#565 NANOVNA Coax Velocity Factor and Coax Length

Episode 565
Since shooting this video, I've seen some RG8x at 66%VF but nothing as low as 64%
at 3:40 I say 180deg is 1/4. that is 180deg in the smith chart. it is a 90 deg phase change in the wave. once around the smith chart is 180 deg (1/2 wave) and a full wave (360 degrees) travels around the smith chart twice.
Be a Patron:

[warplanner]

Mr IMSAI person, thank you for your video. This shocking revelation may only amuse smug observers with a big "so what?" but I can assure you that folks like myself who are putting together phased vertical array ham antenna systems and calculating the coax lengths for the feed and delay lines find this highly interesting. The so-called Christman calculator (elsewhere on the Interwebs) relies on VF as one of the inputs. A variance of this size will ruin the pattern of the antenna!
Thanks muchly yet again yet again!

[ohaya]

Great video and thank you for clarifying in the description the 180 degrees as 1/4 wavelength in the smith chart as I was initially confused as to why you said it

[kium]

I have DX Engineers RG8X. I measured it with my NanoVNA, it calculates to 79.8 VF. Per DX Engineers data on this coax, they claim 82 VF. I have 50 feet of it using it for y VHF\UHF antenna. Seems to be pretty good quality imo. 73, and tnx for the videos .

[hammer]

Some people blame the dog, other people blame the chair ;)

[lmamakos]

It's very reassuring that the numbers all work out to the same result. Physics continue to work. Why the slow velocity factor? You might need to cut a connector off the end to find out.

[RC-Heli]

How would you use the Nano to cut a perfect half wave or multiples of half wave length coax?
Thanks for sharing!

[filipinodiynot]

just wonder how did you compute the 51.1n/s- thanks

[OwenDuffy]

You do not disclose details of the test fixtures, and they are relevant. That said, it does appear that you have a sample of line that uses solid PE (not PP as you mention) dielectric and if good quality, VF at higher frequencies is probably very close to 0.67, but not at 3MHz as measured.

A lot of transmission line theory is taught based on distortionless lines, and practical lines are not usually distortionless.

Zo and VF are often assumed to be independent of frequency, but that is not true and testing practical coax cables at 3MHz as you have done will be subject to error under those assumptions.

Your analysis of the time domain test is too simple to reveal that vf=f(freq). The line is not distortionless, the returned step will be a different shape to the incident step and your measurement does not take that into account.

Your CW test is more interesting, but while the phase of s11=180° (wrt Zo=50+j0 ohms) at a quarter wavelength of OC Zo=50+j0 ohm line, that is not what you measured, Zo is unlikely to be 50+j0 ohms.

Essentially the VNA measures incident and reflected voltages and calculates s11 wrt 50+j0 ohms and displays that, but it is exactly not applicable to your cable which almost certainly has a different Zo... particularly a significant Xo component at that frequency which gives rise to the measurement problem.


A test @ say 50MHz using a good fixture should indicate VF closer to 0.67 as Zo is closer 50+j0 (the calibration Z of your VNA), not perfect, but closer.

[bill-]

I can't see how you measured coax length with the VNA. I watched it several times.
You physically measured it as 64" which converts to 15.34m. I don't see that anywhere calculated or shown on the VNA then you jump to using it for finding velocity factor. Am I missing something?
The accuracy of decimal points and nanoseconds then relying on a tape measure?

[kylefreemason]

T=d/s. Yet if I divide 15.34 by 300 million, I don’t get 51.1. Where did I mess up?

[terry]

Did you also measure the cable's impedance and loss? I would guess they are way out of spec as well.

[TheRealBanana]

Hi Im currently trying to make a quarter wavelength stub with cable that has a velocity factor thats out of spec. The cable Im using only has a connector on 1 end while I snip the other to length, so I can only do S11 measurements. Dumb question but If I do those S11 delay measurements, do I need to divide the delay by 2 since its measuring the length twice (there and back)? Also when calculating the time of one wavelength and comparing to measured values, do we need to consider velocity factor? Thanks for the video, its been extremely helpful.

[ianharvey]

OK a little help here please, I have just over a meter of cable connected to the nvna with pl259 and so239 to sma adapter and not terminated on the other end,
Phase peak to peak (full 360 degrees on Smith chart is 123mhz, devide that by 2 to get 61.5mhz (for 180degree on Smith chart)
1 devide 61.5 =0.01626 this is rounded to 16.3ns then devided that by 4 is 4.075 velocity factor?
Can someone help here I either have a really high shielded coax with the worst ever velocity factor or the math is wrong somewhere?
Does the cable need to be long?

[DavidLopez-bzrj]

Sorry, but I dont get the last step, where you divide 1/freq and you get nanoseconds.
Also, the floating coax is open at the end or shorted? Thanks in advance, good job.

[paulcohen]

Now It's the time to dissect the cable and examine the dielectric and conductors.

[zeljkoadzic]

For my point of view and experience ... look for me like rg58 sorry for word... peace of shit ... the worst coax i ever use in my life!

[joeschroedl]

When you introduce the nanoVNA the first measurement you say is from 0 degrees to 180 degrees. But, on the nanoVNA screen, it's going from -180 to 180 degrees. Wouldn't this be a change of 360 degrees and would be the same as going all the way around the smith chart?