Feedback Amplifiers Questions and Answers

Answer: C

This is the Barkhausen criterion. Nyquist's test examines encirclement of the critical point in the complex plane, Bode's construction is the straight-line magnitude and phase plot, and Routh's is an algebraic stability test on a polynomial. Four names attached to four distinct tools for the same question, and using the right name is what makes a stability argument followable by someone else.

Answer: D

It lowers the output impedance, because sampling the output voltage makes the amplifier work to hold that voltage steady whatever the load draws. The two effects follow from two independent rules - series raises at the input, voltage sampling lowers at the output - and together they define the voltage amplifier this topology produces. Reading a topology's two words as instructions gives both impedance directions without any memorisation. Framing the question as which quantity the change acts upon makes the answer follow without calculation.

Answer: A

Everything feedback provides scales with the loop gain, which is the open-loop gain times the fraction, so a very large open-loop gain leaves a large loop gain even when the fraction is small and the closed-loop gain high. A gain of a hundred thousand still leaves a loop gain of a hundred at a closed-loop gain of a thousand. Load driving, offset and output swing are all separate specifications addressed by other parts of the design. Designing one parameter to excess so that a whole family of others can be traded from it is the economics of the operational amplifier.

Answer: B

An internally compensated device has one pole placed deliberately at a few hertz so that the gain has fallen to unity before any second pole contributes phase, which makes it stable at any closed-loop gain down to one. The devices are perfectly capable at higher frequencies, and neither the stage count nor the output stage is the limitation. The pairing of enormous gain with a tiny bandwidth is a deliberate design choice rather than a shortcoming, and the product of the two is what the user actually buys.

Answer: C

Without the dominant pole its open-loop response extends much further, so at a closed-loop gain above the specified minimum it delivers considerably more bandwidth than a compensated device would. Offset and low-frequency gain are unrelated, and stability becomes more of a concern rather than less. A device that shifts a design constraint onto the user in exchange for performance is a reasonable bargain when the user knows the constraint, and a trap when they do not.

Answer: D

The internal compensation capacitor is charged from a limited current, so the output cannot change faster than that current divided by the capacitance, whatever the small-signal bandwidth suggests. It is a large-signal limit quite separate from the small-signal roll-off, and a sine wave of large amplitude can trigger it well inside the nominal bandwidth. That two independent limits govern the same amplifier, one for small signals and one for large, is why both are specified and why exceeding either produces quite different distortion.

Answer: A

The steepest slope of a sine wave is at its zero crossing:

\[ \begin{aligned} 2\pi f A &\le 10^{6} \\ f &\le \tfrac{10^{6}}{2\pi \times 10} \approx 16 \ \mathrm{kHz} \end{aligned} \]

giving about 16 kHz. Above that the output becomes a triangle rather than a sine, however wide the small-signal bandwidth. The limit falls in proportion to amplitude, so the same amplifier reaches ten times the frequency at a tenth of the output.

Answer: B

Once the demanded slope exceeds what the internal current can deliver, the output simply ramps at its maximum rate until the input reverses, and a succession of constant-slope ramps is a triangle. Amplitude clipping flattens the peaks instead, which is a different and distinguishable distortion. The loop does effectively lose control during the ramp, though that is a description of the symptom rather than its cause. A distortion whose shape identifies its mechanism is worth recognising, since it names the specification being exceeded.

Answer: C

Slew limiting is a limit on how fast the output can change, set by an internal current and a capacitance, while clipping is a limit on how far it can go, set by the supply rails. Both occur at the output, neither is a temperature effect, and both affect any waveform. The distinction matters practically because the remedies differ entirely: a faster amplifier for one, a higher supply or a smaller signal for the other. A limit that survives making every external component ideal is an intrinsic one, and worth knowing for that reason.

Answer: D

Distortion is divided by one plus the loop gain, and since the open-loop gain falls steadily above the dominant pole, the loop gain and hence the correction fall with it. Device linearity is roughly frequency-independent over the audio range, the network's phase affects stability rather than distortion directly, and the bias current does not fall. A specification quoted at one frequency and assumed at all others is a common source of disappointment, and this is the mechanism behind it.