Test and Measurement Equipment Calibration and Uncertainty Informational

How do I calculate the expanded uncertainty of a noise figure measurement?

Q: Why is NF uncertainty larger than S-parameter uncertainty?

S-parameter measurements (VNA) benefit from full vector error correction (12-term model removes systematic errors). NF measurements (Y-factor) use a scalar correction (only magnitude, no phase information for the noise signals). The noise signals are random and cannot be phase-referenced. This fundamental difference means NF measurements have inherently higher uncertainty than S-parameter measurements of the same device.

Q: How does DUT gain affect NF uncertainty?

Higher DUT gain: the receiver (NF meter) noise contribution is reduced by the DUT gain. For gain > 20 dB: the receiver noise is negligible (< 1% of the DUT output noise). The receiver NF correction is small, and its uncertainty is small. Lower DUT gain: the receiver noise is significant relative to the DUT output noise. The receiver NF correction is large, and its uncertainty is amplified. For gain < 0 dB (passive device): the receiver noise dominates, and the NF measurement becomes very sensitive to small errors in the receiver NF calibration.

Q: What is the best achievable NF measurement uncertainty?

Y-factor method (optimized): ±0.15-0.25 dB (k=2) for DUT gain > 20 dB. Cold source method (PNA-X, optimized): ±0.10-0.15 dB (k=2). Radiometric method (cryogenic reference, national standards lab): ±0.02-0.05 dB. The radiometric method is used only by national metrology institutes (NIST, PTB, NPL) as the primary NF standard.

How do I calculate the expanded uncertainty of a noise figure measurement? Noise figure measurement has multiple uncertainty sources, and the total expanded uncertainty is typically ±0.2-0.5 dB for a well-executed measurement (significantly larger for less controlled setups): (1) Uncertainty sources: ENR uncertainty (u_enr): from the noise source calibration certificate. Typical: ±0.15-0.20 dB (k=2). Standard uncertainty: 0.075-0.10 dB. This is often the dominant or second-largest contributor. Mismatch uncertainty (u_mm): caused by impedance mismatch between the noise source, the DUT, and the receiver. Input mismatch: between noise source and DUT input. Output mismatch: between DUT output and NF meter input. Can be calculated from measured reflection coefficients using: u_mm = |Γ_source × Γ_DUT_in| × ENR_linear for the input side. Typical total: 0.05-0.15 dB (standard). Receiver NF correction uncertainty (u_rx): the NF meter second-stage correction introduces uncertainty. Depends on the DUT gain (higher DUT gain reduces the impact of receiver NF). For DUT gain > 20 dB: u_rx < 0.05 dB. For DUT gain < 10 dB: u_rx can be 0.1-0.3 dB. Instrument uncertainty (u_inst): NF meter noise and linearity. Typical: 0.02-0.05 dB. Connector repeatability (u_conn): from connecting and disconnecting the DUT and noise source. Typical: 0.02-0.05 dB per connection (3 connections in the measurement path). (2) Combination: u_c = √(u_enr² + u_mm² + u_rx² + u_inst² + u_conn²). Example for a 15 dB gain LNA at 10 GHz: u_enr = 0.10 dB. u_mm = 0.08 dB. u_rx = 0.04 dB. u_inst = 0.03 dB. u_conn = 0.04 dB. u_c = √(0.010 + 0.0064 + 0.0016 + 0.0009 + 0.0016) = √(0.0205) = 0.143 dB. Expanded uncertainty (k=2): U = 2 × 0.143 = 0.29 dB. (3) Key insight: for a high-gain DUT (> 30 dB): u_rx becomes negligible, and ENR + mismatch dominate. Total U ≈ ±0.25-0.35 dB. For a low-gain DUT (< 10 dB): u_rx becomes significant. Total U ≈ ±0.4-0.6 dB. For passive devices (attenuators, filters): the NF = attenuation, and the measurement uncertainty is highest (u_rx dominates because gain < 0 dB).

Category: Test and Measurement Equipment

Updated: April 2026

Product Tie-In: Calibration Kits, Standards, Cables

NF Measurement Uncertainty

Noise figure uncertainty is particularly important for receiver system design, where the front-end NF directly determines the system sensitivity and range.

Performance verification: confirm specifications against the application requirements before finalizing the design
Environmental factors: temperature range, humidity, and vibration affect long-term reliability and parameter drift
Cost vs. performance: evaluate whether the application demands premium components or standard commercial grades

Common Questions

Frequently Asked Questions

Why is NF uncertainty larger than S-parameter uncertainty?

S-parameter measurements (VNA) benefit from full vector error correction (12-term model removes systematic errors). NF measurements (Y-factor) use a scalar correction (only magnitude, no phase information for the noise signals). The noise signals are random and cannot be phase-referenced. This fundamental difference means NF measurements have inherently higher uncertainty than S-parameter measurements of the same device.

How does DUT gain affect NF uncertainty?

Higher DUT gain: the receiver (NF meter) noise contribution is reduced by the DUT gain. For gain > 20 dB: the receiver noise is negligible (< 1% of the DUT output noise). The receiver NF correction is small, and its uncertainty is small. Lower DUT gain: the receiver noise is significant relative to the DUT output noise. The receiver NF correction is large, and its uncertainty is amplified. For gain < 0 dB (passive device): the receiver noise dominates, and the NF measurement becomes very sensitive to small errors in the receiver NF calibration.

What is the best achievable NF measurement uncertainty?

Y-factor method (optimized): ±0.15-0.25 dB (k=2) for DUT gain > 20 dB. Cold source method (PNA-X, optimized): ±0.10-0.15 dB (k=2). Radiometric method (cryogenic reference, national standards lab): ±0.02-0.05 dB. The radiometric method is used only by national metrology institutes (NIST, PTB, NPL) as the primary NF standard.

Keep Reading

How do I calculate the expanded uncertainty of a noise figure measurement?

NF Measurement Uncertainty

Frequently Asked Questions

Why is NF uncertainty larger than S-parameter uncertainty?

How does DUT gain affect NF uncertainty?

What is the best achievable NF measurement uncertainty?

Related Questions

Related Terms

Request a Quote