Phase Measurements 1

[feetoes]

My company has an 8720A network analyzer. I have not personally used this piece of equipment for 15 years. My question is after calibrating for VSWR, Transmission, and Isolation, how do I measure the phase of a cable assembly?
Once phase is measured, I want to make that my standard and check other Cable assemblies of the same length against it. This would be for flexible cable assemblies.
Frequency would be from 100MHz to 11 GHz.
I have tried to interpret the HP manual, but their manuals never have explained anything to well.

 

[Higgler]

Just connect your gold standard between 8720A ports one and two and do a
Calibrate
Response cal (which is S21 from port 1 to port 2)

The display shows a flat line for amplitude. One of the display options is for Phase (in stead of amplitude), so display phase and thats a flat line too (since you have just calibrated). It's good at this point to wiggle your cable a bit, to see just how much phase change versus wiggling the cable you will get. Most good cables at 18 ghz have only 3 degrees phase change with slight wiggling. This is an important step, cuz if your gold cable is phase sensitive, everything else becomes a nightmare.

Remove this cable and put your next cable on, if they are perfectly matched, you will get a flat line again (or nearly flat line).

This should be good enough for 99% measurements, but if you need one more improvement in accuracy, then you need to do a full two port calibration. I can explain that if needed.

Sometimes "gold standards" need their absolute phase measured. For this, if you have time domain on your analyzer, you need to switch to time mode and measure the rf travel length and record that. Again, you may not need absolute phase length. You should have two or three gold standards if you plan to have a large production run.

kch
8720A's are really old, hope it still works ok.

 

[MKimagin]

If you can create golden sample with time domain , why just not measure that cable assembly in time domain?
Or measure return phase with open port (need cal kit Open on that port) and divided by 2.
Another way is the following.
Let say we have we have cable assembly with SMA M-M connectors. That mean that we need to used adapter exchange method (the only way available on 8720).
So the first step is to connect one good cable to port 1
and connect adapter (from the calkit will be good) 3.5mm F-F , the port 2 of the network analyzer need probably 3.5mm or SMA F-F adapter. Perform full 2 port cal. Then exchange the 3.5mm F-F adapter with 3.5mm F-M adapter from the same cal kit (important to have 2 similar length adapter with similar electrical characteristics). That in most cases should be sufficient to perform the phase measurement on your cable assembly.
The cable connected to the network analyzer during calibration should be in similar position as it will be used during measurement of your cable assembly. That might be impossible to maintain and addition port extension can be perform on port 1 with the cable properly position.
To do port extension you need to set the meas to S11 set the formant to phase set the marker to last frequency point, connect open from the cal kit (connect to that exchanged adapter), go to adapter exchange function , enable it, press port 1 and rotate the knob utile the marker show zero degree.
That should minimized the phase error.
I never used the 8720A, the oldies from that series that I had in my hand was 8720C. I am not sure if the port extension is on the A version.