These days, anything with copper in it is expensive. If you doubt that, a walk into any Home Depot electrical department, where the wire is locked up tighter than Fort Knox, will prove otherwise. Coaxial cable is a particularly expensive species, which is a pity for hams and other radio enthusiasts since it’s the only thing we can use for antenna feedlines.
Or is it? [Steve (VE6WZ)] has found a way to use ordinary Cat 6 Ethernet cable for antenna feed lines that seems pretty clever. As he points out, Ethernet cables are designed to handle frequencies that coincide nicely with most of the interesting amateur radio bands, and their insertion losses are acceptably low, especially for Cat 6 cable. The twisted pairs are also a balanced system that’s good at rejecting common mode noise. Cat 6 cable also has four pairs of conductors, allowing you to feed multiple antennas with one cable, or to distribute power to amplifiers and switches along with antenna feeds. rvvb cable
The downside? Cat6 conductor pairs have a characteristic impedance of around 100 ohms, which isn’t a match for the 50-ohm feedline impedance universally expected by ham radios. Also, the relatively small wires probably aren’t up to the job of carrying much current, limiting their use to feedlines for receive-only antennas. That works for [Steve] since he uses Cat 6 to support his massive Beverage antenna farm (Beverage antennas are non-resonant horizontal antennas that live close to the ground and point in the direction of the signal, rather than broadside to the signal as with a resonant antenna like a dipole.) Each antenna in his farm has a transimpedance amplifier that needs to be powered, plus switching relays so he can turn the correct antennas on for the signals he wants to receive. He describes the amps in detail in the video below, along with the custom impedance-matching transformers he uses and the combining gear.
Coax will probably still be the cable of choice for most feedline applications, but it’s nice to know there are alternatives. And who knows—if you stick to QRP work, maybe Cat 6 could even be used for transmitting.
Oh the irony: For some microwave-and-shorter applications, you put the radio and the antenna in the same place, then run CAT5/6 (plus power) back to the shack where your laptop is controlling the radio.
Also, who says you need coax for HF? It may not be the 20th century anymore, but twin-lead still works.
Twin lead is hard to find these days. I haven’t seen it in a store for a long time, you have to order it from a ham radio site. It’s also a pain to use because it must be kept away from metal and excess cable can’t be rolled up.
You can’t make some spacers and run it?!
Yes, it’s just a lot more work to run it than coax. It’s not cheap anymore either. If I was going to go through the trouble, I would use 450Ω ladder line and a balanced antenna tuner.
Yes, make your own twin lead. Easy, cheap and fun. If you are unsure, Google is your friend.
Can the coaxial cable be used to feed a terrestrial antenna to the rear of the TV?
I wish I had a “massive Beverage antenna farm”, my lot is to small. I may try this for a LOG though, it would make getting power to the LNA easier!
Do you mean, ‘Too small’?
IIRC one thing about balanced feedlines is that you need to keep them away from metal, especially grounded metal (remember the little screw in “lolipop” insulators for 300 ohm TV feedline?
So zip-tying your balanced feedline to a metal fencepost is probably not helping your signal loss numbers
Twisted pair is not affected by nearby metal. You are thinking of the old parallel twin lead. It would really be a problem for computer networks of we had to worry about nearby drop ceiling grid, metal studs, wiring and many other facts of life in your typical modern office.
Want 50 ohms? Put two pairs in parallel, and halve your losses while you’re at it. You still have two pairs to run 100 Mbit ethernet and can still have POE too.
Do transmission line impedances work that way?
Yes. That’s even a standard hack in how to create a quarter-wave transformer to combine antennas: a pair of 50-ohm antennas in parallel present a 25-ohm load, and you match that to a 50-ohm transmitter with a 37-ohm quarter-wave transmission line, made with two 75 ohm lines in parallel.
Mind the velocity factor of the cable when calculating the quarter wave, though.
Yea that’s what makes the idea of transmission line impediences useful!
My thoughts exactly, it’s all to convenient, is there a reason this wasn’t considered? Don’t know much about radio
Yeah, the highest POE type says something about a max current of about 1 amp although with significant droop possible, and it delivers up to 90 watts IIRC before you start going out of spec. Which is enough that a small amplifier could probably be used at the end, which would keep power from reflecting back into the transmission line.
The application here is for receiving…
Does not work that way, it is not a resistor.
Not really. They’re 100 ohm differential, not single ended. If you do it with UTP, it’ll work pretty well, but if you do, say, CAT6E, if you try tying one side of each pair to ground and also have it shielded to your ground, it’ll be well below 50 ohms. The single ended impedance (to shield) is already ~50 ohms in CAT6E. You could run 8 signals in one if you wanted (which I’ve done).
You really want a balun.
Yes, correct. The resulting 50 ohm two-paired line would still be a balanced line and you’d want a balun.
Trying to run eight RF lines down CAT6, expecting 50 ohms, is going to result in a horrific amount of crosstalk between each conductor of each twisted pair. If you run low enough frequency that crosstalk isn’t an issue, then neither is the 50 ohm impedance. But yes, for signalling or power or even audio, 8 signals works fine.
I’d love to see those waveguides for the 160 m and 80 m bands he’s using.
Sounds perfectly practical. When you’re not working the bands, you can park airliners in the 160 meter waveguide. Keep one around as a tuning slug!
🤣..perhaps add a bus for fine tuning!
Wonder how well it’s work to run a parallel twisted pair? Cut the impedence in half and get more conductor for current carrying.
Won’t a bigger concern be that most inputs expect unabaanced input?
This reminds me of a customer who complained about our software not working in a network. After investigation, it turned out that their ethernet network was made of telephone wire – untwisted pairs.
Experiments to save money don’t always work well.
Ethernet over telephone wire should work … over slow speeds and very short distances.
I would expect a 6-inch 10mbps link to work well enough that most people won’t notice you aren’t using CAT5 or better. :)
Bad luck that was long distances, noisy environment and high speed… Gave us a good laugh though !
I was involved in one of those “let’s see if we can double the number of computers in the room” situations a long time ago. There was just about enough CAT5 in the ceiling to handle the existing computers. It would be awhile before we could do it right by adding more wires. Yup, you guessed it, we used the 2 unused pairs and doubled the number of computers on the LAN. It “worked” well enough for our purposes, but I think we had to crank the speed down to 10Mbps or make other compromises. We eventually did it right and added more wires.
This was one of those unfortunately-not-so-rare cases where DOING THINGS WRONG AND OF COURSE DOCUMENTING WHAT YOU DID AND BUDGETING FOR A DO-OVER AS SOON AS PRACTICAL was the right thing to do.
I hope I’m not in a similar situation again, but it’s nice to know that my 1337 HAXX0R SK1LZ let me get the job done without killing anyone or anything. “Joe, hand me that piece of CAT5. No not the red one with the wall plug on the end, that’s the EtherKiller adapter we made last April 1st. Hand me the blue one. [pause] What do you mean you used the blue one already? What for? [RUNS TO THE SHUT OFF POWER TO THE BUILDING BEFORE IT’S TOO LATE]”
I’m a bit surprised you had any problems – assuming individual pairs won’t untwisted too far, etc. As you know, gigabit does 250 Mbps over each of four pairs using extremely similar cable. It’s basically four 250 Mbps links, using the four pairs.
Depending on the connection there shouldn’t be two unused pairs with traffic spread across all of them.
However, IIRC, depending on the setup you should be able to do 2 100mbps links on each cable with 2 pairs for each.
I have run across flat satin aka telephone cable stapled down a wall after running right next to fluorescent lights. It still managed 100Mbits. I highly suggested to the manager that she get with her telephone guy and get it rerun with Cat; away from those lights. She told me that one of the other techs from the “Big Corp” I worked for at the time did the original flat satin; ugh. In a different senario I have also used Cat in an industrial plant to provide 12VDC. We had join two pairs for negative and positive due to the distance but it worked ok. The other pairs ran serial comm.
yeah i had a lot of paired duplex jacks like this just for convenience…i actually had plenty of wire on hand i just didn’t want to pull extra wires for all the different jacks that were piling up under my tv. one for the android tv, one for the raspi, one for the nuc, etc.
worked great until my switch died and then i bought a new one and noticed every single device in the whole house — including the modem! — had upgraded to gigabit without me even desiring it…so then i felt the need to run proper 4-pair to every jack
trivia…depending on the endpoints, there is a minimum length for gigabit ethernet cables. took me an hour of futzing around before i realized the difference between the ‘bad’ cable and the ‘good’ cable was just that i had cut it to the exact length between my pc and modem, which was too short. felt good at least to see verification that i really have become mostly a “works the first time” guy with the crimp tool (finally! in my 40s)
This sounds like something from the Bastard Operator From Hell stories. Though he would not worry about shutting off power to the building.
As others have said, if you were using two pairs each from real Cat5, you should’ve been able to do 100 Mbit with no problems at all. The only thing I can think of that might’ve impacted your speeds is if you didn’t have the pairs wired correctly and the lines weren’t protected from crosstalk. (Whether you’re using T-568A or B, pins 1&2 should be one pair, and pins 3&6 are the other pair.)
I had a client who had untwisted cat 3 between buildings (like 30 ft away). To my shear amazement they got 900mbps throughout with almost no packet loss.
FYI, CAT 3 is an early variation of unshielded twisted pair.
Did they manually untwist the cat3? I thought that was always twisted, could be wrong
Twisted pair impedance isn’t constant along length and isn’t regulated – it can be anywhere between 100 and 200 Ohm. Result would be random. It’s not a waveguide.
With all respect- you’re 100% wrong. Twisted pair impedance is definitely constant and regulated- nominally 100 ohms. The impedance is determined by the distance between the conductors and the dielectric. Ethernet cable has different twists on each pair to reduce crosstalk, but the twists are slow enough to not significantly change impedance. NB: I built a test lab for and did wired ethernet testing (to the 802.3 specification) for a major computer manufacturer (rhymes with Hell). I know what I’m talking about.
With fence as antenna you will get very high resistance definitely not good for CB radio or ham radio
The twisted wires’ 100 Ohm impedance can be easily solved. Instead of using one wire pair, connect two pairs in parallel – the result is 50 Ohms. On the two samples that I have tested, the VSWR <1.8 in the HF range. https://sites.google.com/site/randomwok/Home/electronic-projects/cables-connectors/lanethernet-cable-for-connecting-to-active-loop-aerial
The 100 ohm impedance is easily solved. Instead of using 1 wire pair, use 2 and connect them in parallel. Testing 2 samples, the results are better than -14 dB return loss in the HF band. https://sites.google.com/site/randomwok/Home/electronic-projects/cables-connectors/lanethernet-cable-for-connecting-to-active-loop-aerial
The problem is that with Ethernet cable the twist rates are different between the four pairs, so the electrical length between the pairs differs slightly. This makes connecting them in parallel have anadditional problem.
The difference in electrical length is small and the wavelength for HF is very long. Might work fine in practice?
So, if you’re using pairs X and Y, at the exact hafway point of the cable break and connect pair X incoming to pair Y ongoing, and vice versa. The electrical lengths of the two composites are then equal. QED
This is already done in the twist schedule of the wires: they alternate in pitch with distance to equalize the electrical length between pairs. I can’t recall or quickly find the repeat interval, but it’s on the order of a few meters, so even if you pick a “wrong” location on the wire to cut it you won’t be off by more than a fraction of a nanosecond between pairs.
Was there an S12 figure in the video? Skimmed through but didn’t see, I assume radiative losses will only become relevant at higher frequencies?
I have measured the insertion loss (S21) of 15m cables (n =2). https://sites.google.com/site/randomwok/Home/electronic-projects/cables-connectors/lanethernet-cable-for-connecting-to-active-loop-aerial
Do you know what is better than a nice long piece of even the most expensive coax money can buy?
Just move the radio to the antenna and use the Cat-6 to remote it.
This isn’t unusual for microwave and above, where feedline losses can be high if you don’t use very high quality (read: expensive) coax and connectors.
For your typical 100W HF rig though, it’s frequently cheaper and more convenient to put the radio in your heated/air-conditioned shack and run a feedline out to the antenna than to buy a weather-hardened 100W radio then run both power and coax back to your shack.
That said, if you are going to put a radio and antenna up at your vacation cottage (which I don’t have and probably never will) and control it remotely year-around over the internet, then a weather-hardened remote-controlled HF radio that “lives at the antenna” is certainly worth looking into.
I wonder why coax is the go to solution for some applications such as experimental vibrations measurements, despite the sensor manufacturers documentation showing better EMI performances for for shielded twisted pairs. The same may apply here.
This will be lossy for uhf tv channels. CAT6 is good up to 250mHz, TV UHF us 470-600mhz
Stuff being locked away in big box stores is not because of increase in raw materials prices… It’s because of recent experiments in societies without law enforcement
Please be kind and respectful to help make the comments section excellent. (Comment Policy)
This site uses Akismet to reduce spam. Learn how your comment data is processed.
pv wire for solar panels By using our website and services, you expressly agree to the placement of our performance, functionality and advertising cookies. Learn more