I've designed a matching network for a PCB Loop Antenna and I've tried to match it to a source of 465 Ohm impedance with 1pF parasitic capacitance i.e.
Source Impedance = (166 -j223) Ohm
Ideal Load Impedance = (166 + j223) Ohm.
I've used Tapped Capacitor Matching network as suggested in this Atmel AppNote.

I want to tune the antenna using a VNA. I'm confused if the tuning process is the same as that for tuning a match to 50 Ohm. Usually the return loss is measured using a 50 Ohm input VNA and 50 Ohm matched antenna. Can the return loss be interpreted in the same way for non-50 Ohm matched antenna?

 

I have a non-50 Ohm loads what must be passed anyway. If to use VNA with brute force, it will be overinhabited with hundreds of resonances what in reality are reflections of all possible corners of transmission line. I have tried to use hardly short lines like 1 mm long etc (less than lambda/40) but can say the problem may be perfectly solved that way. But can applying the one pi filter to pass the signal source with 50 Ohm line and device, and other filter to adjust this 50 Ohm line with real load.

 

The adapter board properly connected to a T fitting so you are at a safe distance away depending on the output level.
https://www.nonstopsystems.com/radio/frank_radio_coax-sw.htm

 

When working in a none 50 ohm system but using a 50 ohm VNA you are no longer aiming for maximum return loss at 50 ohms. Instead you are trying to get as close as possible to the required impedance point on the smith chart. A way to quantify the quality of the match is to load measured S-parameters and then change the measurement port impedance to that required an simulate return loss.
Best regards from Chris.

 

When working in a none 50 ohm system but using a 50 ohm VNA you are no longer aiming for maximum return loss at 50 ohms. Instead you are trying to get as close as possible to the required impedance point on the smith chart. A way to quantify the quality of the match is to load measured S-parameters and then change the measurement port impedance to that required an simulate return loss.
Best regards from Chris.

Hello Chris,
I think this is the way to go. Thank you for your input.