Precision Power Level Accuracy Measurements Reduce Risk for Wireless Receiver Manufacturer

Company

Designing reliable receivers for military applications demands precise and accurate measurements. When designing a new receiver for the aerospace and defense market, a leading manufacturer needed to ensure its products could capture the intended signal in a crowded RF environment, even when operating at the maximum distance. To ensure performance, the company tests receivers during production to verify receiver sensitivity meets design parameters. The device must receive the signal and successfully demodulate it to pass the test.

Key Issues

The company measures receiver sensitivity using a signal generator. The signal generator supplies an input signal to the wireless receiver under test. A test engineer decreases the signal generator power level until the wireless receiver can no longer demodulate the signal. The results are recorded as a power level accuracy measurement. Power level accuracy determines receiver sensitivity. During a power level accuracy measurement, there are four possible test outcomes, as shown in Figure 1: true pass, true fail, false fail, and false pass. True Pass and True Fail are expected outcomes. However, two additional outcomes are possible when measurement uncertainty is considered. False fail - when a perfectly good product fails the test, and false pass — when a defective product passes the test.

Solution

Keysight evaluated the manufacturer’s test procedure, and determined that the manufacturer’s calibration procedure was contributing unnecessary measurement uncertainty (MU) in their test process. From the customer’s calibration report, it could be seen that the manufacturer was using a state-of-the-art signal generator, but their calibration service provider was calibrating it using an older portable spectrum analyzer. The older portable spectrum analyzer, a Keysight 8563E (9 kHz to 26.5 GHz) had larger MU or less accurate specifications than a modern Keysight E4440A PSA (3 Hz to 26.5 GHz). The differences resulted in substantially more measurement uncertainty, increasing the risk of false pass and false fail measurements. The older unit used by the customer’s service provider for calibration had significantly higher MU at < ± 1.5 dB, versus < ± 0.25 dB for the new instrument. The striking difference is driven by analog components inside the older spectrum analyzer, which contribute to the much higher MU in the older instrument as shown in Figure 2.

The ROI on Calibration

The customer understood the additional risk of calibration, but needed a business case to help justify the investment in calibration services. A mathematical equation defines these test outcomes for calculating risk as established by the Joint Committee for Guides in Metrology (JCGM), a worldrecognized organization formed by seven international standards organizations in JCGM 106:2012 per section 9.5, Calculation of Global Risks.1 Keysight calculated risk by modeling an industry-based scenario that resonated with the customer per JCGM 106:2012. The highest risk of missing production volume targets comes from false passes, and the most significant costs are due to good filters mistakenly failed during testing. The risk assessment determined that a manufacturer producing 100,000 units per year could reduce its volume of false passes from 500 per year to under 10 per year if a more accurate calibration is performed. Improving power level measurement accuracy, a critical measurement parameter for receivers, can reduce test result risk by 50 times. And reducing false fails and false passes could result in $1.5M annual savings (see Figure 4).