. Shorter than (here ) a shorted line behaves as an inductor, , and stubs are cut to that length as small tuning inductances.
Lossless 50 ohm line with |Gamma| = 0.6: maximum input impedance magnitude moving towards the generator
Answer: Official key: 200 to 200
- On a lossless line the reflection coefficient at a distance from the load is . Its magnitude stays and only its phase turns, once round every .
- The input impedance is . For a fixed , is largest when is real and positive (phase ): the numerator is as large and the denominator as small as they can be.
- That point is a voltage maximum, where the impedance is purely resistive: , which is with .
- The smallest magnitude is at : . Moving along the line, the impedance visits both every , whatever the load.
Trap: the value comes from using only and forgetting to divide by , and is the minimum, not the maximum. Both are outside the official key. The code built the line as lumped -sections per wavelength and terminated it in loads with at phases , , and , scanned half a wavelength of , and found the same maximum , , , and minimum every time. The travelling-wave solution , gave the same and .
A VSWR of on an antenna feed means the line impedance swings between and along its length on a system, which is why antenna and amplifier datasheets specify a maximum VSWR. The ratio of the voltage maximum to the voltage minimum is what a slotted line or a directional coupler reading reports.
Average power delivered to a 50 - j75 ohm load on a lossless 50 ohm line when the incident power is 10 mW
If the average incident power on the line is , then the average power delivered to the load (in mW, rounded off to one decimal place) is __________.
Answer: Official key: 6.3 to 6.5
- Reflection coefficient at the load: .
- Magnitude squared: , so .
- The line is lossless, so the reflected power travels back with the same size it had at the load: .
- Power delivered to the load .
Trap: mW is the reflected power, and it is outside the key range to . The code took for mW, found the load power directly as mW, and as incident minus reflected ( mW). It then built the line as lumped LC sections ( wavelengths, ABCD matrices) driven by a matched source: the load power is mW with , the same as the closed form mW.
Mismatch loss is what a VNA return-loss reading turns into: |Gamma| = 0.6 is a return loss of about 4.4 dB, and the 36% of the power sent back is the reason a power amplifier or antenna feed is matched to its 50 ohm line. The reactive part of the load (here -j75 ohm) is what a matching network or stub is added to cancel.
Percentage of power dissipated in a lossy two-port at the point where |S11| and |S21| intersect at -4 dB
For a lossy passive two-port network, and intersect at dB.
The percentage of power dissipated in the lossy network at the intersection frequency is ______.
(rounded off to two decimal places)
Answer: Official key: 20.00 to 21.00
- Magnitudes in dB are , so in linear power is . For a wave entering port 1 (port 2 matched), the fraction reflected is and the fraction delivered to port 2 is .
- Lossless check: the two curves meet at dB, so each power fraction is . Their sum is : all the power is reflected or transmitted and none is lost.
- Lossy case: at dB, .
- Sum of reflected and transmitted: .
- The rest is absorbed in the network: . So of the incident power is dissipated.
Trap: using as the power fraction (the amplitude factor, not the power) gives , which is more than , and a negative loss of , which is impossible for a passive network. The code checked the lossless case with a series reactance (), where dB and exactly, then built a lossy pi network (shunt , series in units of ) solved to dB. Adding up the real power taken by each resistive element for an incident wave gave , equal to .
On a vector network analyser, a filter, attenuator or cable is characterised by S11 and S21 in dB; the quantity 1 - |S11|^2 - |S21|^2 is the share of power that was absorbed in the part, which is how insertion loss is separated into mismatch and true dissipation. A passive part with |S11|^2 + |S21|^2 greater than 1 on the trace is a sign of a bad calibration.
Load with reflection coefficient 0.5 at 30 degrees: input impedance at the diametrically opposite Smith chart point
A co-axial cable with a characteristic impedance of is connected to the load. The new input impedance of the load now moves to a diametrically opposite point on the same circle on the Smith chart.
Which option is the nearest input impedance of the cable connected load (in )?
- (A)
- (B)
- (C)
- (D)
Answer: (B) Official key: B
- Moving along a lossless line keeps fixed and turns clockwise by . A diametrically opposite point on the same circle is a turn of , so the new reflection coefficient is . (This needs , a quarter-wave cable.)
- .
- Normalised impedance: .
- Multiply by : , nearest to option (B), .
- Check with the quarter-wave rule. The load is and , the same value.
Trap: option (D), , is the load itself, that is, the point before moving. Option (C), , is its complex conjugate, which is the load mirrored about the real axis () and not rotated. Both were confirmed by the code. It also built the cable as a cascade of short lumped and sections (total electrical length , ) ending in : the input impedance came out as and to within . A scan of the cable length from to found that the shortest length that puts at is (the next is , since the pattern repeats every ).
A quarter-wave section of line turns an impedance Z into Z0^2/Z, the basis of quarter-wave matching transformers and of the way a 50 ohm cable length can make a load look very different at the radio end. On a VNA the Smith chart shows this as the trace rotating about the centre by twice the electrical length.