On Mon, 29 Aug 2011 08:51:12 GMT, Clifford Heath wrote:

On Sun, 28 Aug 2011 22:09:00 +0200, Helmut Wabnig wrote:
All started because I could not get the PVC tubing which the many
internet sites propose.

Hi. I foxhunt regularly with the Melbourne ARDF community,
and we have the design of 2m sniffer antennas down to a fine art.

The best material for the shaft is 12mm square or u-channel aluminium.

The common antenna element length and spacing recommended with the
tape-measure antenna is wrong - very wrong. For a start, the lengths
vary by too much (in the hope of a wider effective bandwidth), and
the reflector is too close to the driven, the director too far away.
This causes large rear side lobes, which makes sniffing incredibly
frustrating.

For effective sniffing, you *don't* want maximum F/B ratio, and you
don't want maximum gain - what you want is a cardioid pattern with
just one lobe - the front one. Anything else will give you a bum
steer.

The best pattern we've found so far is the VK3VT beam:
http://www.vk3vt.net/VTbeam.html. Most of the recent ones use a
gamma match instead of the one shown.

We build this using elements made up from 3mm fiberglass rod and
co-ax braid, covered by heatshrink. The element goes through a
short copper tube before the heatshrink goes on.

These elements are sufficiently stiff to keep their shape, but
whippy enough to run with through a tree.

Clifford Heath.



Do you think the matching affects the cardioid or F/B pattern?
So far I have neglected any influence from that.

With the YAGICAD optimizer I could not come to a positive end.
This program finds 100 different "optimums" therefore
I turned to 4nec2 now.
The patterns which YAGICAD calculates look good to me.
When I compare the F/B patterns, of course the 4nec2 has a more
detailed "resolution" showing minor side lobes where YAGICAD
is just approximating.

I wonder how important matching is for a pure
direction finding antenna, methinks rather it should not.
Only the coax cable currents are not considered
in the simulator diagrams, there might be an influence.
Therefore matching is important :-)

Thank you for the vk3vt link.

w.
http://www.qsl.net/oe8uww/