EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: electrolust on August 19, 2020, 03:54:20 am
-
Is there any device other than FieldFox that can do a one port cable loss measurement, without requiring access to the other end of the cable? It doesn't have to be a portable device.
-
Yes, any VNA or scalar network analyser will do it - a scalar NA is just a spectrum analyser with a tracking generator and return loss bridge.
It should be possible with an oscilloscope and a pulse generator too - aka TDR.
Cable loss is frequency dependent so you need the instrument to be capable of the frequency range of interest obviously.
The loss you measure will be twice the actual loss of the cable - assuming the remote end of the cable is open or short circuit.
However, if the other end of the cable is already connected to something, and you cannot independently measure that something, or you really cannot access it to disconnect that something, then I don't think it will be possible.
NanoVNA is likely to be the cheapest way, they go up to 3GHz in the V2 for under $100.
-
Yes, my understanding is that you have to calculate it from S11. I'm not crystal clear on the calculation yet, as I've seen 2 different versions.
I saw a nanoVNA video on youtube that explains a method that requires measuring both open and shorted, then calculating the cable loss. It's based on a QEX paper by AI1H, and specific for the nanoVNA type of analyzer. Whereas the fieldfox does not require 2 measurements, and performs its calculation as part of its "app". From Keysight:
"Once the reflected measurement is complete, FieldFox, using a built-in model for coaxial cable dispersion
and knowing that the measured signal contains twice the cable insertion loss (due to the round trip path), can now report the cable insertion loss as a function of frequency."
I found references to "coaxial cable dispersion model" which basically describe how to model the dispersion effects vs frequency, important if you want to simulate the cable. I'm not sure why you'd need to apply such a model when you are measuring the real-world effect of it. ie, the RL measurement already includes the dispersive effect.
I have a nanoVNA on the way for a few weeks now -- supposed to be here any day. It's very challenging to reach even one end of the installed [set of] cables I want to measure. I was hoping to find a way to do it with a single measurement like the fieldfox. When I get the nanoVNA I'll see how the 1 port vs 2 port measurements differ on a sample cable and maybe I can just wing it.
Also I was just curious what other devices out there might have similar cable measurement convenience to the fieldfox.
-
Deepace KC901V (i guess the C, S and M version got too) VNA has the DTF function built in the FW... for Nano VNA, not sure about the latest FW but from this link, you have to do it manually or using external script... https://nuclearrambo.com/wordpress/accurately-measuring-cable-length-with-nanovna/ maybe you can ask OwO engineering (a member here who got involved in Nano VNA development) to add the DTF/FDR functionality into the FW if it hasnt been added already. ymmv.
-
Yes, my understanding is that you have to calculate it from S11. I'm not crystal clear on the calculation yet, as I've seen 2 different versions.
I saw a nanoVNA video on youtube that explains a method that requires measuring both open and shorted, then calculating the cable loss. It's based on a QEX paper by AI1H, and specific for the nanoVNA type of analyzer. Whereas the fieldfox does not require 2 measurements, and performs its calculation as part of its "app". From Keysight:
"Once the reflected measurement is complete, FieldFox, using a built-in model for coaxial cable dispersion
and knowing that the measured signal contains twice the cable insertion loss (due to the round trip path), can now report the cable insertion loss as a function of frequency."
I found references to "coaxial cable dispersion model" which basically describe how to model the dispersion effects vs frequency, important if you want to simulate the cable. I'm not sure why you'd need to apply such a model when you are measuring the real-world effect of it. ie, the RL measurement already includes the dispersive effect.
I have a nanoVNA on the way for a few weeks now -- supposed to be here any day. It's very challenging to reach even one end of the installed [set of] cables I want to measure. I was hoping to find a way to do it with a single measurement like the fieldfox. When I get the nanoVNA I'll see how the 1 port vs 2 port measurements differ on a sample cable and maybe I can just wing it.
Also I was just curious what other devices out there might have similar cable measurement convenience to the fieldfox.
Ah nice, you have something more advanced in mind. Sorry I assumed it was just a basic measurement you were after.
Is this the paper?
http://www.arrl.org/files/file/Technology/tis/info/pdf/0505qex044.pdf (http://www.arrl.org/files/file/Technology/tis/info/pdf/0505qex044.pdf)
I suspect that with typical ham frequencies, cable lengths and losses (i.e. HF, long and high) there won't be a very big difference between return loss, open/short, open/short + attenuator, and the 25ohm/100ohm approach. The loss in the cable itself will bring the impedance back closer to 50 ohms.
I found this article with some details on the Fieldfox methods and their uncertainties, but it doesn't mention the coaxial cable dispersion model which I see in the marketing blurb.
https://www.keysight.com/upload/cmc_upload/All/GNO2_ERTA_FINAL.pdf (https://www.keysight.com/upload/cmc_upload/All/GNO2_ERTA_FINAL.pdf)