Cheap cables from China


 

I got a closeout deal, five 3-foot RG-58 BNC male to male cables for $4.95 (shipped!). The ebay ad did not say where they were made, but the price should have told me. They have tags with a p/n and "Made in China", of course.

Anyway I figured they'd be OK up to HF, at least, but not high speed. So I hooked up my avalanche generator and GR power divider, and put one of the cables in the path to the faster (S-4) sampling head. Plus a GR-BNC adapter on one end, and an SMA-BNC on the other end. Figured it should look like crap.

Well... there are NO differences in the dual traces, once I adjusted the 3S2 (B Delay knob) for the extra 5 ns or cable! There is a slight difference in transient response but I have not attempted to adjust either head (and the S-4 has a new hybrid diode assembly). That was a real surprise to me. Figured there would be reflections, sags, etc. everywhere. Both at 1 ns/div (photo in Sampling with 3S2 album) and zoomed out (slower sweep speeds). Nothing but the artifacts from my still-in-development pulser.

I figured I couldn't get too hurt for $5...
Thoughts?


cmjones01
 

I've had similar experiences with cheap RF cables. As far as matching and
signal quality goes, they seem fine at least for my uses though I wouldn't
want to use them for any quantitative measurements higher than a couple of
GHz (for the SMB cables).

What I have noticed is that the cheap "RG58" cables often have terrible
screening. This became apparent when I used some to connect AM503 current
probe amplifiers to a scope in an environment where there were high-current
(>200A)transients from switching a large battery. These transients found
their way in to the cheap cables making proper current measurements
impossible (the scope still triggered even with the AM503s powered down!).
The problem went away when I used old scruffy cables with a heavier screen.
Some of this may have been due to a lower earth loop impedance.

Chris

On Sun, 14 Mar 2021, 03:18 Charles, <charlesmorris800@centurytel.net> wrote:

I got a closeout deal, five 3-foot RG-58 BNC male to male cables for $4.95
(shipped!). The ebay ad did not say where they were made, but the price
should have told me. They have tags with a p/n and "Made in China", of
course.

Anyway I figured they'd be OK up to HF, at least, but not high speed. So I
hooked up my avalanche generator and GR power divider, and put one of the
cables in the path to the faster (S-4) sampling head. Plus a GR-BNC adapter
on one end, and an SMA-BNC on the other end. Figured it should look like
crap.

Well... there are NO differences in the dual traces, once I adjusted the
3S2 (B Delay knob) for the extra 5 ns or cable! There is a slight
difference in transient response but I have not attempted to adjust either
head (and the S-4 has a new hybrid diode assembly). That was a real
surprise to me. Figured there would be reflections, sags, etc. everywhere.
Both at 1 ns/div (photo in Sampling with 3S2 album) and zoomed out (slower
sweep speeds). Nothing but the artifacts from my still-in-development
pulser.

I figured I couldn't get too hurt for $5...
Thoughts?






Jean-Paul
 

The work we do involves a sytem of probes, cables, scopes, etc.

The weakest point in the signal chain will determine the BW, risetime, CMRR etc of the observations.

Cables are often assumed to be OK but there are many differences.

We noticed that the BNC cables poorly made in China easily fail, bad connections, improper crimps, poor quality connectors, etc.

I bought a bunch of original Tektronix USA made calibrated BNC><BNC 0.5M cables. 012-0480-00 from an Ex TEK man.

These are the very best. They have served me for decades.

Kind Regards,

Jon


 

Thanks for the info. I'd forgotten about the shield coverage! Can anyone help me devise a simple test that doesn't require switching 200 amps? :) I can generate 500 watts or more of RF at 50 ohms with my ham gear, but only up to 30 MHz.

These cables actually do look reasonably well made, but of course I would not expect them to hold up to daily continuous use like Tek or GR cables. For 99 cents each, they'll probably be fine for occasional tinkering.


Mark Goldberg
 

I've tended to use double or triple shielded cable or hardline for
interconnects. I happened on some really good ones. I suppose you
could transmit into a dummy load and see what leaks out the cable, but
it depends on the leakage in the dummy load and transmitter also. I've
definitely seen differences in cable leakage but have not taken
qualitative measurements.

Regards,

Mark

On Sun, Mar 14, 2021 at 8:07 AM Charles <charlesmorris800@centurytel.net> wrote:

Thanks for the info. I'd forgotten about the shield coverage! Can anyone help me devise a simple test that doesn't require switching 200 amps? :) I can generate 500 watts or more of RF at 50 ohms with my ham gear, but only up to 30 MHz.

These cables actually do look reasonably well made, but of course I would not expect them to hold up to daily continuous use like Tek or GR cables. For 99 cents each, they'll probably be fine for occasional tinkering.





Greg Muir
 

I think I may have mentioned this before.

When attending a seminar at NIST (then NBS) I learned quite a bit about coaxial cables and their connectors. With regards to cables carrying the MIL RG numbering system not all manufacturers cables are the same. For a true RG cable meeting the original military specifications for shield coverage, dielectric, size and even impedance the designation will specifically state “RG-xx” or “Mil-C-17 RG-xx”. Those cables that contain the designation “RG-xx” followed by “Type” are cables that look like the military styles but do not necessarily meet the published specifications.

As for Chinese cables, well, you are on your own. They don’t seem to follow the rules of the game when it comes to labeling things.

Greg


SCMenasian
 

Not all RG58s are the same. Depending upon whether a cable is RG58, RG58/U, RG58A, RG58B, RG58C, etc. the impedance can vary from 50 ohms to 53.5 ohms without even taking into account differences between various manufacturers and each manufacturer's process control. RG223 might be a better choice, particularly for high frequency applications. It's double shielded and microwave power loss is lower. IIRC, the impedance is more dependably closer to 50 ohms.


 

I'm learning a lot about cables ;)

The cable jackets do say "RG-58/U" but, as mentioned, the offshore factory will print whatever you want. There have been documented cases where they lightly sand the top of whatever batch of IC's they have on hand (or even empty packages), and print the "correct" marking for the chips the customer ordered! They will go to the trouble of putting fake undersized die in TO-3 transistor cases, so they pass a superficial test but fail in use... Lots of stories here: https://hackaday.com/2018/01/15/is-that-part-a-counterfeit-here-are-a-few-pointers/
There used to be a great web page that had pictures of all the phony semiconductors but it seems to be offline.

Anyway. For comparing cables I would use the GR ones as my standard, thus compensating for any leakage from the transmitter and dummy load (unless, of course, said leakage swamps the hopefully small amounts from the cable under test). I also can generate about 20 watts at 148 MHz, which would be easier to detect than 30 MHz leakage...