[MLB-WIRELESS] Antenna Trials

[Richard Talbot]

Well, me and a mate sat down on the weekend to build some omni's.  We used the Collinear 
design posted on the Melbwireless site here (http://melbwireless.dyndns.org/pages.php?page=2.4Ghz+Vertical+Colinear).  Building was reasonably straight forward once we worked out the more than cryptic documentation and sorted through all the measurements.

A few points of note... 
A 6mm screw with 16 threads per inch didn't exist at our hardware shop.  Even if it had, the pitch would be wrong.  There is just no way one could coil a 3/64 brass rod around the scree without the coils touching.
The design looks seriously strange to us.  There is nothing like it in the usual sources, ARRL experimenters guide and the like.  With no maths provided it is almost impossible to verify the theory.

We cut all the parts accurately and measured with verniers.  With soldering complete and attached to a SMA connector, we set about to test it.  Initial testing was done with a Compaq WL100 and pigtail.  We were talking to a WAP11 upstairs.  We couldn't see any appreciable results so we figured it would be easier to test if we walked up the street till we nearly ran out of signal and plugged it in.  Once in position 0m up the street we plugged it in and to our horror saw the SN ratio go down (- ive scale) about 15db.  The card then alternated up and down every second or so on diversity.  The internal antenna presumably doing better than our "5db" omni.

We rechecked all our measurements, construction and integrity of the wiring.  Still no go!  Just to make sure we weren't dreaming we plugged in one of the Galaxy antennas I have previously modified from the same place.  After moving the antenna around a bit we notice a 14-15 db gain on the internal antenna.  Considering we had 5m of RG 213 on the end, one of those terrible Lucent connectors and the Galaxy antenna isn't really designed for ISM frequencies, we thought this was pretty good.  The score thus far Galaxy 1, homebrew omni's 0.

We then set to building another design.  This time a 26 element omni.  The maths we could understand and the design involved simply bending the brass... so this time we thought we might be more successful.  We didn't actually use the full 26 elements, more like 7.  We felt this antenna couldn't be any worse than the last in any case.  Anyway, back we walked to our original testing place and no joy on this attempt either although we didn't drop quite so many db's.  The score now Galaxy 2, homebrew omni's 0.

Having got this far we were a bit discouraged, but we decided to press on and build a Pringles antenna.  This looks promising, although we haven't finished it yet.  The design is a well known ARRL one, used by hams for quite some time.  I'll keep you posted.  

I'd also be interested in hearing from anyone that has managed to successfully build an omni with a verifiable gain....

Cheers
Richard.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.wireless.org.au/pipermail/melbwireless/attachments/20020121/06ce2219/attachment.html>