Automated and Adaptive RF Effects Testing

Testing electronics for vulnerability to radio frequency (RF) radiation is time-consuming, due to the large number of source variables of interest. One typically searches for the minimum electric field that causes upset, as a function of center frequency, pulse width, pulse repetition frequency, number of pulses, and bandwidth. It is impossible to test all combinations of all the variables, so one must intelligently select the source parameters most likely to expose a vulnerability. To select source parameters, we propose using standard techniques from minimization theory. We first select a few frequencies where the test object seems most vulnerable; either where the object is resonant, or at its operating frequency. At each frequency, we iteratively select source parameters, and test to find the E-field upset threshold. We then measure the local gradient of the threshold function and follow a steepest descent path. The source variables will all be controlled electronically. Also, one can automatically determine whether the test object has been upset, and send a reset command. Eventually, this will lead to a completely automated system, which will intelligently select the test parameters, monitor the status of the device, reset it as necessary, and converge on a minimum electric field upset threshold. BENEFIT: This research will lead to an efficient procedure to test components for vulnerability to RF radiation. During Phase I, a partially automated experiment will be set up to test the vulnerability of a mobile phone. If this research is successful, then Phase II would involve the complete automation of the procedure developed in Phase I. It would also involve developing the technique to automate the testing of more complicated devices, which might not be as easily tested as a mobile phone.