I was scanning that rfcafe.com list to see if there were any anomalies (remember my suspicions about the 75 ohm cables with higher than the 52pF/m I measured on old style semi-air spaced UHF co-ax and the more typical 55pF/m for its modern day replacement, CT100 PE foam dielectric?) and spotted a value of 12pF/Ft (40pF/m) for the RG223 cable.
Since this cable is only slightly thicker than RG58 cable, allowing it to be confused with other similar gauge 50 ohm cables and, if the pF/Ft were actually true, a simple capacitance test would lead to its rejection by reason of it measuring as a 93 or 95 ohm cable, I decided to search for datasheets for RG223 cable, landing up on Pasternack's site to download this datasheet from here: -
https://www.pasternack.com/images/ProductPDF/RG223-U.pdf
(this was before I spotted the RG223 in your list of 50 ohm coax cables with typical 101pF/m figures).
I repeated my search and downloaded this pdf from a link on the Farnell web site:-
http://www.farnell.com/datasheets/2921248.pdf
All of which confirmed my suspicions that not all of those entries in that rfcafe.com list can be trusted. It's a handy list but you need to check out any entries that show unusual values for the pF/Ft figure which is what the 'sanity check' capacitance measurement is relying upon as a means of estimating the most likely Zo value for an 'unknown cable'.
I suspect this was most likely simply a transcription error made when that list had first been compiled. It's quite possible there may be some more such errors giving the impression that some of those cable have somehow stepped outside of the laws of physics (as we understand them 
).
Even datasheets can contain (not too few) errors, all the more so webpages that collect a vast amount of data in one big table…
And then there is the sad fact that data sheets are often incomplete. For example, I could not find a specification for the capacitance of the very popular CS29 cable – otherwise I would have included it in my table.
Anyway, even with correct data it’s not straight forward if you have cables from a dubious source. Take my standard lab cable for example, HyperFlex 5 (from a reputable source of course). Despite the huge difference in quality, one might confuse it with RG59 because of the similar diameter, same color – even the minimal bending radius is the same, although the HyperFlex 5 actually feels notably stiffer.
A capacitance measurement has to be fairly accurate to clearly distinguish 67.6 pF/m from 74 pF/m. This should not be a problem as long as you have, say, 10 meters of each cable available. But in practice your piece of cable is less than 2 meters, sometimes even only 25 cm. In this case, stray capacitances might become a major problem - and you have plugs on either end of the cable, which also add some capacitance (that would be roughly the same for any cable), thus additionally swamping the difference in measurement.
In case of BNC cables (and there will hardly be any other connectors used in a lab and for video alike), we can have a look at the pin of the plug. It is thinner for the 75 ohms connector. That was the final evidence for me that the “RG316” was actually some 75 ohms cable.
One mystery remains: my crappy RG58 cables. I bought them as “new” (and they looked like new indeed), sealed in plastic bags, from a professional German eBay seller some ten years ago. If I look at the jacket, I can see “RG58 COAXIAL CABLE 50OHM” printed on it. The plug has the thicker pin (although totally corroded!), so it has to be 50 ohms indeed. Yet this cable performs at least as bad as the 75 ohms cable did.
Of course I have good quality RG58 too. Just because it was for a different location, I didn’t want to take everything away from my lab but rather thought why not finally put these brand new cables out of this old, untouched purchase to good use at home?
For more serious (and precise) measurements, we don’t want to use standard cables with their high insertion loss anyway. We want low loss and high shielding, which neither RG58 nor RG316 can provide.
In the light of this, it was a rather thoughtless action to include a (supposed) 1.2 mtr RG316 coax connection as reference for the probe measurement. Apart from the ripple, the difference in insertion loss should be obvious.
Here’s the insertion loss at 600 MHz and shielding at a certain frequency for some popular cables:
HyperFlex 5: ~0.20 dB/m, >105 dB @ 100-2000 MHz
RG400: 0.36 dB/m, >81 dB up to 6000 MHz
RG223: 0.43 dB/m, >83 dB up to 1000 MHz
RG58: 0.50 dB/m, >38 dB up to 1000 MHz
RG316: 0.60 dB/m, >38 dB up to 1000 MHz
Home
Help
Search
Login
Register
