Low cost test and tuning of RF circuits and systems

Abstract

The demand for multi-faceted wireless applications has led to increased levels of device integration and migration of RF to CMOS technologies. However, the resulting circuits are increasingly susceptible to manufacturing process variations, coupled noise (substrate, power planes) from on-chip digital signal processing circuitry, thermal fluctuations (V t sensitivity to temperature), resulting mismatch effects and device wear-out (V t degradation) phenomena. In the recent past, “alternative” testing methods have been used to perform parametric testing of RF transceivers. These allow simple tests to measure complex RF specifications while at the same time allowing catastrophic failures to be detected. All (or most) of the device under test (DUT) specifications are evaluated from a single test application which may be repeated multiple times to average out noise effects. As an example, the EVM specification for a transmitter or receiver can be predicted accurately from a carefully constructed multi-tone signal with specified magnitude and phase for each tone. The process of using such techniques requires the use of supervised learning algorithms that are necessary for determining the test specification values and associated pass/fail conditions from the alternative test measurements. These learning techniques also account for nonidealities in the test stimulus generation and data acquisition hardware. Once the infrastructure for such algorithms is put in place, production testing is greatly simplified and allows significant trade-offs between the complexity of test instrumentation (stimulus generation, response analysis), the complexity of the test procedures and the complexity of the back-end signal processing algorithms needed for specification prediction and pass/fail analysis. As another example of the benefits of this approach, simple high frequency digital pulse streams from a logic circuit appropriately filtered (shaped) can be used as stimulus to RF devices and the response, down converted by an envelope detector, can be used to accurately predict the gain and nonlinearity specifications of the RF DUT.