Maker Pro
Maker Pro

GigaAnt Mica 2.4 GHz SMT antenna

S

Steve

Jan 1, 1970
0
Hi all, new to this group, looks interesting. I'm a newbie to
2.4GHz design, now working on a board with an Airwave 2.4GHz
video/audio Tx module which I hope to couple to the GigaAnt SMT
antenna. I'm a bit challenged by RF stuff! This SMT antenna has
feed and ground pads at the business end, close together. The RF is
applied here. I'm confused to find that applying a multimeter to these
input pads on the unsoldered device (3 examples tested) gives a short.
I expected high DC resistance. Can anyone explain this for me?

Steve
 
J

Joel Kolstad

Jan 1, 1970
0
Steve said:
Hi all, new to this group, looks interesting. I'm a newbie to
2.4GHz design, now working on a board with an Airwave 2.4GHz
video/audio Tx module which I hope to couple to the GigaAnt SMT
antenna. I'm a bit challenged by RF stuff! This SMT antenna has
feed and ground pads at the business end, close together. The RF is
applied here. I'm confused to find that applying a multimeter to these
input pads on the unsoldered device (3 examples tested) gives a short.
I expected high DC resistance. Can anyone explain this for me?

Hmm... I'm guessing you've never studied transmission lines, right?

You have to start thinking of the copper on that antenna just as a medium
which guides an oscillating wave. If you send a 2.4GHz wave down a line
that's long enough such that its shifted 90 degrees in phase by the time it
hits the end (the ground plane), it'll bounce back and be 180 degrees out of
phase with the input when it makes it back to the feed. However, there was an
additional 180 degree phase shift to the 'bounce' because -- to make the end
'grounded' -- the reflected voltage must be the opposite (180 degree phase
shift) of the incident voltage, right? Therefore, back at the feed, the
reflected voltage is really 360 degree outs of phase with the incident
voltage -- the same as 0 degrees or _in_-phase -- and the total current is
zero (since the reflected signal has the same magnitude of current as the
incident signal). This means that, at 2.4GHz, you're looking at an open
circuit!

In other words, there are many devices that are 'short circuirts' at DC yet
which do 'something interesting' at higher frequencies. (A real antenna can't
just 'appear' to be a pure open circuit, however, because that would imply
that no energy is transmitted... something like a half-wave dipole, at
resonance, appears as a 77 ohm load, for instance. The energy that goes into
that 'load' is what's being transmitted out into space.) The key is that the
physical dimensions of the device typically have to be 'significant' (perhaps
10% or more) compared to the wavelength of the signal you're using. At
2.4GHz, the free-space wavelength is 125mm; in circuit board material it'll
typically somewhere between 1.5 and 3 times that. (People do build antennas
much smaller than a 'significant' fraction of a wavelength, but it's
difficult -- physically bounded, as a matter of fact -- to get high efficiency
out of such designs. These designed are deemed to be 'electrically small.')

Real antennas can be quite complicated things that realistically few people
could analyze much behind a 'gut feel' without the use of computer simulation;
this is particularly true with antennas that are attempting to be useful over
a wide bandwidth and yet still be compact. On the other hand, when this isn't
the case, you see a lot of the very similar designs over and over again, with
only small variations -- dipoles ('regular', sleeve, folded, etc.), loops,
Yagis, log-periodic designs, etc.

This is a really gross simplfication (that can get you into trouble later on
:) ), but 'loop'-type antennas are close cousins to inductors (which, you'll
recall, have little DC resistance but plenty of impedance at finite
frequencies) that happen to radiate whereas 'dipole'-type antennas are close
cousins to capacitors which do so. Intuitively most people find it a little
easier to view how a simple coil of wire ends up being an antenna; for dipoles
the usual progression is to ask you to to think of a piece of twin-lead cable
where the two leads are flared out and eventually become parallel to one
another.

You might want to download Chipman's "Transmission Lines" book from
http://oregonstate.edu/~kolstadj/ . For more than you ever wanted to know
about antennas, there's John D. Kraus's "Antennas For All Applications," or a
Google search to various amateur radio-related web sites for a more
introductory look. Oh, and visit rec.radio.amateur.antenna!

---Joel Kolstad
 
S

Steve

Jan 1, 1970
0
Thanks indeed Joel for that excellent clarification!


Hmm... I'm guessing you've never studied transmission lines, right?

Right indeed. I'm an amateur, and RF is pretty foreign territory.
And this freq is way higher than anything I've handled before. The
minute you mentioned "transmission line" all became much clearer
though, I need to think more deeply when dealing with this sort of
wavelength!
incident signal). This means that, at 2.4GHz, you're looking at an open
circuit!

Makes sense. A bit frustrating when you are soldering the thing by
"kitchen reflow", no easy way to check the integrity of the soldering
job I guess.
much smaller than a 'significant' fraction of a wavelength, but it's
difficult -- physically bounded, as a matter of fact -- to get high efficiency
out of such designs. These designed are deemed to be 'electrically small.')

Yes the Giga thing is one of those I think.
You might want to download Chipman's "Transmission Lines" book from

Great, thanks for the URL. Will do some reading.

Steve
 
Top