Hi Johann

There is no doubt a lot of info on internet articles you could wade.
through. Since you have a practical requirement though I'll try and answer.

The length and diameter of an antenna is related to the frequency. In
rough terms a half wave in free space on 2.4GHz is calculated as
300/2400MHz/2 metres or about 60mm. A half wave dipole antenna tends to
be shorter than this by a percentage between about 85% and 98%. The
variation being a function of the material diameter vs its length (ie
the fatter the antenna element the shorter its length) If you etch the
design into a PCB the dielectric around it (PCB insulative material)
will lower that percentage more. I don't know how much off hand.

Within limits however you can match any transmitter to any simple
antenna length. Small variations and errors in length (say 20%) wont
deter very much from the performance provided you have done a good job
coupling the transmitter to it. If however you use a simple element of
more than a half wavelength you can get an undesirable pattern of
radiation from it. Gain antennas like yagi's will suffer badly from this
much variation though.

Coupling/tuning in this sense means getting a maximum power transfer
from the transmitter to the antenna.

Antennas for longer distances on 2.4GHz are almost always devices built
for gain. They can use a cluster of dipoles phased to concentrate the
radiation into a narrower pattern, hence the received signal at the
other end is stronger than from a simple dipole. The downside is of
course if either end is moving/rotating in respect to the other. Loss of
signal results. (Think of a searchlight)

Yes etching an antenna on a PCB is a good way to do it. "Patch antennas"
might be the term to look for, as well as "slot antennas" with PCB as
part of the search criteria.

One of the conditions you didn't specify is whether the ends of the link
are fixed or moving. If they are fixed you can get away with lower power
and gain antennas. If they are moving you probably want an antenna that
radiates equally in all directions. This is actually difficult to do as
there is always variation in different angles from the antenna. A dipole
for instance radiates best at 90 degrees to the element. There is
however virtually no radiation directly off the ends. This doesn't make
the system unusable though. Some of the signal will get through via
reflection off other objects and a very useful amount will come off at
(say) 10% from off the ends. If you have enough power margin it will
still work. It will be something like (say) 5m range at almost any
attitude and 50m range if you hold the antenna just right!

Smallest antenna? A slot would be maybe 60mm long and 20mm wide if
etched to a PCB. A quad type element etched on a PCB may me 30mm x 30mm.
You could also make a 1/4 wave groundplane. That's a 30mm or so wire
perpendicular to a ground plane area (eg PCB) a minimum of 60mm
diameter. If you can suffer some efficiency loss you can etch a loaded
dipole pattern on a PCB maybe 15 x 10mm but it would be a good idea to
test whether the distance loss will be okay in your environment.

If you hand make the antenna the coax etc attachment becomes a large
lumped reactance/resistive load that has to be allowed for when
adjusting for best coupling. Likewise a etched PCB antenna will be
affected by components mounted on the reverse side as well as the (say)
plastic etc box it is mounted in. The rule is to tune it with everything
attached.

For the purists out there I have watered this down some and tried to
keep heavy theory out of it. I hope however you'll still find it useful.

Cheers Bob W5/VK2YQA

wrote:


What I need is the smallest antenna possible!

Thank you for your insight.
Johann Blake