John - KD5YI wrote:

Actually, on elevated antennas (as in the usual VHF setup), just two
quarter-wave radials 180 degrees apart is almost indistinguishable from
4 or more radials. EZNEC shows very little change in terminal impedance
and pattern by removing two radials from a 4 radial ground plane.

I once used copper tape on a window to make a ground plane vertical like
that for 70cm. It worked very well.

George Brown, the inventor of the ground plane antenna, found that only
two radials were necessary. But when his company went to sell it, the
marketing department decided that no one would buy a two-radial ground
plane antenna in the belief that it would be omnidirectional. So they
added two more to make it "look" more omnidirectional. The four-radial
ground plane persists to this day.

Just a few weeks ago, I designed what amounted to a two-radial ground
plane antenna as part of a consulting job. It was made from copper tape
on a Duroid dielectric material, a lot like the window antenna John
described. An omnidirectional pattern was a requirement, and I was
concerned that either the flatness of the tape or the presence of the
dielectric might have some impact on the circularity of the pattern. So
I had it tested at a local lab. It was the most circular pattern they'd
ever seen, having about 1 dB maximum difference between any two directions.

Roy Lewallen, W7EL