De-embedding
- Updated2023-12-27
- 2 minute(s) read
De-embedding
NI-RFPM always performs calibrations at the blind mate connectors of the STS. The DUT is not directly connected to the blind mate connectors, and the signals pass through other hardware interfaces, such as the load board, that connect the DUT to the blind mate connectors. Therefore, to generate and measure accurate signals at the DUT, the hardware interfaces must be de-embedded. To de-embed the hardware interface you need to characterize the S-parameters of the paths that go from the DUT to the blind mate connectors. These paths are considered as two ports. So, de-embedding is achieved by providing NI-RFPM with an S-parameter (S2P) file which relates the reference plane from the blind mate level to the DUT level for each port in the system.
You can use the NI-RFPM Configure Deembedding VI or the niRFPM_ConfigureDeembedding function to load the S-parameter file in NI-RFPM for each measurement port.
The quality of your measurement is influenced by both the accuracy of the STS as defined in the specifications–which are the results of a custom calibration–and the accuracy of the data in the de-embedding file. The accuracy of the de-embedding file also influences the relevance of the error coefficients as the requirements at the DUT are transformed by NI-RFPM internally to the requirements at the blind mate. NI-RFPM assumes that the S-parameters as a function of the frequencies in the file are accurate. As such, you can get the best accuracy when the frequencies used to calibrate match the frequencies in the S2P file. If the frequencies do not match, the system performs a third order cubic spline interpolation on both the real and imaginary part, which decreases the accuracy.