A coax cable horror story | Hacker News Reader

A coax cable horror story

qrper.com

65 pointsthcipriani49 comments

> I have bought a couple of different lengths, mainly from Chinese online sellers at AliExpress.

You can stop reading right there...

This is not a new problem however. 20 years ago you might go to the local ham store and pretty much any cable in a plastic bag was going to be total crap.

I got a NanoVNA for about $60 and it works way better then I had any right to expect. If you have any need to figure out some kind of RF issue it is well worth the money for the amount of time you will save.

I used it to diagnose a VHF issue on a boat. The AIS was reporting high VSWR but only sometimes... It turned out that the antenna was seriously affected by it's angle relative to the metal poles that support the Bimini top.

Also, a lossy cable is not always a bad thing. A lossy cable can stabilize a PA if you are forced to have bad match at the load end.

recently picked up a nanovna as well. I am having a lot of fun characterizing antennas and cables. suddenly it’s clear to me why this cheap 144mhz whip on my desk is only useful on a very narrow set of frequencies.

nanovna: great buy!

Bending coax that tightly is usually bad. I once put a time-domain reflectometer on some RG-58 and, and it would show a 3" radius bend. The center conductor to braid distance has to remain constant. The price of high-flexibility coax is some loss.

This is very true. Good terminations and good materials matter. By far the best coax is silver semi-rigid -- brilliant for the bottom of my lHe cryostats, not exactly a cable in any other terms.

One thing this article didn't mention is hope the length really matters if that meter isn't perfectly impedance matching. Stub and strip lines and all that.

You've touched on what is probably the real issue, but without further measurements, we cannot know for sure. The author said (in the comments) that they used an antenna as the load. With a minimum SWR of 2:1, it's obviously not a very good match. If the match is bad, the length of the cable could have a big effect on the measured power. Other comments suggested using a resistive dummy load instead of an antenna. THAT is the way to go if you are measuring cable insertion loss.

The two cables with the poor insertion loss were 3 and 10 meters in length. None of the good cables were of these lengths.

A geophysics prof I did a project with one summer gave me this sage advice:

"Be good to your coax, and your coax will be good to you."

I don't if I would call it a "horror story". A horror story would be that someone bought cheap coax cables off Aliexpress, and when they hooked them up all of the smoke came out of the radio.

In this case, the radio operator hooked up the cables and weren't transmitting as well as they thought they should be. Some measurements indicated that the cables were signal-attenuating crap. Cables that were purchased from AliExpress. It's almost as if one gets to the end your campfire story, and one of the campers pipes up and says, "don't tell me, the call was coming from inside the house?", only I piped up at the beginning and said to myself, "coax cables from AliExpress? Don't tell me, you found out that about half the transmission power went to creating heat and not radio signals."

I don't know, about the requirement of smoke to be a horror story. I have been in my own cabling hell, and no smoke was ever present. Cabling issues can be tricky, especially when they are intermittent. Everything works except when you sit in a chair. That chair with weight puts pressure on something that then squeezes something else. Specific example of something that happened.

I've managed post production facilities where there were thousands of cables running in multiples to/from various pieces of gear. You can pre-terminate, test, label all at a work bench with whatever test equipment you like, but once you start running the cables in, through, under, over other cables, plumbing, etc, things can just get weird.

We are remodeling our house so we fished coax and ethernet through the office, to TVs, and to strategic mesh hotspot areas.

When I hooked up my modem to the new coax in my office, the modem appeared to connect. However, my TPLink Deco (mesh network) wouldn't connect to the modem. I could hardwire my laptop into the modem and it worked. Confused, I spent a few hours, unplugging, resetting the Deco (never want to do that again), and finally giving up.

I told my electrician about it when he passed by a few days later and he cut off the coax tip and put a new one on. Lo and behold it worked. The Deco connected.

I'm still confused. His insight was that maybe the Deco was tracking the signal and the tip didn't have a clean connection.

Is there a reason you’re willingly using coax in 2023 when you wired Ethernet as well? There’s hdbase-t if you want to pass media

I read it as coax from outside to the cable modem, not between 2 Ethernet capable devices.

Yes this.

Coaxial is useful if want to put cable box or modem in another room.

I have charter/spectrum but not by choice. If I were buying/building a house, I would give preference to places that have fiber.

Here is the cable company recommends you only put the modem within a couple meters of the outside connection and not run it through the house.

That is for 1.4gbit though, so the requirements might be tighter.

That might explain why I only get 850M down instead of 1G...

However, I wanted my modem in my office so I can know if my network goes out during work (I teach online) and try to mitigate the issue.

Is coax still popular in the US (and presumably in other countries)? Over here in the UK I don’t think I’ve seen it used for 10 or so years now as terrestrial TV antennas have fallen out of use. Genuine question as I never even considered it as used for anything other than A/V so TIL.

It's used for cable television and cable internet (which is most high-speed internet in the US). Generally you don't have to think about it too much: you'll have a coax jack in the wall and then you connect it to your TV and/or cable modem and then leave it alone, but yes.

Outside of cable internet / TV service, it's frequently used by amateur radio operators, though in a 50 ohm resistance. Most (all?) cable/TV coax is 75 ohm.

Amateur Radio operators do not use cables with 50 ohm resistance. Perhaps you are thinking of “impedance?”

what are you talking about? Drive _anywhere_ in the UK and you'll see an aerial TV antenna on almost every roof.

sorry to be snarky there. Your post read to me like, "I haven't seen a banana in the US in 10 years. We don't eat them anymore."

I'm sure some (plenty) of Brits do still plug-in their TVs to those antennas, but anecdotally I can't think of a single friend or family member in the UK who I've noticed still doing that.... most either use a TV package from their ISP, or no TV at all. And yet we all still have those antennas on the roofs because why bother removing them?

On the other hand, doesn't Virgin Media (pretty much the only cable provider in the UK) use coax into routers/TV boxes?

A common issue with outdoor coax cable is when the connectors are not sealed well, water seeps in and causes gain reduction from water being in the cable. Especially with diy sealed cables, I've run into this. The giveaway is weirdly high SWR ratios.

Another problem, one that I experienced as I was new and unschooled: It was a long run to the antenna and I left too much slack before the transmission equipment. I just bunched the slack into a jumble and then hooked it up. Of course the engineer that lectured me about conductivity, magnetic fields and SWR remains a memory. He neatly coiled the coax and my problems disappeared. Even something as simple as coax benefits from experience and a small bit of knowledge.

Uh. I shudder to remember all the problems I had with 10Base2. And the electric shocks. I think I still have PTSD from all the electroshocks.

Any standard cable, wallwart, etc from Alibaba cannot be trusted. They take liberties with the specifications in order to cut costs. Worst case scenario you're outright scammed.

"I can not say, if it is a loose connection, a faulty plug or if my handling was too rough. It could be that all cables arrived fine, and I damaged the cables during my usage. But to be honest, I don’t care."

That doesn't seem a good test, aside the fact that the RG316 is a quite a lossy cable, Chinese or not.

Heh, I just wrote a blog post this weekend where I tested, among other things, the loss of some coax cables:

https://tomverbeure.github.io/2023/05/06/HP-11720A-Pulse-Mod...

My results are much worse, with a minimum of 2.5dBm, but I'm measuring frequencies that start 1.5GHz and that go all the way up to 20GHz.

It'd be interesting to redo these measurements at the frequencies of this blog post. I think they're doing the measurements at 28MHz?

Hi, I apologize for being the RF pedantic, but did you mean "2.5 dB"? dB is a ratio; "dBm" is just one way to declare a given power, in this case because it is referenced to 1 millwatt (the 'm' in dBm). Ergo, 0 dBm = 1 milliwatt; 30 dBm = 1 Watt, etc.

I hope you were very careful with your fixturing going up to 20 GHz; poor fixturing can be masking the cable's true performance.

Yeah, I meant dB.

And, no, I wasn't very careful at all. This was done with an uncalibrated 30+ year old sweep generator with questionable attenuator, a 40 year old power meter that's only rated up to 10GHz, AliExpress connector adapters, and an assortment of coax cables that I took from my stack.

At 20GHz, it shows a trend at best. (I wrote as much at the bottom of the blog post, but I forgive you for not reading that far!)

Also beware of bad quality SMA connectors from cheap Chinese sources. They may have significant dimension error, and have the potential of destroying mating good ones.

I had no idea what QRP is so I looked it up. Wikipedia says it's for HAM radio lower power devices.

> The term "QRP" derives from the standard Q code used in radio communication, where QRP is used to request "Reduce power" and QRP? is used to ask "Should I reduce power?"

2/3 cables with > 4dB drop!? You kidding me?? It's a sabotage!

I was hoping this would be about vampire taps on coax Ethernet cables.

Reply on news.ycombinator.com