[Post #14/38] Source-Degenerated CS and Common-Gate (CG) Amplifiers

CONTENT_START [HERO_HERE: Schematic illustration showing the transition from a standard Common-Source amplifier to a Source-Degenerated version by inserting an R_S resistor.]

📘 Microelectronic Circuits Series — Post #14/38 — 7.2-7.3 (Theory)

Source degeneration and the Common-Gate (CG) configuration are the primary tools used by analog designers to trade gain for stability and bandwidth. While the basic Common-Source (CS) amplifier is the "workhorse" of voltage gain, it is inherently nonlinear and susceptible to process-voltage-temperature (PVT) variations; these two topologies solve those limitations by introducing negative feedback at the source.

1. Overview & Background — Why this matters

Source degeneration is like placing a stiff spring on a hydraulic valve handle: the more you push, the more the spring pushes back against your finger, making the valve's response smoother, more predictable, and less jerky. In an amplifier, the "stiff spring" is a resistor R_S placed at the source of the MOSFET. It turns an unstable "high-gain/high-sensitivity" stage into a linear, reliable block. This is fundamental in high-fidelity audio pre-amps and CMOS RF front-ends where distortion must be kept to an absolute minimum.

Figure 1. Source-degenerated CS amplifier with R_S

The Common-Gate (CG) amplifier, by contrast, is like a bouncer at a club door who refuses to let anyone enter unless they have a very specific, low-impedance signal. It doesn't provide the massive voltage gain of a CS stage, but it excels at isolating input signals from output fluctuations. It is the architectural spine of the high-speed cascode amplifier used in virtually every high-bandwidth communication chip today.

[DIAGRAM_1_HERE: Standard CS amplifier vs. Source-degenerated CS stage showing the degeneration resistor R_S at the source node.]

2. How it Works (Physical & Circuit Principles)

In a standard CS amplifier, the transconductance g_m is entirely dependent on the transistor's bias current and physical geometry, leading to massive gain fluctuations as the temperature changes. When we insert R_S, any increase in gate voltage v_gs raises the source voltage v_s via negative feedback. This "self-limiting" action reduces the effective v_gs, thereby reducing the effective transconductance g_m,eff. The circuit essentially fights against the input signal to ensure the output current is dictated more by the resistor R_S and less by the nonlinear g_m of the MOSFET.

The Common-Gate (CG) stage operates by keeping the gate constant and driving the source directly. Since the input current flows into the source, the input impedance is remarkably low, approximately 1/g_m. Because the signal is injected at the source, the output at the drain is in-phase with the input, resulting in a positive, non-inverting gain. This non-inverting nature makes it the perfect partner for the "inverting" Common-Source stage in folded-cascode architectures.

where g_m is the intrinsic transconductance (2I_D / V_OV), and R_S is the degeneration resistor. As R_S becomes large, the gain approaches R_D / R_S, becoming almost entirely independent of the transistor properties.

💡 Intuition: If g_mR_S ≫ 1, the amplifier gain becomes defined by a simple ratio of two resistors (R_D/R_S), just like an Op-Amp. This is the "Gold Standard" for linearity in analog design.

3. Key Design Equations

For a source-degenerated CS stage, the voltage gain is:

Figure 2. Common-gate (CG) amplifier circuit with signal-flow arrows

where g_mR_D is the intrinsic gain, and (1 + g_mR_S) is the factor of degeneration that trades gain for linearity and bandwidth.

For the Common-Gate (CG) stage, the input impedance looking into the source is:

where g_m determines how easily current can be pulled from the source node.

The voltage gain of the CG stage (assuming r_o is large) is:

where the sign is positive because an increase in input voltage at the source reduces V_GS, decreasing the drain current and causing the drain voltage to rise.

4. Worked Numerical Example — Calculate it yourself

Consider a 180-nm CMOS NMOS transistor with a transconductance g_m = 5 mA/V. We wish to design a gain stage with a drain resistor R_D = 2 kΩ.

Step 1: Calculate the intrinsic gain without degeneration.

A_v = g_m · R_D = (5 mA/V) · (2 kΩ) = 10 V/V.

Step 2: Add R_S = 200 Ω to improve linearity.

g_mR_S = (5 mA/V) · (0.2 kΩ) = 1.0.

Step 3: Calculate the effective gain.

A_v,eff = 10 / (1 + 1) = 5 V/V.

The gain is halved, but the linearity has improved significantly because the feedback loop now compensates for the MOSFET's nonlinear V_GS vs I_D curve.

[DIAGRAM_2_HERE: Small-signal equivalent circuit of a CG amplifier showing the input resistance 1/g_m at the source.]

5. Design Considerations & Trade-offs

• Linearity: Higher R_S linearizes the transconductance but burns voltage headroom and reduces gain.
• Output Resistance: Source degeneration boosts the output resistance of a CS stage to approximately r_o(1 + g_mR_S), allowing for a better current source approximation.
• Input Impedance (CG): The low 1/g_m input impedance is a double-edged sword; it is excellent for current-mode signaling but terrible for driving high-impedance voltage sources.
• Noise: Adding a real physical resistor R_S adds thermal noise (4kT R_S), which can degrade the signal-to-noise ratio in sensitive low-power receivers.

6. Where it Shows Up in Practice

The CG amplifier is the core of the telescopic cascode and folded-cascode operational transconductance amplifiers (OTAs) used in high-precision ADCs (like the AD7606). Source degeneration is ubiquitous in the output stage of Class-AB power amplifiers and the input pair of wideband instrumentation amplifiers, where stable gain over process corners is mandatory.

7. Common Pitfalls & Debugging Tips

• ⚠️ Headroom Crunch: Forgetting that R_S consumes DC bias voltage. If your V_DD is 1.2 V, a large R_S might push your transistor out of saturation.
• ⚠️ Body Effect: In a bulk CMOS process, the body effect (g_mb) reduces the effective degeneration, as the bulk node is typically tied to ground, not the source.

8. Exam & Interview Hot Spots

• 💡 "If you have a CS stage and want to increase gain without increasing R_D, what do you do?" (Answer: Increase g_m by increasing bias current, or decrease R_S).
• 💡 "Why is the CG amplifier used in RF?" (Answer: Its low input impedance matches 50 Ω lines naturally, and its high reverse isolation prevents local oscillator leakage).

9. Key Takeaways

• Source degeneration trades gain for improved linearity and higher output impedance.
• The Common-Gate amplifier provides non-inverting gain with very low input impedance.
• g_m,eff = g_m / (1 + g_mR_S) is the governing equation for degenerated stages.
• The CG amplifier is an essential building block for high-speed, wide-bandwidth cascode architectures.
• Always check for DC headroom constraints before adding a source degeneration resistor.

Educational content only. Always verify with datasheets and SPICE simulation before production design.