Debugging a helical antenna installation

[Alan Wilkie]

My father and I have just finished building a pair of helical antennae
as per the design on Jason Hecker's web site, but we're having some
problems getting a link going.  Does anyone have experience debugging
these things?

We're using Compaq WL200 cards which work fine when they're in the same
house with their original aerials plugged in.  At the moment we've just
soldered an extra piece of coax straight onto the board to connect the
external antenna.  The cable runs we're using for testing are fairly
short (5 or 6 metres) so I don't think cable losses are going to be a
problem.

There distance is about 500m and there is only one tree that prevents us
from getting absolute LOS.  I did some calculations on Fresnel zones as
well and there is nothing in the way that should affect us.  We can get
perfect LOS if it becomes necessary, it's just existing mast placement
that's made us go this way.

Needless to say, the reason I'm writing this is that we've had no
success getting the cards to associate, even when they're 10m apart. 
It's as if there's no signal at all getting into/out of the helicals. 
We've checked all the obvious things and there are no shorts, the right
pins are connected to the right places, etc.

What are the critical factors that affect these things?  Accuracy of the
construction?  Connectors?  Soldering directly to the board?  How do you
test them (i.e. who's likely to have a 2.4GHz signal generator and
spectrum analyser)?  Do they saturate if you have them too close
together?

TIA,

Alan Wilkie