[Antennas] Some thoughts on baluns

[Charles Greene]

Kees,

If it works, use it.  Every installation using the type 77 is not going to 
overheat at any frequency.  It it does that's a sure indication that you 
have excessive losses.  My experience is that under certain power 
conditions and balun locations, the balun will overheat.  My latest attempt 
was to use a choke balun in an EH antenna where the balun was in the field 
of the antenna.  A very small balun made from some small type 43 beads on 
RG174 got real hot at 10 watts but only warm using a larger choke balun 
using larger beads on RG58.  The loss was still there, but you didn't 
notice it as much.  When I used the balun at the other end of a length of 
coax, the balun did not overheat.

I have found that a balun of coax wound on a ferrite core or air wound have 
a much smaller bandwidth than a W2DU type choke balun using beads.  I was 
going to investigate this phenomenon this Winter and come up with a better 
design, but I have been overtaken by events so far.  One way around this is 
to use two or more baluns in series if you are going to cover 160 to 
10.  Place one that covers 20 - 10 in series with one that covers 160 - 30 
for example.  Same applies to a choke balun.  You may have noticed that the 
Palomar choke balun kit recommends using two in series for 160.  You can 
measure the bandwidth by connecting the output to a 50 ohm dummy load or 
non-inductive resistor and using a SWR bridge with a power source connected 
to the input.  Or you can use an antenna analyzer.  Generally, satisfactory 
performance will be indicated by a SWR of 1.2:1 or less.  If the balun 
isn't going to work well into a dummy load, chances are that it will not 
work well into an antenna either.  This is not the final test.  You need to 
measure the impedance of the coax outer conducter too.  1000 ohms is a 
target, but 500 ohms and over 1000 ohms is ok for most installations.  You 
need to watch out for resonances if the balun has turns anything including 
air.  If the balun is resonant, it could burn up pretty fast.  An antenna 
analyzer should show you resonances.  This is another reason why it is 
difficult to build a coiled balun that has adequate performance from 160 
through 10.  That's why I prefer the W2DU choke balun over the balun with 
turns on anything including air for wide band use.  However, you should be 
able to use a coiled balun with an air core or ferrite core over a limiter 
bandwidth, like from 20 to 10 for a triband beam, or 160 - 40.  I really 
haven't done a lot of experimentation on coiled baluns for wide band use, 
but they are great for a more limited bandwidth.  One of my favorite baluns 
is a 1/4 wavelength at lowest freq of RG58 wound on a 1/2 gallon GatorAide 
bottle.  Works fine over about a 2:1 bandwidth. but it's a little ungainly 
to put up in the air.  I use it at the ground level and use open wire feed 
line up to the antenna.  For an balun at the antenna, I have had best 
results with a 4:1 current balun or a W2DU type choke balun.

At 08:55 PM 1/11/2003 -0600, Sandy and Kees Talen wrote:
>Thanks for the response. You are right in that a consideration is
>overheating the beads and changing characteristics. What I was
>trying to do was use a choke balun in a <200w 160m to 10m tuner.
>So far I'm at two large type #77 beads with 4 turns, 4.5" dia, of
>RG/58U through them. The reasoning is that the 4 turn coil fixes
>some shortcomings at 21-30Mhz, the two type #77 work well at
>160m, running multiple turns through them increases their
>effectiveness approx 16 times (and increases heat ?  ...I thought
>more choke impedance decreases the heat)

Depends.  Spreads it out for one thing.  If you use very large cores or 
beads, the loss is still there but you just will not notice it.

>....the down side is
>any heat generated  ...which I hope to offset some by the fact
>that the 4 turn coil will take some of the heat due to it's impedance.

Nope.  Any heat at all is a loss.  What heats the balun up anyway?  It's 
not inductive reactance. The losses in an inductor are due to its 
resistance at RF (not impedance), dielectric losses; and if it has a 
ferrite or powered iron core, losses due to circulating currents in the 
core.  A good balun will be about 97% or higher if air wound efficient into 
its characteristic impedance (50 ohms for a 1:1 balun), and about 94% or 
higher if air wound into a reasonable SWR, say 5:1.  These are losses of 
tenths of a dB; undetectable in an antenna system.

>Maybe add a few (2-3) more ferrites to distribute whatever heat.
>I've made impedance measurements and it looks good from 160m
>to 10m. I have NOT done any power testing ....and may have to
>get some hot dogs and fixins ready.
>
>I did try more mass with a larger, #43 material, flat ribbon cable
>ferrite spaced further apart with additional ferrite material but
>the results were poor at 160m. I was going to try two of these
>but haven't had a chance.


Try #14 enamel insulated wire, close spaced.  You want to get the 
charactistic impedance of the coiled transmission line close to the 
impedance you are working into and large enough wire such that ohmic losses 
are not a factor.


>I don't want to use a piece of coax with 50+ beads (W2DU), plus
>all indications are that they could be improved at 10m. Is that
>where they get hot ...10m and 15m ? (low impedance) or is it
>80m ? (high impedance) ....maybe add a 6 turn 4" coil to fix the
>10m low impedance ?

You may have some resonance problems by using more turns at 10 (or lower), 
or using a balun that works well using turns on 160.  I'd break the balun 
in two parts.  One that works well at 160 with one that works well at 10, 
but watch the low frequency one for heat on the higher freqs.  An heat at 
all is a sign of loss.  Measure the SWR of the total combination of baluns 
into a dummy load.


>The backup plan is the use a large 30ft long piece of coax (like
>in the AG6K tuner) with a "T" BNC connector at 6 turns and
>figure out how to switch that so the 20m-10m performance is
>satisfactory.
>
>Great fun.


Good luck Kees.  You probably can come up with a combination that 
works.  In my opinion, it is not worth the hassle; just use one or two W2DU 
baluns in series, or a well designed 4:1 current balun.  It depends on what 
you want to do: operate or investigate baluns.


>73  Kees K5BCQ






73, Chas, W1CG
K2 #462