Differential Amplifiers Questions and Answers

Answer: A

A low source resistance generates little thermal noise of its own, so what matters is the amplifier's input noise voltage, and a bipolar pair run at a good current has the lowest available. Field effect devices excel with high-impedance sources where their negligible input current dominates, and degeneration raises the noise voltage. Which family is quieter depends entirely on the source impedance, which is why low-noise design begins by measuring the source rather than by choosing a device.

Answer: B

It is the tail source, and its dynamic impedance is what decides the stage's common-mode rejection. A mirror is the arrangement often used to build it and to form the active load, a reference is what sets the current it copies, and a bootstrap is a different technique for raising an apparent impedance. Several related ideas appear in one input stage, and naming each by the job it does keeps the schematic readable.

Answer: C

Slew rate is set by the tail current and the compensation capacitance, neither of which depends on how well the two devices match. The other three all improve directly with matching, since each arises from an imbalance between the halves. Sorting a stage's specifications into those limited by matching and those limited by absolute values is what decides where design effort belongs, and slew rate falls on the other side of that line. Reading the definition literally, rather than recalling a formula, is the reliable route through questions of this shape.

Answer: D

Emitter-coupled logic steers a fixed tail current into one branch or the other according to which input is higher, and never lets a device saturate - which is why it switches so fast. Rejection, active loads and trimming belong to the linear applications. That the fastest logic family ever built is a differential pair operated as a switch is worth knowing, and it explains both its speed and its considerable appetite for power.

Answer: A

The tail current is drawn whether the gate is switching or resting, so a gate consumes the same power idle as busy - the exact opposite of complementary logic, which draws almost nothing at rest. Its devices deliberately never saturate, which is where the speed comes from. That the fastest family is also the hungriest is the trade that eventually lost it to complementary logic as circuit densities rose.

Answer: B

The Gilbert cell stacks two differential pairs on top of a third, so one signal steers the tail current and the other steers within each upper pair, giving a four-quadrant multiplier. The other options describe simpler arrangements. It is the standard analogue multiplier and balanced mixer, and recognising it as three pairs rather than an unfamiliar block makes its operation follow from what a single pair already does.

Answer: C

Four-quadrant working means either input may be positive or negative and the output's sign follows correctly, covering all four sign combinations. Amplitude, device count and frequency range are unrelated. A multiplier restricted to one quadrant can only handle unidirectional signals, which is why the distinction matters when choosing one for a mixer that must handle a bipolar waveform.

Answer: D

Each input appears equally at both sides of the output pair, so taking the output differentially cancels it while the product term, which appears in antiphase, survives. No filter is needed for the inputs themselves, the tail source does not block them, and they are applied differentially rather than in antiphase to each other. That the suppression falls out of the balance rather than needing a filter is exactly why the arrangement is called balanced and why it displaced the simple mixer.

Answer: A

A single-device mixer passes its local oscillator straight through to the intermediate frequency stages, where it is far stronger than any wanted signal and must be filtered out; a balanced mixer cancels it instead. Conversion loss is comparable, an oscillator is still needed, and frequency range depends on the devices. Removing a large unwanted output by symmetry rather than by filtering is both cheaper and more effective.

Answer: B

Integrated processes control the ratio between adjacent components far better than any absolute value, and every important property of the pair - offset, drift, rejection - depends on matching. It uses more devices rather than fewer, needs a tail source, and single-rail operation is a design matter. That integrated design rests on matching while discrete design rests on component values is the single division that explains why the two look so different.