Richard:

Thanks, I think you have at least one problem I created in check. I forgot
to place ferrite beads over the coax, at the antenna, I guess I was thinking
the gamma was some sort of increased isolation. It is now so apparent--now
that you mentioned it...

Regards

--
Hay, if'n ya'll cun't konstructivly partecipete in this har disscusion, haw
aboot speel-checkin it fer me?


"Richard Clark" wrote in message
...
On Sun, 3 Apr 2005 20:06:45 -0700, "John Smith"
wrote:

Lenght of coax from rf source to ant seems critical, when coax is
shortened
or lengthened, gamma must be adjusted (but is always around 3 ft)

Hi Brett,

This is a classic symptom of feedline interaction with the feedpoint
Impedance. Basically, the exterior of the cable is acting as an
uncontrolled tuning element that is bridging your attempt to tune the
antenna. To correct this one indication, you need to choke the
feedline. This is accomplished by one of several methods.

We frequently recommend what is called a "Current BalUn" which is a
series of ferrite beads strung onto the coax near the feed point.
Another method is to simply coil the coax with half a dozen turns of
about 6 inches diameter. Either method will decouple the feedline
from the feedpoint. There's more to be said, but that can be
discussed at another time, or you can research that in the archives
using the keywords in quotes above. This is successful with Low-Z
loads because the choke is Hi-Z in comparison - your drive point
choice presents a problem in that regard.

That problem is that a halfwave is extremely sensitive to environment,
especially in the face of unchoked lines. This is a fact of life in
that for a Hi-Z load, nearly everything nearby looks like a short
circuit unless you can hoist the antenna out of the way. The choking
action will have to be extremely good to overcome this (or you will
have to insure that the external coax length is also halfwave in
length - this is one of those CB antenna/coax issues that rarely gets
discussed and becomes lore instead of learning). Hi-Z loads are
usually matched with Hi-Z circuits through voltage matching, the Gamma
match is more a Low-Z device working as a current transformer. Its
standard application, nearer the middle of the halfwave antenna, would
be more suitable.

To achieve this, isolate the bottom end of the halfwave (mount it on
an insulator, the details which follow will be challenging). Make
sure you can snake the coax up inside the antenna (this presumes it is
tubular of sufficient diameter to pass cable up inside). Break out
the cable halfway up (care must be given that this does not
substantially weaken the whole of the antenna). Build the gamma match
(external of course) here and drive it against the radiator body with
one lead of the coax going to the gamma structure, the other side
going to the break out hole.

All in all, a SWR of 1.7 is not shabby for a first pass approximation
for a physical solution. If you decide to stick with your current
implementation, you might try making the gamma tube larger than the
diameter of the radiator (capacitance stays the same). We are
departing from the regime of nutshell math.

By this (classical design), the work revealed in full sized, folded
dipoles (and monopoles) teaches that the size ratio of the two
elements (in this case approximated by the radiator and the gamma
element), and their proximity, yield a step-up or step-down
relationship in drive point Z (depending on which, larger or smaller,
is being driven). This is like conventional transformer winding
ratios, except in RF it is embodied in diameters and separation with
complex results (not as simple as counting turns and being done with
it). This is further complicated by the structure not being a classic
folded design - I am merely extending the metaphor, returning this to
nutshell theory.

This means it is up to you to close the gap through experimentation,
observation, and correlation. There may be a limit in what you can
achieve, but you seem close enough to vary a few parameters to see if
there is any progress.

73's
Richard Clark, KB7QHC