开云体育

Hi Norbert
If I understand you correctly you calibrated the Ch0 port directly and then just connected the various cables to Ch0 port and terminated at the far end with a female 50 ohm load.
If so it is only a measurement of how well the various cable can "transport" the 50 ohm load to the Ch0 port and obviously there is various impedance transformation pending the quality and characteristic of the different cables. It has nothing to do with impact on calibration.
If you had calibrated at the end of the various test cables then you should see no difference but for the NanoVNA you will because is has no build in error corrections, it just relies on the performance of the measurement bridge.
The purpose of error correction is to compensate for the imperfection of the output impedance of the Ch0 port and the input impedance of Ch1 port, as you see for you test 5, which is the Ch1 port input impedance transmitted thru the test cable incl impedance transformation caused by that cable.
A golden rule is: Consider always a test cable as part of the equipment - the NanoVNA - when calibration at the end of the test cables and the purpose of the calibration process is to remote the influence of these test cables now being part of the calibration data saved in the NanoVNA/NanoVNA-saver. However for the NanoVNA there is till a small or large impact from the test cables as no internal error 6 term error correction (12 term error correction not possible as requires measurements of all 4 S parameter in one go. So second golden rule use high quality test cable and as short as possible
Sorry .. my contribution tends to be long and is actually missing info on how to calibrate correctly so S11 and S21 phase are in sync and the calibration is done at the SMA male adaptros reference plane and the FF adaptor phase delay also is removed in the calibration process
Kind regards
Kurt

-----Oprindelig meddelelse-----
Fra: [email protected] <[email protected]> P? vegne af Norbert Kohns
Sendt: 23. november 2019 14:46
Til: [email protected]
Emne: [nanovna-users] Coax cable length influence on calibration

Hi all,
today I performed a NanoVNA calibration test to find out if the coax cable length and quality has any negative influence specially on the S11 return loss.

The calibration setup is 10kHz-900MHz at 808 points.
For the test I used high quality coax cable (M17/152) and standard coax cable (RG174U).
The reference test was done directly on the NanoVNA port connectors. For the thru test a 5 cm long semi rigid coax cable was used (picture 1).
Next, a 20 cm long M17/152 coax cable was tested (picture 2).
Then, a 93 cm long M17/152 coax cable was tested (picture 3).
And last, a 70 cm long RG174U coax cable was tested (picture 4).
Picture 5 shows the test result for the RG174U coax cable for S21.

All test conditions have been repeated. The differences between the two tests were negligibly.

One can see, that the coax cable length has some negative impact on S11, specially with the 20 cm long coax cable.
The quality of the coax cable plays obviously a minor roll.

Do the experts among us have an explanation why this is as such?

Best regards
Norbert, DG1KPN