by Lichun Wang & Yifan Zhang, Keysight Technologies
In time response tests involving “non-ideal” pulse-modulated signals for wideband power sensor characterization, result correction becomes essential. Traditionally, the Root Sum Square (RSS) calculation is employed for this purpose. However, due to inherent limitations in the RSS method, deconvolution emerges as an alternative for result correction. This paper outlines a fundamental process for obtaining a power sensor’s time response using deconvolution. Additionally, we delve into two critical aspects of the deconvolution process: enveloping trace extraction from the incident pulse-modulated signal and effective result denoising.
