LNA
Understanding Low Noise Amplifiers
The LNA is the most critical component in any receiver. According to the Friis cascade formula, the noise performance of every subsequent stage is suppressed by the LNA gain. A well-designed LNA with high gain and low noise figure effectively masks the noise of the mixer, IF amplifier, and all downstream components.
Key LNA Parameters
- Noise figure (NF): The noise added by the amplifier. Lower is better. State-of-the-art room-temperature LNAs achieve below 0.5 dB at L-band and below 2 dB at Ka-band.
- Gain: Typically 15 to 35 dB. Must be high enough to suppress downstream noise but not so high that the LNA compresses on strong signals.
- P1dB (compression point): The input power at which gain drops by 1 dB. Determines the maximum signal the LNA can handle linearly.
- IP3 (third-order intercept): Characterizes linearity. Higher IP3 means better performance with multiple strong signals present.
- Stability: Must be unconditionally stable across all frequencies to prevent oscillation.
LNA Technologies
- GaAs pHEMT: The workhorse for microwave LNAs. Excellent noise performance from 1 to 40 GHz.
- InP HEMT: Best noise performance above 40 GHz. Used in radio astronomy and satellite receivers.
- GaN HEMT: Higher linearity and robustness than GaAs. Increasingly used where survivability matters (radar front ends).
- SiGe BiCMOS: Lower cost, integrates with digital circuits. Suitable for consumer applications up to ~30 GHz.
Design Tradeoffs
LNA design involves a fundamental tradeoff between noise figure and input match. The minimum noise impedance (Zopt) of a transistor is different from its conjugate match impedance. Designers must choose between noise-optimal matching (lowest NF, imperfect input VSWR) and power-match (perfect VSWR, slightly higher NF). Many designs use a compromise that achieves near-minimum NF with acceptable return loss.
NF_sys = NF_LNA + (NF_mixer - 1) / G_LNA
Example: NF_LNA = 1 dB (F=1.26), G_LNA = 25 dB (316x)
NF_mixer = 8 dB (F=6.31)
F_sys = 1.26 + (6.31-1)/316 = 1.277
NF_sys = 1.06 dB
Input P1dB typically: -20 to 0 dBm
Output IP3 typically: +10 to +30 dBm
LNA Technology Comparison
| Technology | Frequency | Noise Figure | Gain | IP3 |
|---|---|---|---|---|
| GaAs pHEMT | DC - 40 GHz | 0.3 - 2.0 dB | 15 - 30 dB | +15 to +25 dBm |
| InP HEMT | DC - 110 GHz | 0.5 - 3.0 dB | 15 - 25 dB | +10 to +20 dBm |
| GaN HEMT | DC - 40 GHz | 1.0 - 3.0 dB | 15 - 25 dB | +25 to +40 dBm |
| SiGe BiCMOS | DC - 30 GHz | 1.0 - 4.0 dB | 10 - 25 dB | +10 to +20 dBm |
| Cryogenic InP (4K) | DC - 100 GHz | 0.02 - 0.3 dB | 25 - 40 dB | -10 to +5 dBm |
Frequently Asked Questions
What does an LNA do?
A low noise amplifier amplifies weak received signals while adding minimal noise. It is placed as the first active component in a receiver chain so that its high gain suppresses the noise contribution of all subsequent stages. The LNA noise figure directly sets the sensitivity of the receiver.
What is a good noise figure for an LNA?
It depends on frequency. At L-band (1-2 GHz), state-of-the-art LNAs achieve 0.3 to 0.8 dB. At Ka-band (26-40 GHz), 1.5 to 3.0 dB is typical. Cryogenically cooled LNAs used in radio astronomy achieve below 0.1 dB.
Why is the LNA placed first in the receiver?
The Friis formula shows that the first stage noise dominates the cascade. Every dB of loss before the LNA (from cables, switches, or filters) adds directly to system noise figure. Placing the LNA as close to the antenna as possible minimizes this pre-LNA loss and maximizes receiver sensitivity.