Home / Courses / CSIR NET Physics / Practice / Amplifiers and feedback

Amplifiers and feedback

Electronics · 77 questions and templates for CSIR NET Physics, in order from warm-up to top rank. Check each answer, then read the full solution.

Loading questions… (they need JavaScript; nothing you type is sent anywhere)

Written questions with solutions

The same questions as a list, so you can print them or read without the practice box. Templates (fresh numbers) appear only in the practice box above.

  1. MCQWarm-up

    An amplifier of gain A has a fraction β of its output fed back to subtract from the input. The closed-loop gain is
    1. A/(1−Aβ)
    2. A/(1+Aβ)
    3. A(1+Aβ)
    4. 1/(Aβ) exactly
    Answer and solution

    Answer: B

    The amplifier sees xi=xs−βxo and gives xo=Axi. So xo=A(xs−βxo) and xo/xs=A/(1+Aβ). When the loop gain Aβ≫1 this tends to 1/β, set by the feedback network alone.

    Why people get this wrong: A/(1−Aβ) is positive feedback; 1/β is only the limit for large loop gain.

  2. MCQWarm-up

    A voltage gain of 100 expressed in decibels is
    1. 20 dB
    2. 10 dB
    3. 40 dB
    4. 100 dB
    Answer and solution

    Answer: C

    For voltage (or current) ratios, gain in dB =20log10⁡(100)=40 dB. A power ratio of 100 would be 10log10⁡100=20 dB.

    Why people get this wrong: Using 10log⁡ (the power formula) for a voltage ratio.

  3. MCQWarm-up

    Well above the corner frequency of a single-pole low-pass amplifier, the magnitude of its gain falls at
    1. 20 dB per decade (6 dB per octave)
    2. 40 dB per decade
    3. 3 dB per decade
    4. 10 dB per octave
    Answer and solution

    Answer: A

    |A|=A0/1+(f/fH)2≈A0fH/f for f≫fH: ten times the frequency, one tenth of the gain, which is −20 dB. Doubling the frequency halves it: −6 dB per octave. Each extra pole adds another −20 dB/decade.

    Why people get this wrong: The 3 dB figure is the drop at the corner itself, not the slope.

  4. NATWarm-up

    An amplifier with open-loop gain 1000 has negative feedback with β=0.009. Its closed-loop gain is ______.
    Answer and solution

    Answer: 100

    Loop gain Aβ=9, so Af=1000/(1+9)=100.

    Why people get this wrong: Using 1/β=111 ignores that the loop gain here is not large.

  5. MCQWarm-up

    Which of these is reduced, rather than improved, when negative feedback is applied to an amplifier?
    1. Bandwidth
    2. Harmonic distortion
    3. Sensitivity of the gain to transistor changes
    4. The magnitude of the gain
    Answer and solution

    Answer: D

    Negative feedback trades gain for everything else: the gain falls by the factor 1+Aβ, while bandwidth grows by about that factor, and distortion and sensitivity to A fall by it.

    Why people get this wrong: Reading "bandwidth" as something feedback reduces: for a single-pole amplifier it rises.

  6. MSQWarm-up

    In voltage-series (series-shunt) negative feedback, the output voltage is sampled and a voltage is fed back in series with the input. Which statements are correct?
    1. The input resistance rises by the factor 1+Aβ
    2. The output resistance falls by the factor 1+Aβ
    3. The voltage gain is stabilised near 1/β
    4. The output resistance rises by the factor 1+Aβ
    Answer and solution

    Answer: A, B, C

    Series mixing opposes the input voltage, so the source pushes less current for the same voltage: Rif=Ri(1+Aβ). Shunt sampling of the output voltage fights any change in it, so the output looks stiffer: Rof=Ro/(1+Aβ). The non-inverting op-amp amplifier is the textbook example.

    Why people get this wrong: Output resistance rises only when the output current is sampled (series sampling).

  7. MCQWarm-up

    Three voltage amplifier stages with gains of 20 dB, 15 dB and −5 dB (an attenuating filter) are cascaded without loading. The overall voltage gain is a factor of about
    1. 31.6
    2. 30
    3. 1000
    4. 1500
    Answer and solution

    Answer: A

    Decibel gains add: 20+15−5=30 dB. As a voltage ratio, 1030/20=101.5=31.6.

    Why people get this wrong: 30 is the dB number itself; 1000 is what 30 dB means for a power ratio, not a voltage ratio.

  8. NATStandard

    An amplifier has loop gain Aβ=99. Its open-loop gain A drifts by 20%. The closed-loop gain changes by about ______ % (small-change estimate).
    Answer and solution

    Answer: 0.2 % (±0.01)

    Differentiate Af=A/(1+Aβ): dAfAf=11+AβdAA. So the change is 20%/100=0.2%.

    Why people get this wrong: Dividing by Aβ=99 instead of 1+Aβ=100 changes little here, but matters when the loop gain is small.

  9. NATStandard

    An amplifier has midband gain 2000 and a single upper cut-off at 50 kHz. Negative feedback with β=0.0045 is applied. The new upper cut-off is ______ kHz.
    Answer and solution

    Answer: 500 kHz (±1)

    Loop gain Aβ=9, so 1+Aβ=10. With one pole, A(s)=A0/(1+s/ωH) gives Af(s)=A0/(1+A0β)1+s/[ωH(1+A0β)]: gain 200, cut-off 50×10=500 kHz. The gain-bandwidth product stays 108 Hz.

    Why people get this wrong: Dividing the cut-off by the factor instead of multiplying (that is what happens to a lower cut-off).

  10. MCQStandard

    A capacitance of 2 pF connects the input and output of an inverting stage with voltage gain −99. Seen from the input, it behaves like a capacitance to ground of
    1. 2 pF
    2. 198 pF
    3. 100 pF
    4. 200 pF
    Answer and solution

    Answer: D

    The voltage across the capacitor is vi−(−99vi)=100vi, so it draws 100 times the current a grounded 2 pF would: CM=C(1−K)=2×100=200 pF.

    Why people get this wrong: C|K|=198 pF forgets the input side of the capacitor also moves.

  11. NATStandard

    A common-emitter stage has gm=40 mA/V, collector resistor 5 kΩ and an unbypassed emitter resistor of 200Ω (ro and the load ignored). The magnitude of its voltage gain is ______ (to one decimal place).
    Answer and solution

    Answer: 22.2 (±0.1)

    |Av|=gmRC1+gmRE=40×51+40×0.2=2009=22.2. The emitter resistor is local series feedback with loop gain gmRE=8.

    Why people get this wrong: Using RC/RE=25 is the large-loop-gain limit, not exact.

  12. MCQStandard

    A 10μF coupling capacitor feeds an amplifier whose input resistance is 2 kΩ (ideal voltage source driving it). The lower 3 dB frequency it sets is about
    1. 50 Hz
    2. 8.0 Hz
    3. 0.8 Hz
    4. 80 Hz
    Answer and solution

    Answer: B

    The capacitor and the input resistance form a high-pass RC: fL=1/(2πRC)=1/(2π×2000×10−5)=7.96 Hz.

    Why people get this wrong: Forgetting the 2π gives 50 Hz.

  13. NATStandard

    Three identical non-interacting amplifier stages, each with a single upper cut-off at 1 MHz, are cascaded. The overall upper 3 dB frequency is ______ MHz (to two decimal places).
    Answer and solution

    Answer: 0.51 MHz (±0.005)

    Each stage gives 1/1+x2 with x=f/f1. Three in a row: (1+x2)−3/2=1/2, so 1+x2=21/3 and x=21/3−1=0.510. The overall bandwidth is 0.51 MHz.

    Why people get this wrong: Thinking the bandwidth stays 1 MHz, or dividing by 3.

  14. MCQStandard

    The highest possible collector efficiency of an ideal class B push-pull amplifier with sinusoidal output is
    1. 25%
    2. 50%
    3. 78.5%
    4. 100%
    Answer and solution

    Answer: C

    At full swing Vp=VCC: Pout=VCC2/2RL and the supply delivers Pdc=2VCC2/(πRL). The ratio is π/4=78.5%. Class A reaches 25% with a resistive load and 50% with a transformer.

    Why people get this wrong: 50% is class A with transformer coupling.

  15. MCQStandard

    A differential amplifier has differential gain 200 and common-mode gain 0.02. Its common-mode rejection ratio is
    1. 80 dB
    2. 40 dB
    3. 100 dB
    4. 20 dB
    Answer and solution

    Answer: A

    CMRR =|Ad/Acm|=200/0.02=104, which is 20log10⁡104=80 dB.

    Why people get this wrong: Using 10log⁡ gives 40 dB.

  16. MSQStandard

    An amplifier has a single high-frequency pole at fH and is driven by a voltage step. Which statements are correct?
    1. Its 10-90% rise time is about 0.35/fH
    2. Doubling fH halves the rise time
    3. The output overshoots before settling
    4. Under negative feedback its gain-bandwidth product stays the same
    Answer and solution

    Answer: A, B, D

    One pole gives vo∝1−e−t/τ with τ=1/(2πfH); from 10% to 90% takes τln⁡9=2.2τ=0.35/fH. A first-order response never overshoots. Feedback divides the gain and multiplies fH by the same 1+A0β.

    Why people get this wrong: Overshoot needs at least two poles (a complex pair).

  17. NATStandard

    An amplifier with open-loop gain 2000 gives a closed-loop gain of 50 with negative feedback. The feedback factor β is ______ (to four decimal places).
    Answer and solution

    Answer: 0.0195 (±0.0002)

    1+Aβ=A/Af=2000/50=40, so Aβ=39 and β=39/2000=0.0195.

    Why people get this wrong: Taking β=1/Af=0.02 assumes infinite loop gain.

  18. MCQStandard

    An amplifier of gain A=10 has a fraction β=0.05 of its output added in phase to its input (positive feedback). The closed-loop gain is
    1. 6.67
    2. 10
    3. 20
    4. infinite (it oscillates)
    Answer and solution

    Answer: C

    With positive feedback Af=A/(1−Aβ)=10/(1−0.5)=20. The loop gain Aβ=0.5 is below 1, so the circuit is stable, but the gain now depends strongly on A: a 10% rise in A would raise Af to 11/0.45=24.4.

    Why people get this wrong: 6.67 is the negative-feedback answer A/(1+Aβ); oscillation needs Aβ=1.

  19. NATStandard

    A power transistor may run with its junction at up to 150∘C. Its thermal resistances are junction-to-case 1.5∘C/W, case-to-sink 0.5∘C/W and sink-to-air 3.0∘C/W, in air at 25∘C. The largest power it may dissipate continuously is ______ W.
    Answer and solution

    Answer: 25 (±0.1)

    Heat flows through the three thermal resistances in series: θ=1.5+0.5+3.0=5.0∘C/W. Tj=Ta+Pθ, so Pmax=(150−25)/5.0=25 W.

    Why people get this wrong: Using only the sink-to-air resistance gives 41.7 W and overheats the junction.

  20. NATExam level

    A transistor with β=100 is biased at IC=1 mA (VT=25 mV) in a common-emitter stage with an unbypassed 500Ω emitter resistor. The small-signal resistance looking into the base is ______ kΩ (to one decimal place).
    Answer and solution

    Answer: 53 kΩ (±0.1)

    rπ=βVT/IC=2.5 kΩ. The emitter resistor carries (β+1) times the base current, so it appears (β+1) times larger from the base: Rin=rπ+(β+1)RE=2.5+101×0.5=53.0 kΩ.

    Why people get this wrong: Adding RE itself rather than (β+1)RE.

  21. MCQExam level

    A common-emitter stage is driven from a 1 kΩ source (take rπ≫1 kΩ). It has Cπ=10 pF, Cμ=1 pF and a midband gain of −49. The input pole is at about
    1. 15.9 MHz
    2. 3.18 MHz
    3. 2.65 MHz
    4. 159 MHz
    Answer and solution

    Answer: C

    Miller: Cμ appears at the input as Cμ(1+49)=50 pF. Total input capacitance 10+50=60 pF. fp=1/(2π×103×60×10−12)=2.65 MHz.

    Why people get this wrong: 15.9 MHz ignores Cμ; 3.18 MHz forgets Cπ. The tiny feedback capacitor dominates.

  22. NATExam level

    A voltage amplifier has output resistance 2 kΩ and open-loop gain 500. Voltage-series feedback with β=0.1 is applied. The output resistance becomes ______ Ω (to one decimal place).
    Answer and solution

    Answer: 39.2 Ω (±0.2)

    Sampling the output voltage divides the output resistance by 1+Aβ=51: Rof=2000/51=39.2Ω.

    Why people get this wrong: Multiplying by 51 (that is what series mixing does to the input resistance).

  23. MCQExam level

    An amplifier with input resistance 5 kΩ is given shunt (current) mixing at its input with loop gain 24. The input resistance becomes
    1. 125 kΩ
    2. 200 Ω
    3. 5 kΩ
    4. 208 Ω
    Answer and solution

    Answer: B

    Shunt mixing draws extra current from the source through the feedback path, so the input looks smaller: Rif=Ri/(1+Aβ)=5000/25=200Ω. The inverting op-amp input is the extreme case (a virtual ground).

    Why people get this wrong: 208 Ω divides by 24 instead of 25.

  24. MCQExam level

    The loop gain of a feedback amplifier falls to magnitude 1 at a frequency where its phase is −135∘. The phase margin is
    1. 135°
    2. −45°
    3. 90°
    4. 45°
    Answer and solution

    Answer: D

    Phase margin =180∘+∠T at the frequency where |T|=1: 180−135=45∘. The loop would oscillate if the phase reached −180∘ with |T|≥1; 45° is a common minimum, 60° gives little ringing.

    Why people get this wrong: Reading the phase itself as the margin.

  25. NATExam level

    A loop gain has three identical real poles: T(s)=T0/(1+s/ωp)3. The largest T0 for which the closed loop is stable is ______.
    Answer and solution

    Answer: 8

    The phase reaches −180∘ when each pole gives −60∘: ω/ωp=tan⁡60∘=3. There |T|=T0/(1+3)3/2=T0/8. Stability needs |T|<1 at that frequency, so T0<8. (This is also the gain condition of the three-section RC phase-shift oscillator with buffered sections.)

    Why people get this wrong: Taking ω=ωp, where each pole gives only −45∘.

  26. MSQExam level

    Which statements about power amplifier classes are correct?
    1. In class A the transistor conducts over the whole cycle (360°)
    2. In class B each transistor conducts for half the cycle (180°)
    3. Class C conducts for less than 180° and is used with tuned loads at radio frequency
    4. Class AB conducts for more than 360°
    Answer and solution

    Answer: A, B, C

    Conduction angles: A 360°, AB between 180° and 360°, B 180°, C below 180°. A class C collector current is a train of pulses; a tuned circuit picks out the fundamental, giving efficiencies above 78.5%. No class conducts for more than a full cycle.

    Why people get this wrong: AB sits between A and B; it cannot exceed the full cycle.

  27. NATExam level

    A class B push-pull stage runs from ±20 V into an 8Ω load and delivers a sine wave of 16 V peak. Its efficiency is ______ % (to one decimal place).
    Answer and solution

    Answer: 62.8 % (±0.2)

    Pout=Vp2/2RL=256/16=16 W. Each supply delivers current only on its half cycle, average Ip/π, so Pdc=2VCCVp/(πRL)=2×20×16/(8π)=25.46 W. Efficiency =16/25.46=62.8%=(π/4)(Vp/VCC).

    Why people get this wrong: Quoting 78.5%, which needs a full swing Vp=VCC.

  28. MCQExam level

    Crossover distortion in a class B complementary push-pull output stage is caused by
    1. the dead band near zero input, where neither transistor is conducting until |vBE| reaches about 0.6 V
    2. saturation of both transistors at the peaks
    3. unequal current gains of the npn and pnp transistors
    4. the output coupling capacitor
    Answer and solution

    Answer: A

    Each transistor needs about 0.6-0.7 V of base-emitter drive before it conducts, so the output sticks at zero while the input crosses through that band. Biasing both transistors slightly on (two diodes or a VBE multiplier between the bases) gives class AB and removes the dead band; overall negative feedback reduces what is left.

    Why people get this wrong: Clipping at the peaks is a different distortion.

  29. MCQExam level

    A BJT differential pair has a tail current of 1 mA from a current source of output resistance 50 kΩ; the output is taken from one collector (VT=25 mV, large β). Its CMRR is about
    1. 40 dB
    2. 54 dB
    3. 66 dB
    4. 60 dB
    Answer and solution

    Answer: D

    Each transistor carries 0.5 mA, so gm=20 mA/V. Single-ended: Ad=gmRC/2 and Acm≈RC/(2REE). CMRR =gmREE=20×50=1000, that is 60 dB.

    Why people get this wrong: Using the whole tail current in gm (40 mA/V) gives 66 dB.

  30. MCQExam level

    An amplifier has a midband gain of 60 dB and real poles at 1 kHz and 100 kHz. On the asymptotic Bode plot, its gain at 10 kHz is
    1. 60 dB
    2. 20 dB
    3. 40 dB
    4. 57 dB
    Answer and solution

    Answer: C

    From 1 kHz the gain falls at 20 dB/decade, so one decade later (10 kHz) it is 60−20=40 dB. The second pole only starts to act near 100 kHz; beyond it the slope is 40 dB/decade.

    Why people get this wrong: 57 dB is the true gain at the first corner (1 kHz), not at 10 kHz.

← All topics · Timed mock test →