I am using a NanoVNA and testing a SMA (Connected to an external GPS) to IPX / U.FL (connected to PCB, PCB has male) wire. I have done proper calibration of NanoVNA (short, open, load, through) and measuring S11 and S21 parameters on NanoVNA. I got little surprised that at higher frequency my S21 Gain is positive. Why is it ?

My Setup is:
Nano VNA Port 1 ---> SMA Male to Male Wire (to make DUT connector compatible) ---> SMA to IPX / U.FL (DUT) ---> IPX/ U.FL (to make DUT connector compatible) ---> Nano VNA Port 2
All are 50 Ohms.

 

What would the physical connection need to measure S21 of a piece of wire look like?

Thanks for your reply, This is my setup.

 

This is likely due to port mismatches on the Rx and/or Tx ports. NanoVNA doesn't even specify the port return losses as far as I can see. What is most probably happening is that the port impedances vary from 50 ohms and you are accidentally providing a better match to the port impedance at the higher frequencies. The calibration routine probably assumes the ports are matched.

Put 10dB pads on both the Tx and Rx ports and recalibrate. You'll lose 20dB of dynamic range, but you will more accurately measure the S21 of whatever you place between the Tx and Rx ( for low loss S21).

 

This is likely due to port mismatches on the Rx and/or Tx ports. NanoVNA doesn't even specify the port return losses as far as I can see. What is most probably happening is that the port impedances vary from 50 ohms and you are accidentally providing a better match to the port impedance at the higher frequencies. The calibration routine probably assumes the ports are matched.

Put 10dB pads on both the Tx and Rx ports and recalibrate. You'll lose 20dB of dynamic range, but you will more accurately measure the S21 of whatever you place between the Tx and Rx ( for low loss S21).

Thank you for your input.