Geoffrey S. Mendelson wrote:
Jim Lux wrote:
They do make a nifty $100-200 or so WiFi spectrum analyzer which has a
calibrated receiver. And, if you have a wifi card with an exernal
antenna, you can do even better.


First of all, a $100 item in the US is $200-$300 here (or more, depending
upon the regulation required to get it in, the rarity, etc). For example,
cheap MP3 players are around 25% more expensive, an iPod is double.

Ok.. so the Wi-Spy is going to be a bit on the pricey side for you.



Second, I have never seen a real meaningfull number from a computer. Whether it
is MacOS, Linux, or Windows XP, they all give some sort of relative number,
for signal strength.

What software are you using to get the signal strength from the WiFi
card? AND, is your WiFi interface one that actually supports scanning
and measuring signal strengths (not all do.. it's less and less common)

However, if you have a card that supports it, you can probably get a
quantitative (but not necessarily calibrated) measurement.


In some cases it has been useful, for example using Kismac, I found that the
signal strength as measured by my WiFi card in my Mac, was in the mid 50's
with a bluetooth dongle on the computer, and in the 60's without one.

No matter what the number really is 50 is about where it gives up completely.

However, it does not really help me. What I need is to be able to place a
device at a specific distance from a 2.4 gHz antenna and read the EIRP.



I can assume that a 100mW signal into the little whip on the back of a router,
produces and EIRP of 100mW, which makes all gain antennas illegal.


That's not a valid assumption. You'd have to look at the FCC
registration data for the device (assuming the configuration is the same
in the US as you have) and see what their analysis is. Gain antennas
aren't necessarily a problem, even under Part 15.

Or I can
make an attempt at measuring it. It does not have to be 100%
accurate, 10% or even 25% off would be good enough.

Your problem is going to be finding a way to calibrate your measurement.
Do you have ANY calibrated signal source around? If so, you can build
a standard gain antenna, and generate a known field to measure with your
measurement device.





For example, I found that placing my meter at about 1 inch from a WiFi router
running "full tilt" produced a reading of .36mv/cm2 on the meter. However the
only real information I can get from that is yes, the router is transmitting
a signal, and yes, I can read it, but no, I have absolutely no clue as to
what it means.

I assume you mean 0.36 mW/cm2... That's probably an average power (with
the averaging done by the meter movement).



I may be asking the impossible, but what I would like to get is a number
for example, if I hold the meter an inch from the antenna, 100mW EIRP is
1 mv/cm2. Then I can do something useful with it, as I already know how
to compensate for distance. :-)



If you hold the meter an inch from the antenna, the meter is going to
alter the performance of the antenna.


You need something with some gain (or better sensitivity), so you can
put the probe farther away (e.g. more than a meter), and still make your
measurement. If your device under test is radiating 100 mW, the
surface area of a 1 meter radius sphere is about 130,000 cm^2, so 8E-4
mW/cm2. (call it -30dBm/cm^2) At 2.4 GHz, a wavelength is 12.5cm, so
a dipole will be half that, or 6cm. The effective aperture will be
around 10-20cm^2 (I can't remember off hand), so a receiver hooked to a
dipole probe needs to measure -20dBm sorts of power levels, which isn't
too bad. There's a bunch of inexpensive power detector ICs around that
have sensitivities spanning this (like the AD 8314.. goes -45dBm to
0dBm, 2.7GHz, $1 each in quantity)

You could probably get a sample or two for free. Build a tiny board that
combines the AD log detector and some sort of display or IR telemetry
and a small battery, with the dipole sticking out.





Geoff.