Manufacturing and Production Assembly and Test Informational

How do I design an RF calibration procedure for a production test station?

A well-designed calibration procedure ensures that production RF test measurements are accurate, traceable, and consistent over time: (1) Calibration scope: the calibration must cover: the test equipment (VNA, power meter, spectrum analyzer, noise figure analyzer), the test fixture (cables, connectors, probes, adapters), and the DUT interface (the connection between the fixture and the device under test). Each element contributes measurement error; calibration removes or quantifies these errors. (2) VNA calibration for S-parameter measurement: SOLT (Short-Open-Load-Thru): the most common calibration method. Uses four calibration standards: short, open, load (50 Ω), and thru. Applied at the cable ends (at the connector interface with the test fixture). Corrects for: cable loss, connector reflections, and VNA systematic errors. Does not correct for the test fixture (the fixture loss and reflections must be de-embedded separately). TRL (Thru-Reflect-Line): uses calibration standards embedded in the test fixture. More accurate than SOLT for fixture-based testing (the reference plane is moved to the DUT interface). Requires designing TRL standards into the fixture (a thru connection, a reflect (short or open), and a line of known length). (3) Power calibration: calibrate the power meter with a traceable reference (NIST-traceable power sensor). Measure the power at the DUT input plane (including the cable and fixture loss). Record the loss from the test equipment to the DUT plane. Apply this loss as a correction factor in the test software. (4) Calibration interval: daily: verify calibration using a known check standard (a golden DUT or a characterized attenuator). If the check standard measurement is within tolerance: the calibration is still valid. If out of tolerance: recalibrate. Monthly: full recalibration of the VNA and test equipment. Annually: send the calibration standards and test equipment to an accredited calibration laboratory for NIST-traceable calibration. (5) Traceability: all calibration records must be traceable to national standards (NIST, PTB, NPL). The calibration chain: NIST primary standards → calibration lab secondary standards → production test equipment → production measurements. Each link in the chain has a known uncertainty. The total measurement uncertainty is the combined uncertainty of all links.
Category: Manufacturing and Production
Updated: April 2026
Product Tie-In: Assembly Materials, Test Equipment

RF Production Calibration

Calibration is the foundation of production test accuracy. Without proper calibration, a production test station may pass bad units or reject good ones, both of which cost money.

  • 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

What is a check standard?

A check standard (also called a verification standard) is a characterized passive device used to verify the test station calibration without performing a full recalibration. Examples: a precision 10 dB attenuator with known S-parameters (±0.05 dB). A filter with known center frequency and bandwidth. A golden DUT with known measured performance. Usage: measure the check standard before each production run. Compare the measured values to the expected values. If the agreement is within the acceptance criteria (e.g., ±0.2 dB for S21): the calibration is valid. If not: recalibrate. Check standard values should be traceable to a reference measurement (lab-grade VNA, NIST-calibrated).

How do I de-embed the test fixture?

De-embedding removes the effect of the test fixture from the DUT measurement: (1) Physical de-embedding: measure the fixture characteristics (S-parameters of the fixture without the DUT). The VNA software mathematically removes the fixture response from the total measurement. The fixture is modeled as two 2-port networks (one on each side of the DUT). (2) Calibration-based de-embedding (TRL): the TRL calibration moves the measurement reference plane to the DUT interface. The fixture effects are automatically removed during calibration. This is more accurate than physical de-embedding. (3) Time-domain gating: use the VNA time-domain mode to "gate" out the fixture reflections. Only the DUT response within the time window is retained. Less accurate than TRL but requires no calibration standards in the fixture.

What about automated calibration?

Electronic calibration (ECal): an automated calibration module (e.g., Keysight N4691-60004) that contains multiple impedance states internally. The ECal module is connected to the VNA ports, and the VNA automatically steps through the states and computes the error correction. Advantages: faster than manual SOLT (30 seconds vs 5 minutes), more repeatable (no human error in connecting standards), and less wear on the fixture connectors (one connection vs four for SOLT). Cost: $3,000-15,000 per ECal module (frequency-dependent). For production: ECal is strongly recommended. The time savings and improved repeatability pay for the ECal module within weeks of production use.

Need expert RF components?

Request a Quote

RF Essentials supplies precision components for noise-critical, high-linearity, and impedance-matched systems.

Get in Touch