On the Elecraft-KX mailing list, an item popped up on a topic that I’ve been pondering for a while. Titled Balun necessary? Barry N1EU asked the question:
I’m planning some 40M-15M portable operation with a 44ft doublet fed with 25ft of twinlead into the KX2 ATU. Is it, or is it not, necessary to use a balun between twinlead and KX2 antenna terminals for this application?
I’ve been pondering this very question myself, as my go-to antenna for portable operation with my KX-3 is a 66-ft. doublet fed with twisted pair. I’ve just been connecting the twin-led feed line directly to the radio with a BNC-binding post adapter, but it occurred to me last summer that I should perhaps be connecting the unbalanced output of the KX-3 to the balanced feed line. The antenna seems to work just fine without the balun, but I wonder in the antenna system would be more efficient with one. At 10 – 15 W out, you want your antenna system to be as efficient as possible.
As you might expect, this question generated a lot of replies. Many of the replies simply said, “A balun is preferred,” without really giving a reason for this. Many write that using a 1:1 balun or current choke would prevent common-mode currents on the feedline. The purpose of this being to reduce the chance of the feedline radiating and causing RFI or to reduce noise pickup.
I jumped into the conversation when it was well underway and mentioned that I have been using a doublet with twisted-pair feed line. That prompted a private reply from Barry, and we struck up a conversation on the relative merits of twisted-pair feedline vs. 300 Ω twinlead and whether or not to use a balun. In one of the emails, he wrote:
FYI Dan, did a little experiment. I set up a 44ft doublet on a summit yesterday and ran some 1 watt WSPR transmissions on 40, 20, and 15m using first a 300-ohm twinlead feedline and then a twisted pair feedline (PTFE insulated, harvested from surplus CAT5e cable). I’ve always been intrigued by the super lightweight twisted pair and several years ago used it once with very good results (high score in 2014 QRPTTF/SOTA event). But I always wondered about the loss involved, especially with the high SWR encountered in a non-resonant antenna.
I lacked sufficient time on the summit to be very thorough and methodical but I believe the data is valid. 30-45 minutes separated the transmissions for the different feedlines.
- 40m – 24 reports for both – twisted pair averaged -3.75dB down from twinlead
- 20m – 27 reports for both – twisted pair averaged -1.56dB down from twinlead
- 15m – 13 reports for both – twisted pair averaged -4.15dB down from twinlead
Perhaps one day I’ll do a more thorough and methodical test but I think these results show that PTFE twisted pair is a viable balanced feedline and makes for an ultralight option for SOTA activation, although twinlead is preferred when weight is not an issue.
Other equipment used included an Elecraft KX2/ATU and Android phone running WSPR Beacon. The 300-ohm twinlead was connected using a homebrew dual FT140-43 core 4:1 Guanella current balun and the twisted pair was connected using a homebrew FT140-43 common mode choke (12 turns).
I should probably do some of my own testing. While scouring the shack for stuff to take to a hamfest a couple of weeks ago I ran across some quality 300 Ω twinlead, so I could make up a 66-ft. doublet with the twin lead feedline. I also have ferrite cores and try winding the twisted pair feedline around it to see if that makes any difference. Barry’s test show that the twin lead is more efficient, but twisted pair is easier to handle and does wind up into a smaller and lighter package for transporting to and from a POTA or SOTA site.
It would be interesting to do some testing with the 44-ft. doublet. In general, of course, the more wire in the air the better, but less wire again makes it more manageable. I’m going to have to download WSPR Beacon to my Android phone, too. In any event, it looks like I’ll have a lot of fun stuff to play with this summer when I can get out and do some activations.
Ed Woodrick says
I think that you need to go back to what a balun is, a balanced line to unbalanced line converter.
AFAIK binding posts to coaxial is a BALUN.
But what’s missing is the impedance. The lines generally don’t have the same impedance and that creates an issue with power transmission.
Is it a problem, not necessarily, can it be made better? Absolutely
Clive G6TDI says
Dan,
All wires are at the end of the day transmission lines or waveguides, even single wires can carry power from one end to the other without radiating it (see G-Wire or Goubau line). The thing to remember is that the power is not “in the wire” but “in the field around the wire” which is why we have “waveguides”.
Mostly it’s not important till you realise that the “insulation” around the wires is not just a dielectric it’s also an absorber of power, which is why at high SWR where power bounces back and forth along the line “power loss” is greater in heavily insulated or field captured lines such as coax, but minimal in two wire “twin lines” where the wires are spaced out with only tiny bits of insulation.
The idea with twinelines is that the movment of charges is in opposite directions so cancells out “entirely” only it does not due to either conductors or insulators causing localised changes in balance (remember you can make the equivalent of a microwave yagi with stacks of polystiren and polyprop discs at the end of a waveguide by the dielectric changes acting like lenses).
Now the idea of twisted line especially tightly twisted (at around 93 ohms impedence) is that you are kind of making any near by metal or insulator look likecit is wrapped around the twinwire thus it effects both conductors equaly thus keeds things “ballanced” which is a good thing especially if you have a blasting cap at the other end (twisted pair was originally designed for mining and telephones). The downside though is that insulation causing loss. Not to bad if the impedances are matched but, crap if the SWR is high.
The down side of twisted pair is it’s impedence is ~100R which is about twice that of the 50R of the transmitter. Which means any “impedence match” by transformer has the,”square root of 2″ problem that is the turns ratio is impossible to achive as it’s 1.4142… as,an irrational number with 10:7 or 7:5 giving you about the best approximation. Which might be good for certain common cores at the low end of the HF band but…
The thing about doublets are the are almost always not matched thus the SWR is not going to be good… Thus tight twisted line is going to be lossier than you would like. Worse if you want to work multiple bands you start getting into “random wire” problem territory. As you know the “random wire” is anything but random and needs to be not close to either a 1/4 or 1/2 wave at any frequency of opperation.
But also you have to remember Pythagoras’ little theorom of A^2 = B^2 + C^2 as it tells you your Zo for two lengths of different impedence line on a quaterwave match…
So you can see why life can be fun…
My advice as “balanced matching units” are a pain to make and use as well as tend to be large, and I’m asuming QRP is go for a 7:5 on a pig-nose (two hole) core remembering that each time the wire goes through a hole it’s a turn so what looks like a single full turn from above is realy two turns.
Oh and don’t forget the capacitor on the 50R input side, A few inches of PTFE RG316 can generally be cut to give the best performance.
Tom King KT7SS says
Hi Dan
I just finished your audio book. Loved it.
I’ve worked in the wireless field for over 20years.
Something I thought about and am doing some testing with is using 2 twisted pairs in parallel from a Cat5 cable. They are nominally at 100 ohms and putting them in parallel get you 50 ohms impedance. Then use scotchlok connectors to add a dipole from some of the other pairs harvested from cat 5 and you have a cheap feedline and antenna for portable use. Also very lightweight.
Dan KB6NU says
That sounds interesting. Let me know how it goes and we can throw something up here on my blog.