Manufacturing and Production Assembly and Test Informational

What is the test time and cost tradeoff between full S-parameter testing and go/no-go production tests?

Full S-parameter testing and go/no-go testing represent two extremes of the production test strategy spectrum, with different trade-offs in accuracy, test time, and cost: (1) Full S-parameter testing: measures the complete 2-port (or multi-port) S-parameters across the full frequency range. Data: S11, S21, S12, S22 at typically 201-1601 frequency points. Test time: 30 seconds to 5 minutes per DUT (depending on: number of frequency points, number of averages, IF bandwidth setting, and number of bias conditions). Information: provides complete characterization of the DUT (gain, return loss, isolation, group delay, stability). Enables: detailed failure analysis (the S-parameter data reveals exactly what is wrong with a failed DUT), statistical process control (SPC, tracking parameter distributions over production lots), and correlation with customer incoming inspection (the customer can reproduce the measurement). Cost per DUT: $0.50-5.00 (equipment amortization + operator time). (2) Go/no-go testing: measures a limited set of parameters at a few fixed frequencies. Example: measure S21 at 3 frequencies (low, mid, high band), S11 at the center frequency, and a power measurement at one frequency. Test time: 1-10 seconds per DUT. The VNA sweeps only a few points (or uses a scalar network analyzer). Information: pass or fail against pre-set limits. Does not reveal the root cause of a failure (only that it failed). Cannot provide detailed SPC data. Cost per DUT: $0.05-0.50. (3) Optimal strategy: for development and qualification: full S-parameter testing (100% data). This establishes the baseline and identifies the critical test frequencies. For production (volumes > 1000): go/no-go testing at the identified critical frequencies. Test the parameters that are most likely to fail (e.g., gain at the band edge, return loss at the center frequency). Add one or two "sentinel" frequencies that are particularly sensitive to fabrication variation. The go/no-go limits are derived from the full S-parameter data: tighten the limits by the measurement uncertainty to create a guardband. (4) Sampling strategy: not every board needs full testing. Statistical sampling: full S-parameter test on 1-5% of production boards (to monitor process stability). Go/no-go test on 100% of boards (to screen for defects). If the sampling data shows drift: increase the full-test percentage or trigger corrective action.
Category: Manufacturing and Production
Updated: April 2026
Product Tie-In: Assembly Materials, Test Equipment

Full Test vs Go/No-Go

The test strategy is ultimately a business decision: balancing the cost of testing against the cost of shipping a defective unit to the customer.

  • 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
  • Interface compatibility: verify impedance, connector type, and mechanical form factor match the system architecture
Common Questions

Frequently Asked Questions

How do I choose the go/no-go test frequencies?

Analyze the full S-parameter data from the qualification phase: (1) Identify the parameters that vary most across production samples (e.g., gain at the upper band edge). (2) Identify the frequencies where the specification margin is smallest (closest to the pass/fail limit). (3) These are your critical test frequencies. Also include: the frequency where gain is highest (to check for oscillation), and the frequency where return loss is worst (most sensitive to fabrication variation). Typically: 3-7 test frequencies capture 90-98% of potential failures.

What VNA settings optimize test speed?

VNA test speed optimization: (1) Reduce the number of frequency points: 201 points is the default; 21-51 points is often sufficient for go/no-go. (2) Increase the IF bandwidth: wider IF BW (30 kHz instead of 1 kHz) reduces the measurement time per point by 30×. Trade-off: wider IF BW increases noise (less averaging). For go/no-go: 10-30 kHz IF BW is usually sufficient. (3) Reduce the number of averages: 1 average (no averaging) is fastest. Acceptable for signals well above the noise floor. (4) Use segmented sweep: define frequency segments with different numbers of points. Dense points at critical frequencies, sparse points elsewhere. Total measurement time for 50-point segmented sweep: 0.5-2 seconds (instrument-dependent).

Should I save the test data?

Yes, always. Even for go/no-go tests: save the measured values (not just pass/fail) for every board. Why: (1) SPC: track the parameter distributions over time. Detect drift before it causes failures (preventive action). (2) Traceability: if a field failure occurs, you can look up the production test data for that specific board (using the serial number). (3) Customer requirements: many customers (especially military and aerospace) require full test data for each delivered unit. Store data in a database with: board serial number, test date/time, all measured values, pass/fail result, and VNA calibration date. Data retention: typically 7-10 years (per customer contract or regulatory requirements).

Need expert RF components?

Request a Quote

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

Get in Touch