75:50 ohm UNUN

(1/2) > >>

Bruce Danyluk:
I would like to build a 75:50 broadband transformer (UNUN) to match 75 ohm coax to 50 ohm coax for use on HF.  I found plans for such a UNUN in an old ARRL handbook (1989) but the design is based on a FT-200-61 toroid core which I have been unable to find.  I have checked a number of websites (Amidon, Palomar, etc.) without success.

Does anyone know of a source for this core or know of another design using available components?


Bruce - VE4AEK


Seems like an FT-240-61 is pretty widely available and I think it would work fine.

Is this a high power application?  I accidentally recieved some B-002-61 binocular cores (sort of 1.2 inch square, .6 inch thick binocular cores) from CWS Bytemark


and I used them to build a couple of 75:50 ohm ununs .. think just three or four quintifilar passes through each hole, and they seem to work OK from 160m thru 10m @ 100W level.   I used the wire from inside CAT-5 cable.  I built two of them and wired them back to back and measured something not more than 0.13dB loss apiece, and no noticeable heating maybe 30 second keydown.

Now, I didn't make a very good loss measurement; just a wattmeter, from memory got at least 94 watts out of the back to back pair for 100W in, and I don't remember if I tested them across the whole frequency spectrum.  

Of course, at 2kW they'd each dissipate nearly 60W and soon fail.

I don't necessarily recommend the binocular cores over the toroidal one, but it did make for a very nice, compact transformer that fits inside a BUD minibox.  I think the moral of the story is that it's actually not all that critical what core and wire you use for these, at least at low power ;-)


Bill Burrows:
Why bother, the mismatch is 1.5 : 1.  Most transceivers consider this an acceptable match and the coax losses are not increased significantly.

For a single band you can also match it using a piece of 50 ohm line cut to the right length.  Google Regier match or Branham match.

"Why bother, the mismatch is 1.5 : 1. "

Yeah, from the antenna to the line... which sets the loss, but the transceiver could see something higher.

An odd multiple of a quarter wavelength of RG-6 will transform 50+j0 up to about 110+j0 and some transceivers will balk at that.  It gets even worse if the load is lower impedance than 50 ohms.  35 ohms, 1.4:1 SWR on 50 ohm line, gets transformed up to around 150 ohms by an odd quarter wave length of 75 ohm line, and now the rig sees 3:1 and shuts down the output.

Changing line length will fix that of course, but if you start talking 40m or 80m, an extra quarter wavelength of line is a lot to have lying around.

So I think matching is appropriate here.

"Google Regier match or Branham match."

That doesn't work... but I think that's the same as a "twelfth wave transformer"

And they'll actually work well over a few bands.  My 20m/17m moxon uses such a match cut for 16MHz to transition over to RG-6/QS type coax.

There's another series section matching setup (maybe associated with one of those names) that breaks the transformation into shorter chunks for even wider bandwidth and I used one of those to match 75:50 ohms from a magloop that tuned 5-21MHz and that seemed to work fine.  Those series-section matchers are quite wide, really... but probably won't cover 160m-10m like a ferrite core would.  You also end up with a lot of connections to waterproof or a big coil of coax inside too.

Great technique, though, and certainly something to have in your bag of tricks.


Bruce Danyluk:
Thanks, Dan and Bill, for your comments and suggestions.  I should have described what I am trying to do.

I am putting together a new station for HF after an absence from ham radio of many years.  Initially, I plan to run 100W to an N4GG Array (see ARRL website) for 40 and 20 meters.  The N4GG Array is a fan arrangement of 3/2-wavelength dipoles with the outer 1/4-wavelength of each dipole element hanging vertically.  The feed-point impedance is 73 ohms, hence the 75-ohm feedline.  I will be using a homebrew 1:1 balun for 75-ohms (G3PTN design) at the feedpoint.  The feedline will be around 125 ft and I am looking at using Belden 1694A RG-6 cable with upgraded (Canare) F-connectors.

I was hoping to minimize my losses by using an efficient homebrew UNUN.  (I have to change from the F-connectors to UHF connectors at some point anyway, and I thought this might be done at the UNUN.)  Of course, I don't want to introduce any significant losses with the UNUN since in that case I might have been better off living with the small 75:50-ohm mismatch!

Bruce - VE4AEK


[0] Message Index

[#] Next page