开云体育

When you click on the TDR in Nanovsaver. The length of the cable is already indicated, which I presume is the physical length once you specify the coax type or VF. There are two plots on the graph, but what do they show?

On Mon, Nov 14, 2022 at 12:53 AM, tuckvk3cca wrote:

When you click on the TDR in Nanovsaver. The length of the cable is already
indicated, which I presume is the physical length once you specify the coax
type or VF. There are two plots on the graph, but what do they show?

One graph shows the Time Domain response of the cable attached to CH0 (port1). The other shows the approximate impedance magnitude |Z|. In the annotated Saver graph below I measured 3M of RG-316 connected to 1.8M of 75 ohm RG-6 with an adapter.

Roger

Great, thanks. So you get the same results if the end is either open or shorted right? Reverse the cable and you can confirm where the break is.

With a dummy load the cable length should read very large with no kinks right?

How do I adapt this to other cables? Do I need a balun?

Yes, open or shorted gives equivalent measurements (one is phase-reversed
from the other).
And terminated in the characteristic impedance of the cable will give an
'infinite' measurement (no reflection).

How to adapt to other cables? You can do other cables the same way, even if
they are balanced or not of 50-ohm impedance. You do need to know the
velocity factor.

And you also need to know that the measurement is not exact: you will only
get within several percent of the cable length at best. This is due to the
way the nanovna (or host software) does the TDR measurement. The nanovna
measures at discrete frequencies using its frequency scan, then the
software takes those points and calculates an FFT to change to the time
domain. This inherently uses a 'bin' size, which limits the granularity of
the measurement. You can see this granularity as you step across the TDR
graph - each calculated point is separated by this minimum granularity.

If your vna firmware has the "Measure / Cable" option in the menu, it is
better to use if you want to find the length of the cable (or the distance
to a break, if you have seen one in the TDR). It still is only as accurate
as your estimate of the cable velocity factor, but is accurately measured
(it uses a different technique without any binning effect/uncertainty).
The Measure Cable function will also show the characteristic impedance of
the cable, and the loss in dB along the length of the cable.

On 11/15/22 12:06 PM, Stan Dye wrote:

And you also need to know that the measurement is not exact: you will only
get within several percent of the cable length at best. This is due to the
way the nanovna (or host software) does the TDR measurement. The nanovna
measures at discrete frequencies using its frequency scan, then the
software takes those points and calculates an FFT to change to the time
domain. This inherently uses a 'bin' size, which limits the granularity of
the measurement. You can see this granularity as you step across the TDR
graph - each calculated point is separated by this minimum granularity.

Rather than TDR - (where the resolution is c/(frequency step size))

I'd suggest just measuring phase at your frequency of interest - You can directly display phase vs frequency, set a marker or two, and go to town.
In any of the PC tools, you can put a reference trace up of one of the cables as you measure and cut the other.

Then you don't have any issues with interpolation between bins in the FFT.