Noise Matching
Understanding Noise Matching
Noise matching is the foundation of LNA design. Every transistor has a specific source impedance that produces minimum noise figure. Presenting this impedance, rather than the standard conjugate match for gain, is the key to achieving the lowest possible system noise figure.
Noise Matching Concepts
- Gamma_opt: The source reflection coefficient that gives NFmin. Found on the noise circles on the Smith Chart.
- NFmin: The minimum achievable noise figure of the transistor.
- Rn: Noise resistance. Indicates how rapidly NF increases as the source impedance deviates from Gamma_opt. Lower Rn = more forgiving.
Noise vs Power Match
- If Gamma_opt = S11* (conjugate): no trade-off. Rare.
- Typically Gamma_opt != S11*: must choose between noise match (NF close to NFmin) and power match (maximum gain).
- Inductive degeneration can align noise and power match simultaneously. The standard LNA technique.
Frequently Asked Questions
What is noise matching?
Noise matching presents the optimum source impedance (Gamma_opt) to a transistor to achieve minimum noise figure. This is generally different from the conjugate match for maximum gain, creating a noise-gain trade-off in LNA design.
What is inductive degeneration?
Adding a small inductor in the transistor source/emitter rotates the input impedance on the Smith Chart, aligning the noise match and power match circles. This allows simultaneous good NF and good input match, the standard technique for LNA design.
How much does noise match matter?
The importance depends on Rn. With low Rn (<10 ohms), the transistor is forgiving and NF degrades slowly from Gamma_opt. With high Rn (>50 ohms), NF is very sensitive to source impedance. Modern transistors generally have low Rn.