On Wed, 15 Aug 2007 16:12:29 -0500, "amdx" wrote:

Now the question is: Is the wifi adapter output really 50 ohms? If it's
not 50 ohms
then I need not attempt impedance matching with 1/4 Lambda length coax.

Hi Mike,

There are very, very, very few antenna designs that arrive out of the
box with 50 Ohms Z. You usually have to manipulate something (feed
point or linear loading, or coupling between elements or other
radiators) to get it.

The good thing about microwaves is that you can build and test so
quickly - and throw away what doesn't work. This isn't like you are
making jewelry after all.

The scale of construction is one you can get your arms around at the
kitchen table. If someone reliable built one and documented it enough
(both of these provisos are, admittedly, built on a LOT of trust);
then you don't need any instrumentation at all.

I searched for several hours looking for specs on the RF chip in the wifi
adapter without any success.

It sometimes takes some deep digging.

Regarding the spacing; The literature I've seen shows spacing of
0.2, 0.3 or 0.4 Lambda before grating lobes form. But the Patch is 0.5
Lambda wide so you can't get them closer than 0.5 Lambda.
Here's one source;
http://www.orbanmicrowave.com/The_Ba...nna_Arrays.pdf

Pretty tedious stuff, that. EZNEC can do it all in real time - even
the free version is sufficient for your purposes. For one, don't
confuse matching with the radiation lobes - those issues are separable
and you can ignore matching to examine lobes.

At this point I have a working Patch with good signal strength, so now I'm
just trying to see if I can improve on it,

Give yourself the benefit of the doubt and just move forward - the
rest is analysis paralysis.

73's
Richard Clark, KB7QHC