How it works
Negative feedback feeds a fraction β of the output back to subtract from the input. Closed-loop gain Af = A/(1 + Aβ), where A is open-loop gain and Aβ is the loop gain. If Aβ >> 1, then Af ≈ 1/β, independent of A — gain is set entirely by the passive feedback network. Four topologies exist based on how feedback is sampled at the output and summed at the input: series-series (samples voltage, sums voltage — increases Zin and Zout), series-shunt (voltage amplifier feedback — increases Zin, decreases Zout), shunt-series (current amp — decreases Zin, increases Zout), shunt-shunt (transresistance — decreases both). Negative feedback also reduces distortion by factor (1 + Aβ) and extends bandwidth: BWf = BW × (1 + Aβ).
Key points to remember
Gain with negative feedback: Af = A/(1 + Aβ). Sensitivity improvement: percentage change in Af = (1/(1+Aβ)) × percentage change in A, so a 10% transistor gain variation causes less than 0.1% variation in Af for Aβ = 100. Bandwidth extends by factor (1+Aβ): an amplifier with BW = 100 kHz and Aβ = 9 achieves BWf = 1 MHz. Noise and distortion generated within the amplifier (not at the input) reduce by (1+Aβ). Positive feedback causes oscillation when Aβ ≥ 1 at 0° phase shift — Barkhausen criterion. The four topologies must be identified from circuit diagrams by checking whether feedback is at emitter (series) or base (shunt) and whether output is taken from collector voltage or emitter current.
Exam tip
The examiner always asks you to identify the type of feedback topology in a given amplifier circuit — check whether feedback signal is added in series or shunt at input, and whether output is sampled as voltage (from collector) or current (from emitter resistor), then state all four effects on gain, Zin, Zout, and bandwidth.