Application Note No. 47899.15 10 GHz Slot Antennas
We recently had some slot antennas machined...32-slots for 10.368 GHz per W1GHZ. A friend of ours had a chance to measure the antenna pattern, and as predicted by W1GHZ, there are 4-5 dB nulls in the pattern in the azimuth plane. He added a standard white PVC-pipe radome to it, and the nulls deepened to almost infinite levels at certain angles. Be careful with your 10 GHz radome material...try to use something very thin. I think the bad material was about .250 in thick. Maybe some .062 G-10 would be better. We will post new info here, as it becomes available.
UPDATE... The azimuth pattern (E-plane) of the 32-slot waveguide antenna was found to have -17dB nulls at +/-45deg offset from the broadside direction. W3IP reports that adding wings (around 0.75 inches) to the antenna got the nulls down to -7dB or so. The broadside gain was found to be about the expected 12dBi value. It appears that a uniform omni-directional pattern is in-fact-hard to achieve (as expected). Please email your observations of slot antenna performance to me at firstname.lastname@example.org. Tnx es 73, W3IY
P.S. Here is a very interesting link to some good slot antenna information. It looks like they also had problems with omnidirectionality. PE1CQQ