K-Factor
Understanding K-Factor Stability
Stability analysis is the first step in any amplifier design. Before optimizing for gain, noise figure, or power, the designer must ensure the amplifier will not oscillate at any frequency, with any possible source and load impedance it might encounter. An oscillating amplifier is worse than no amplifier at all, as it generates spurious signals that interfere with all nearby receivers. K-factor analysis provides the mathematical framework for this critical assessment.
The physics behind K-factor is straightforward: an amplifier oscillates when the signal fed back through S12 from output to input arrives in phase with the input and with sufficient amplitude to sustain the loop (Barkhausen criterion). Higher S12 (more reverse leakage) and higher S21 (more gain) increase the oscillation risk. The K-factor combines all four S-parameters into a single metric that captures this risk for all possible termination impedances simultaneously.
Stability Equations
K = (1−|S11|²−|S22|²+|Δ|²) /
(2|S12||S21|)
Δ = S11S22 − S12S21
Stable: K > 1 AND |Δ| < 1
μ-factor (Edwards-Sinsky):
μ = (1−|S11|²) /
(|S22−ΔS11*| + |S12S21|)
Stable: μ > 1 (single test)
Higher μ = more stability margin
Stability circle (load):
Center = (S22−ΔS11*)* / (|S22|²−|Δ|²)
Radius = |S12S21| / ||S22|²−|Δ|²|
Stability Analysis Summary
| Condition | K | μ | Action | Risk |
|---|---|---|---|---|
| Unconditionally stable | >1, |Δ|<1 | >1 | Any ZS/ZL OK | None |
| Conditionally stable | <1 | <1 | Check stability circles | Moderate |
| Potentially unstable | <1 at some f | <1 at some f | Add stabilization R | High |
| Resistively stabilized | >1 all f | >1 all f | Gain traded for safety | Low |
| Out-of-band unstable | <1 out-of-band | <1 out-of-band | Lossy stub / R at low f | High (hidden) |
Frequently Asked Questions
What does K>1 mean?
Unconditionally stable: no oscillation for ANY passive source/load impedance. Safe even with antenna VSWR changes (ice, nearby objects). K<1: certain impedances within stability circles cause oscillation. Must either stabilize (add series/shunt R, trade gain for K>1) or constrain impedances to safe regions. Check K at ALL frequencies, not just in-band.
How do stability circles work?
When K<1: circles on Smith chart divide stable/unstable impedance regions. Load stability circle: load impedances causing instability. Source stability circle: source impedances causing instability. Calculated from S-parameters (center and radius formulas). Stable region: side of circle containing Z_0 (if stable at Z_0). Design constraint: keep matching within stable region.
K vs. μ-factor?
μ-factor: single test (μ>1), no auxiliary condition needed. Higher μ = more margin (distance from instability on Smith chart). K=5 vs K=2 does NOT tell relative stability (|Δ| also matters). μ preferred in modern tools. K remains in literature/datasheets due to history. Both computed from same S-parameters, equivalent results.