Noise Figure Calculator

This tool computes the total (cascaded) noise figure of a multi-stage RF receiver chain using Friis’ noise formula. It converts stage gains and noise figures from dB to linear units, applies the cascade equation, and returns the overall noise figure in dB. The result reflects how much the system degrades the signal-to-noise ratio relative to an ideal noiseless system.

Cascaded noise figure determines the sensitivity limit of receivers—especially critical in low-SNR applications like satellite communications, radar, and 5G front-ends. Since the first stage dominates overall noise performance, this calculator helps engineers prioritize low-noise amplification early in the chain and evaluate trade-offs between gain distribution and noise degradation.

“Gain (dB)” is the small-signal power gain of that stage (e.g., LNA, mixer, IF amplifier). “NF (dB)” is its noise figure—the dB increase in input-referred noise due to internal device noise. Both must be entered as positive values; negative gain implies loss and will correctly increase total noise figure if entered accurately.

Yes—enter the component’s loss (e.g., –3 dB) as its “Gain (dB)”, and its physical temperature-derived noise figure: NF (dB) = Loss (dB) for a matched passive device at room temperature. For example, a 6 dB attenuator has Gain = –6 dB and NF = 6 dB, significantly degrading cascaded noise performance.

Low-noise amplifiers (LNAs) typically have NF = 0.5–2.5 dB and gain = 10–25 dB. Mixers range from NF = 5–12 dB with near-zero or negative gain. IF amplifiers may have NF = 3–8 dB and gain = 15–30 dB. Always verify datasheet specs—excessively optimistic values yield misleading cascade results.

This is expected—and correct—if later stages have low gain or high loss. Friis’ formula weights earlier stages more heavily, but poor gain after a noisy stage allows downstream noise to contribute significantly. Double-check that gain values are positive where amplification occurs and negative only for actual lossy elements.

The current interface supports three stages—the most common configuration for receiver front-ends. For >3 stages, apply the calculator iteratively: compute NF and gain for Stages 1–2, treat that pair as a single “Stage A”, then cascade Stage A with Stage 3 (and so on). All intermediate results must stay in linear units for accuracy.

No—they’re related but distinct. Noise figure (NF) is a logarithmic ratio (dB), while noise temperature (T_n) is linear (kelvin): T_n = 290 × (10^NF/10 – 1). This calculator uses NF exclusively, but you can convert results using standard formulas when interfacing with thermal noise models or cryogenic systems.

No—it assumes perfect impedance matching (50 Ω) and ideal power transfer between stages. Real-world mismatches cause gain ripple, reflection-induced noise, and altered effective noise figure. For precision designs, perform EM simulation or measure S-parameters and use advanced noise parameter models beyond Friis’ assumptions.

Only if those stages exhibit significant analog noise contribution (e.g., ADC input buffers or active filters). Pure digital processing adds quantization noise—not modeled here. For ADC-dominated chains, use effective number of bits (ENOB) and sampling rate to estimate SNR, then reference back to analog front-end noise figure.