Nug wrote:
"I am building an rf transmitter for a short range data link at 433 MHz
and am almost done, but I would like to understand better exactly what I
am seeing with regard to antenna performance."

Your wavelength is about 0.69 meter or about 2.3 feet. In antennas
everything depends on wavelength. If you use a transmitter housing as
the ground plane for your antenna, it needs to be a sizeable part of a
wavelength or the salient part of your antenna must be longer to
compensate for the small ground plane.

If you had an infinite ground plane, a 1/2-wave wire perpendicular to it
would produce up to 50% more volts per meter field strength than a
1/4-wave wire perpendicular to the ground plane. It`s not something for
nothing. Total radiation is the same in both cases. More of the
radiation is perpendicular to the wire in the 1/2-wave and less goes off
at some other angle to the wire.

50% more field in some particular direction is realy not very
significant in most cases, and there are other consequences of using a
1/2-wave wire instead of a 1/4-wave wire.

An end-driven 1/2-wave wire presents a very high impedance. It is
equivalent to a parallel-resonant circuit. It would match a direct
connection to a parallel resonant tank circuit perhaps.

An end-driven 1/4-wave wire presents a very low impedance when worked
against a ground plane, maybe about 30 ohms.

How well you are able to radiate a signal from a wire is likely to
depend on how well it is matched to the transmitter and less about the
bends in the wire. In any case the complete antenna must be resonant to
eliminate reactance which opposes the signal`s entry into the wire.

For a small transmitter operating at a very short wavelength, the size
of the antenna is not onerous and it would be possible to use a
center-fed 1/2-wave antenna. Each half would be just a little over a
half foot in length. Drivepoint impedance is in the 70-ohm range.
Another possibility is a full-wave loop, about 2.3 feet in perimeter
with a drivepoint impedance of about 120 ohms.

Performance of all the suggestions is probably about the same. You can
find the best by trying them.

Best regards, Richard Harrison, KB5WZI

Thanks for all your responses, good stuff here.

Rich, I am using TX modules from Liapac which specify a 50 ohm load.

So putting it all together (I think) by bending the antenna (both 1/4
and 1/2) I am messing with its impedance and subsequent matching? So I
would be better off with a straight 1/4 over ground plane if I had the
choice.

From reading I've done I think that a bend from vertical to horizontal
(closer to ground) would be decreasing the antenna impedance, does that
sound correct? Would I benefit from adjusting (extending) the antenna
length slightly (I don't think so because it would no longer be at
wavelength, but I am not sure?).

Ken said ... "If I've read what you wrote correctly, the antenna spends
more of its length parallel to the surface of the PCB than it does
running 90 degrees away from it"... Actually it's about half and half
(feed up 15mm , loop around the case, up - {clear of case} 130mm).
Note the horizontal loop does not cross over itself.

Sort of like this (~~ is horizontal loop) side view.

|
|
|
~~~~
|
feed

I understand that more of the radiation is perpendicular to the wire in
the 1/2-wave (than 1/4 wave), broadly speaking how will the bend('s) in
the aerial effect radiation pattern (I understand this is tough to
answer)? I can't fit a full loop inside the case.

I have no rf test equipment so can only use trial and error, thanks
again to all who responded. As I said it actually works fine it's just
I don't like not understanding the reason its working, and would like
to make any small changes that may improve performance.

Thanks Again