for our research team we need to have the opportunity to calibrate data offline, i.e. we want to read the error terms after a SOLT calibration off the E8364C PNA and calculate the correction by our own.
So far, we calibrate the PNA using a full 2-port SOLT calibration (with 85052B cal kit) and read the 12 error terms with SENS:CORR:CSET:DATA. Afterwards, we measure a DUT (e.g. a low-pass filter) and switch the correction of the PNA off. This data is then used as raw data. We apply Rytting's equations on these an compare the result with the curves measured by the VNA if the correction is switched on. However, the curves differ a lot (see attached plot "offline_calibr.pdf").
Please note that the sharp peak at about 13 GHz for S21 is an artifact and not of interest. The PNA did not send data for this frequency point, so we added it arbitrarily.
The main questions are:
1./ Is the process we follow correct?
2./ Does the E8364C PNA use the 12-term error model?
3./ Are the equations provided by Rytting correct?
Hopefully, there is someone willing to help.