RF Cascade Budget Analysis: Friis Chains, IIP3 Cascade, and Yield Analysis With Monte Carlo

Why Nominal Values Aren't Enough

Every RF front-end datasheet lists a noise figure and gain at a single operating point — 25°C, nominal supply, center frequency. Production parts arrive with distributions. A 2 dB NF LNA has a distribution that might span 1.5 to 2.5 dB across process corners. If your receiver sensitivity spec requires 2.0 dB cascaded NF with a 0.5 dB margin, that margin evaporates when production corners stack up.

The RF Cascade Budget Analyzer addresses this by computing cascaded NF, gain, IIP3, and P1dB using Friis formulas, then running vectorized Monte Carlo over per-stage tolerances to produce yield statistics and sensitivity rankings. The inputs are a JSON stage list — you define as many stages as your chain has, with any combination of amplifiers, mixers, attenuators, and filters.

Setting Up a 5-Stage Receive Chain

The example chain is a 2.4 GHz receiver: LNA → band-select filter → mixer → IF amplifier → IF filter. Enter the following stage list:

[
  {"name": "LNA",           "gain": 15.0, "nf": 1.5, "iip3": -5.0,  "nf_tol": 0.3, "gain_tol": 0.5, "iip3_tol": 1.5},
  {"name": "BPF",           "gain": -1.5, "nf": 1.5, "iip3": 30.0,  "nf_tol": 0.2, "gain_tol": 0.2, "iip3_tol": 0.0},
  {"name": "Mixer",         "gain": -6.0, "nf": 7.5, "iip3": 12.0,  "nf_tol": 0.5, "gain_tol": 0.5, "iip3_tol": 2.0},
  {"name": "IF Amplifier",  "gain": 20.0, "nf": 5.0, "iip3": 20.0,  "nf_tol": 0.4, "gain_tol": 0.5, "iip3_tol": 1.5},
  {"name": "IF Filter",     "gain": -2.0, "nf": 2.0, "iip3": 30.0,  "nf_tol": 0.2, "gain_tol": 0.3, "iip3_tol": 0.0}
]

The nf_tol, gain_tol, iip3_tol fields specify ±σ tolerances for the Monte Carlo. Passive components (filter, attenuator) use zero IIP3 tolerance since their linearity is essentially fixed by physics.

Nominal Cascade Results

With these inputs the tool computes:

The cascaded NF of 2.31 dB confirms the chain meets a 2.5 dB sensitivity budget with 0.19 dB margin — barely. The IIP3 cascade is dominated by the mixer: despite the LNA's high gain, the mixer's relatively modest IIP3 of +12 dBm referred to its input becomes roughly −9 dBm referred to the system input after accounting for 13.5 dB of gain ahead of it.

Per-Stage Contribution Breakdown

The cumulative gain and NF breakdown shows the LNA contributes 1.50 dB to the cascaded NF (100% of its own NF since it's the first stage), the BPF adds 0.09 dB (attenuated by the LNA gain), and the mixer contributes 0.67 dB. This matches the Friis intuition: the LNA dominates, and every dB of LNA gain directly reduces every subsequent stage's NF contribution.

The IIP3 per-stage breakdown is the inverse: later stages with high gain ahead of them dominate the cascaded IIP3. The mixer at position 3 (13.5 dB gain ahead) contributes the majority of the IIP3 degradation. Improving the IF amplifier's IIP3 from 20 to 30 dBm changes cascaded IIP3 by less than 0.3 dB — it's not the bottleneck.

Monte Carlo: From Nominal to Production Yield

Set 200,000 trials with Gaussian distributions (tolerances are 1σ). The Monte Carlo perturbs all stage parameters simultaneously and computes the full Friis cascade for each trial.

The results:

Against a 2.5 dB NF limit, yield is 78.3%: roughly one in five production boards will exceed the spec with these component tolerances. That's a significant yield problem for a consumer product.

The sensitivity analysis reveals the LNA NF tolerance (±0.3 dB, 1σ) accounts for 47% of the cascaded NF variance. The mixer NF tolerance accounts for 31%. The remaining stages contribute 22% combined.

Improving Yield Without Changing the Schematic

The sensitivity breakdown points directly to the fix: tighten the LNA NF tolerance. Changing from a 0.3 dB to 0.15 dB NF tolerance (1σ) — achievable by tighter incoming inspection or a higher-grade LNA variant — pushes yield to 91.4%. No schematic change, no new components.

Alternatively, shifting the nominal LNA NF from 1.5 dB to 1.2 dB (selecting a higher-performance part) while keeping the same tolerance increases yield to 93.8% and also improves the median cascaded NF to 2.01 dB, providing a comfortable 0.49 dB margin.

The second scenario costs more per LNA but dramatically reduces the tail risk. The tool lets you quantify that tradeoff before committing to a BOM.

SFDR and the Dynamic Range Design Constraint

The cascaded IIP3 determines the spurious-free dynamic range — the range of input signal powers over which neither noise nor intermodulation products dominate:

With a noise floor of −115 dBm (kTBF for 1 MHz bandwidth at 2.31 dB NF) and cascaded IIP3 of −10.8 dBm, SFDR = (2/3)(−10.8 − (−115)) = 69.5 dB. The tool presents this in dB·Hz^(2/3) normalized form. If two co-channel interferers are present at −45 dBm, their IM3 products appear at −10.8 + 2(−10.8 − (−45)) = −44 dBm — right at the interference level. The SFDR result flags this immediately as a potential cross-modulation problem at high input levels.

RF Cascade Budget Analyzer