[TowerTalk] Balun question(s)

[Tom Rauch]

> A fellow ham has built a "Cebik" 44' doublet for his
backpacking outings
> but his tuner won't tune it on 40.  Eznec says the feed
point impedance
> at 20' high is 20.7 - j629.6 ohms.  He has tried inserting
a balun.
> Most
> of the literature on baluns that I have seen never seem to
describe
> what goes on with the reactance part of the impedance.

I think I recall the article, and what I remember is the
main idea was to produce a clean pattern. If so, I have been
in this same discussion with several others who reported the
same problem as your friend has. I really cannot understand
why anyone would want to use that antenna. It is a poor
solution. Feed problems were ignored in the effort to find a
good pattern. The antenna is too short.

It is a design that looks OK in a model (because the model
has no tuner or feedline), but it really stinks when you try
to build it.
I'm going to assume your model is correct, and the feed
impedance is 21 -j630. That is a ***horrible*** feed
impedance.

Here's one way to look at it. Occasionally people whine and
complain about G5RV antennas. G5RV's, with 50 feet of RG-8X
feedline, have an insignificant ~1dB loss on the lower band
(80 meters). The impedance is also reasonable for a tuner.
There is NOT significant loss on 80, 40 and 20 in the
design. Any valid whining about a G5RV being a poor antenna
has to be entirely from a bad install like cluttered
surroundings or low height,  or perhaps from use on a band
like ten meters where SWR is high. The G5RV's reasonable
performance is due to the reasonable SWR presented by the
G5RV to coax on the low end of 80, high end of 40, and 20
meter bands.

In the Cebik dipole, we have further shortened the G5RV
short dipole from just over 100 feet to only 88 feet for 80
meters, which severely impacts ease of feed. Now losses in
the feed system clearly ARE a major problem.

Your friend made the antenna half size, so it starts on 40
meters.  Instead of an SWR in the  3:1 to 5:1 range (G5RV
style), a 50 ohm feeder now sees 266:1 SWR!!! This is not
really a good idea, regardless of how nice the pattern
looks. Through only 35 feet of RG8X, loss is 12dB or higher
and the tuner has to match an impossible load of 2.5ohms
j28. Tuners have the lowest efficiency matching low
impedances like that. It would not surprise me if tuner
losses were 3dB or higher.

Even if you go to ladder line, the SWR is 58:1. The loss is
2dB in only 35 feet of line, and the tuner has to match a
more reasonable 18 j376 and tuner efficiencies in the 80-90%
range are likely.

The suggestion of using an unloaded 2/3rd size doublet is
not good. I think it is a very BAD idea.

> Question 1
> What would the feed point impedance be if a 2:1 balun is
inserted in
> an attempt to raise the impedance towards 50 ohms.  What
really
> transformation of resistance and reactance occurs in
baluns.

The balun SWR would be 197:1.

Resistance would be transformed upwards as would reactance
if the balun could handle that impedance. I suspect it could
NOT. So now you add another major loss.

> Question 2
> According to my limited understanding of Baluns from
Sevick's book
> it is a bit of a misnomer to label baluns 2:1, 4:1 etc. as
they are more
> correctly 100:50 ohm, 200:50 ohm etc.  as baluns have a
characteristic
> impedance?  Furthermore is the 100:50 ohm (or 2:1?)  only
meaningful
> when the connections to it are 100 and 50 ohms?

Sevik is correct when speaking of transmission line type
baluns.

The transformation is reasonably close to 2: 1 for
reasonable load mismatches.  The effect of high SWR is to
tilt the impedance transformation to new ratios depending on
many things, but it's too complicated to get into without
writing a book on the subject.

In a few words, the idea of improving things with a balun is
unworkable when the starting impedance is so absurd that
even resonant networks have a problem handling the load.

> Question 3
> Would a coil sized (+j629.6)  to conjugate the -j629.6
help?  How
> should it be installed?

No, it won't really help unless you add it two 315 ohm
reactance inductors, one in series with each dipole half.
With coil losses included, you have a 50 ohm SWR less than
2:1 on 40 (you could actually get the SWR to 1:1 with a
single tapped j630 inductor or with three separate
inductors).  But you still have an antenna that is useless
on all other bands unless you change the reactance for each
band.

What you are trying to do is fix an antenna that in my
opinion was a very bad idea to begin with. It is very easy
to make a short antenna work well on ONE band, but difficult
to make it work well on multiple bands. The shorter you make
the antenna, the worse the problems become.

There may be some combination of stubs or shunt inductors
and stubs that will make feedline SWR acceptable, you might
ask Cebik if he considered how to feed the antenna
efficiently. My own opinion is you are buying a big bag of
problems just to get a cleaner pattern on higher bands, and
the clean pattern probably doesn't mean that much. The
antenna your friend has now is too short and will cause big
headaches for matching unless you find a special solution.

My suggestion is you make it into a G5RV, or make it into a
nearly full sized dipole for 40 meters (it can be a little
short) and feed it with ladder line.

73 Tom