Phase Measurement

[Squirtle]

Hi,

I have a question about phase measurement of an rf strip with One Input Cable + Two Port Power Divider + Two Output Cables. I use Keysight PNA NA N5224B. So, i have two different methods to measure the phase difference between two output cables;

1-Calibrate VNA using a small cable (Through calibration). Then plug input cable to this cable via fixture. Terminate one of the output ports with 50Ohm. Measure the phase of the remaining output port. Then apply same process for other output port. Finally we got two phase measurement and when we substract these two measurements we shall have phase difference between two output ports. This method is implemented using Port-1 and Port-2 of the VNA.

2-In this approach i used ReceiveA and ReceiveB ports of the VNA. I connected the input cable to source of the VNA which is Port-1, then output cables are connected to receive ports of the VNA. From measurement menu, i switched the response to A/B which is difference (amplitude and phase) between two output cables.

However, although these two methods seem perfectly fine to me, I could not get same result using these two methods. First approach gave to me 105-degree phase difference and second approach 45-degree. So my question is what might be the problem? I must have obtained same results using these two methods.

 

[volker@muehlhaus]

What do you get for setup 2 if you have the exact same thru line connected to channel A, then to channel B? Do they show the same phase?

 

[D.A.(Tony)Stewart]

For your N5242A-400, the best cal method is ;

• The multiport SOLT (Short, Open, Load, Through) calibration, aka TOSM (Through, Open, Short, Match)
• Use 3 ports with semi-rigid cable only.
• Torque all connectors with calibrated tool.
• set IF BW to mid-scale like 1 kHz
• Display S31,S21 then S31-S21 and check S32 isolation.
• Swap ports and compare.
• Expect 1 deg variance unless ports are 180 difference

 

[Squirtle]

What do you get for setup 2 if you have the exact same thru line connected to channel A, then to channel B? Do they show the same phase?

No, I got 45 degree phase difference between two output ports, btw the receiveA and receiveB channels are connected to output cables of the power divider simultaneously.

--- Updated Mar 21, 2025 ---

For your N5242A-400, the best cal method is ;

• The multiport SOLT (Short, Open, Load, Through) calibration, aka TOSM (Through, Open, Short, Match)
• Use 3 ports with semi-rigid cable only.
• Torque all connectors with calibrated tool.
• set IF BW to mid-scale like 1 kHz
• Display S31,S21 then S31-S21 and check S32 isolation.
• Swap ports and compare.
• Expect 1 deg variance unless ports are 180 difference

I do not think problem is about calibration because i did these tests couple of times and there was no big change. Olur vna is two port vna.

 

[volker@muehlhaus]

No, I got 45 degree phase difference between two output ports, btw the receiveA and receiveB channels are connected to output cables of the power divider simultaneously.

You did not understand what I wrote.
I suggested to do a thru-line test, and see if you get the same phase for receiver A and B if you change the input port.

 

[Squirtle]

You did not understand what I wrote.
I suggested to do a thru-line test, and see if you get the same phase for receiver A and B if you change the input port.

so, you are suggesting to connect each output cables to receive A (without involving receive B) and measuring phase difference then doing same process for receive B and comparing the phase difference between receive A and receive B. Right?

 

[volker@muehlhaus]

No, I suggested to leave out your power divider and use a simple thru line, to see if you get the same phase readout no matter which receiver input you use.
Just to make sure they are in sync for phase difference measurement later (with your divider).

 

[Squirtle]

One more question? if there are phase difference between receive A and receive B, is it possible to calibrate them?

 

[D.A.(Tony)Stewart]

No, I got 45 degree phase difference between two output ports, btw the receiveA and receiveB channels are connected to output cables of the power divider simultaneously.

--- Updated Mar 21, 2025 ---

I do not think problem is about calibration because i did these tests couple of times and there was no big change. Olur vna is two port vna.

Did you do all 4 calibrations SOLT? Which port did you calibrate? The splitter or ? Cable A out?

If you do all cal's for both Cable A & B, you can store one or both in memory and do what Volker requested too.