antenna-theory.com

[lslarry86]

I work in RFID, so I see a lot of patch antennas for 902-928 ISM band. Some of them seem to radiate from the antenna cable. That is, say I have a 3 foot RG58 cable between the RFID reader and the antenna. RFID works by transmitting a carrier and listening for changes in reflected power. That should only work within the atnenna pattern. But with some antennas I can run the RFID inlay (antenna + chip) down the coax and get hits on the reader. When it happens, it's usually just at 6 inch or so intervals, so somehow related to the half wavelength. Sometimes the whole cable hits for me. Some of the antennas have quite good return loss, like -20 dB. Some have impedance off, like 35 or 65 ohms.

What is it about an antenna that makes this effect happen?

[admin]

Great Question.

So imagine you have a dipole antenna that is small, say a quarter wavelength (so 31mm at 2.4GHz or whatever). If I want to measure this antenna, I put a coaxial cable up to it with the ground connected to one arm and the feed to the other arm.

What I've done there is actually extended the ground arm of the antenna. There is no way for the antenna to know which part is the antenna dipole ground and which is the outer shield of the coaxial cable - hence it becomes part of the antenna.

So by plugging a cable into an antenna you're naturally making the antenna larger, in your case for a patch, you're increasing the ground plane size. This effect is most signfiicant when the antenna is <= half wavelength in size.

It is also the reason antenna measurements often require 'baluns' or chokes on the feed line, such as this guy: http://www.antenna-theory.com/definitions/bazooka.php

[lslarry86]

That makes sense. And the bazooka balun is just neat. I'll try it when I get back to that project.
Thanks very much!