Question 5 of 5: Common-Gate Amplifier inside a Voltage-Series Feedback Loop — Input and Output Resistance
Nivaar worked solution (AI-drafted; not reviewed by a licensed engineer)
Notes on this paper
Paper format. National Exams, December 2019 — 16-Elec-B5
Advanced Electronics. Three hours; CLOSED BOOK, an approved Casio or Sharp
calculator permitted. Five questions, ALL to be answered, 20 marks each. Op-amps are
ideal and the supplies are ±15 V unless a question states otherwise; in
schematics ground and chassis are common. Where a question leaves an interpretation
open, the paper's own Note 1 invites the candidate to state the assumption inside the
answer — every such assumption below is flagged in a check callout.
Reference texts. A. S. Sedra and K. C. Smith, Microelectronic
Circuits, 8th ed. — Ch. 7 (transistor amplifiers and biasing), Ch. 8
(differential and multistage amplifiers, offset voltage), Ch. 10 (frequency response,
short-circuit time constants), Ch. 11 (output stages and power amplifiers), Ch. 12
(feedback topologies). B. Razavi, Fundamentals of Microelectronics, 2nd ed.,
Ch. 8 (operational amplifier as a black box, non-linear feedback elements) and
Ch. 10 (differential amplifiers). P. Gray, P. Hurst, S. Lewis and R. Meyer,
Analysis and Design of Analog Integrated Circuits, 5th ed., Ch. 3
(single-transistor amplifiers) and Ch. 8 (feedback).
Check — the paper's own MOSFET convention.
The formula sheet printed with Question 1 defines the saturation current as
$i_{DS}=\tfrac12 K\,(v_{GS}-V_{TH})^2$ with $K=K'(W/L)=\mu C_{ox}(W/L)$, and that
convention is used throughout Questions 1 and 5. Question 2 is the one place where it
does not close: with $K=1\ \text{mA/V}^2$ and $V_{GS}=2$ V the sheet would give
$I_D=0.5$ mA, whereas the bias network of Question 2 fixes $I_D=1$ mA by three
independent routes (the gate divider, $V_S/R_S$, and the stated $V_D$). The bias data
therefore governs there, and the transconductance is taken as $g_m=2I_D/V_{ov}$,
which is convention-free. Both readings are quoted in Question 2, Step 2.
Question 5: Common-Gate Amplifier inside a Voltage-Series Feedback Loop — Input and Output Resistance (20 marks)
Find. The resistance looking into the source terminal ($R_{IN}$) and the
resistance looking into the drain terminal ($R_{OUT}$), first with the feedback network
removed and then with it in place.
Figure 5.1 — The common-gate stage of Question 5. The signal enters the source through $C_{IN}$; $I_{BIAS}$ sets the drain current; the divider $R_1$–$R_2$ samples the output voltage at the drain and returns a fraction of it to the gate, in series with the gate–source loop. This is voltage-series (series–shunt) feedback.
Approach. Get $g_m$ from the bias current, write the two
resistances for the bare common-gate stage, then identify the feedback topology and solve
the three-node small-signal circuit exactly — the closed forms that come out are then
cross-checked against the feedback factor $1+A\beta$ and against the limit that must
reproduce part (a).
Fix the operating point and the transconductance. The tail source
forces $I_D = I_{bias} = 2$ mA through $M_1$. Using the paper's own saturation law
$I_D = \tfrac12 K(V_{GS}-V_{TH})^2$,
$$\begin{aligned}
V_{ov} &= \sqrt{\frac{2I_D}{K}} = \sqrt{\frac{2\times 2\ \text{mA}}{1\ \text{mA/V}^2}} = 2\ \text{V},\\
g_m &= K V_{ov} = \sqrt{2KI_D} = 2\ \text{mA/V}.
\end{aligned}$$
$$\boxed{g_m = 2\ \text{mA/V},\qquad \frac{1}{g_m} = 500\ \Omega}$$
Because $\lambda = 0$, the output resistance of the device itself is infinite, $r_o=\infty$,
which simplifies every expression below.
Part (a) — the bare common-gate stage. With $R_1=\infty$ and
$R_2=0$ the gate is grounded and there is no feedback. Driving the source with a test
voltage $v_s$ makes $v_{gs}=-v_s$, so the drain current is $i_d = g_m v_{gs} = -g_m v_s$ and
the current drawn from the test source is $i_x = g_m v_s$. Hence
$$R_{IN} = \frac{v_s}{i_x} = \frac{1}{g_m} = 500\ \Omega .$$
Looking into the drain with the source grounded, the device presents $r_o=\infty$ in
parallel with the drain resistor, so
$$R_{OUT} = R_D\parallel r_o = R_D = 2\ \text{k}\Omega .$$
$$\boxed{R_{IN} = 500\ \Omega,\qquad R_{OUT} = 2\ \text{k}\Omega}$$
The low input resistance is the defining feature of the common-gate connection — it is
a current buffer, well matched to a source that wants to see a low impedance.
Part (b) — identify the topology before computing anything. The
divider $R_1$–$R_2$ is connected across the output node, so it samples the output
voltage; it returns its tap to the gate, which is in series with the
gate–source control loop rather than in parallel with the input node. That is
voltage-series (series–shunt) feedback, and its signature is unambiguous: it should
raise the input resistance and lower the output resistance, both by the
same factor $1+A\beta$. The feedback factor is the divider ratio,
$$\beta_f = \frac{R_2}{R_1+R_2} = \frac{100}{300} = \frac{1}{3},$$
and the open-loop gain of the common-gate stage is $A = g_m R_D = 2\times 2 = 4$, so the
loop gain is $A\beta_f = 4/3$ and $1+A\beta_f = 7/3 = 2.333$.
Write the gate node, which is the only new equation. The gate draws no
current, so $R_1$ and $R_2$ form an unloaded divider between the drain and ground:
$$\begin{aligned}
\frac{v_g-v_d}{R_1}+\frac{v_g}{R_2} &= 0,\\
v_g &= \beta_f\,v_d,\quad\text{where}\ \ \beta_f = \frac{R_2}{R_1+R_2}.
\end{aligned}$$
Everything else follows from KCL at the drain and the source.
Solve for the output resistance exactly. For $R_{OUT}$ the input source
is set to zero, which grounds the source terminal through $C_{IN}$, so $v_s=0$ and
$v_{gs}=v_g$. Injecting a test current $i_t$ into the drain, the three paths that carry it
away are the drain resistor, the feedback divider, and the channel:
$$i_t = \frac{v_d}{R_D}+\frac{v_d-v_g}{R_1}+g_m v_g
= v_d\left[\frac{1}{R_D}+\frac{1}{R_1+R_2}+\frac{g_m R_2}{R_1+R_2}\right],$$
where the middle term used $v_d-v_g = v_d R_1/(R_1+R_2)$. Therefore
$$R_{OUT}=\left[\frac{1}{R_D}+\frac{1}{R_1+R_2}+\frac{g_mR_2}{R_1+R_2}\right]^{-1}
= R_D\parallel(R_1+R_2)\parallel\frac{R_1+R_2}{g_mR_2}.$$
Substituting the numbers, the three conductances are
$0.500\ \text{mS}$, $0.00333\ \text{mS}$ and $0.667\ \text{mS}$, summing to
$1.170\ \text{mS}$, so
$$\boxed{R_{OUT} = 855\ \Omega}$$
Compare $R_D/(1+A\beta_f) = 2000/2.333 = 857\ \Omega$: the two agree to within 0.3 per
cent, the small difference being the direct shunting of the drain by the 300 kΩ
divider, which the ideal feedback formula omits.
Solve for the input resistance exactly. Driving the source terminal with
$v_s$, KCL at the drain gives
$$\frac{v_d}{R_D}+\frac{v_d-v_g}{R_1}+g_m(v_g-v_s)=0
\quad\Longrightarrow\quad v_d\left[\frac{1}{R_D}+\frac{1}{R_1+R_2}+g_m\beta_f\right]=g_mv_s,$$
whose bracket is precisely $1/R_{OUT}$ from Step 5, so $v_d = g_m R_{OUT}\,v_s$. The current
taken from the test source is the channel current, $i_x = g_m(v_s-v_g) = g_m(v_s-\beta_f v_d)$,
and therefore
$$R_{IN} = \frac{v_s}{i_x} = \frac{1}{g_m\left(1-\beta_f g_m R_{OUT}\right)}
= \frac{1}{g_m}+\frac{R_DR_2}{R_D+R_1+R_2}.$$
Numerically $\beta_f g_m R_{OUT} = (1/3)(2\ \text{mA/V})(855\ \Omega) = 0.570$, so
$$R_{IN}=\frac{1}{(2\ \text{mA/V})(0.430)}=1.162\ \text{k}\Omega,$$
and the equivalent additive form gives the same value,
$500 + (2\times 100)/(302) \times 10^3\ \Omega = 500+662 = 1162\ \Omega$:
$$\boxed{R_{IN} = 1.16\ \text{k}\Omega}$$
The ideal-feedback estimate $(1/g_m)(1+A\beta_f)=500\times 2.333=1167\ \Omega$ again agrees
to better than half a per cent.
Check the result against the limit that must reproduce part (a). Letting
$R_1\to\infty$ and $R_2\to 0$ sends $\beta_f\to 0$, so the closed forms collapse to
$R_{OUT}\to R_D = 2\ \text{k}\Omega$ and $R_{IN}\to 1/g_m = 500\ \Omega$ — exactly the
part (a) answers, which is the cheapest available check on the algebra of Steps 5 and 6.
Both directions of change are also the ones the topology predicts: the input resistance rose
by a factor of 2.32 and the output resistance fell by 2.34, both within one per cent of the
predicted $1+A\beta_f = 2.333$, and the closed-loop voltage gain
$v_{OUT}/v_{IN}=g_mR_{OUT}=1.71$ V/V has fallen from the open-loop 4 V/V by the same
$1+A\beta_f$ — a textbook demonstration that feedback trades gain for terminal
impedances.