Answer
The most accurate method is the phase shift method.
We count the phase shifts and show these counts over the frequency using the unwrapped phase method.
Based on the counts of phase shifts and the frequency, we can calculate the length of the cables.
To get the right value for the phase shifts, it is very important that the phase shift between two adjacent
frequency points does not exceed 360°.
For example, if the phase shift difference between two frequency points is 450°, the VNA will not take it as
450° but as 90° (450°-360°). Therefore, the results will be wrong.
Preconditions - you have to insert your specific data :
Cable length 120 m
Velocity factor 0.69
Stop frequency 1 GHz
=> Electrical cable length 120 m / 0.69 = 174 m
Delay time in an S21 measurement 174 m / 3e8 m/s = 580 ns
! Now the important calculation for not getting phase shift differences of more than 360°!
=> Frequency step 1 / 580 ns = 1.7 MHz – we use 1 MHz
Sweep points 1 GHz / 1 MHz = 1000
Settings:
- Preset
- Stop frequency 1 GHz
- Channel - Sweep - Number of Points 1000
- Trace - Format - Unwrapped Phase
- Trace - Scale - Autoscale
- Trace - Trace - Trace Statistics - Phase Delay/El. Length
Make a full two-port calibration with both ports. To be able to do this, you have to insert an additional
short cable in your test setup.
Then connect the reference cable and select
- Trace - Trace - Data -> Mem
- Trace - Trace - Math=Data/Mem
Afterwards you can connect the next cable and you will see the difference in electrical length
relative to the reference cable.
For more information, see also the following ZVR application note: