Hi Moshe,
100 Hz apart is awfully close to separate the signals. I do
a lot of my precompliance testing using a spectrum
analyzer, FET probe, current probe, and H-Field/E-Field
near field probes. With some experience, it is possible
to get "ballpark" ideas as to whether or not the equipment
has a chance of passing at an OATS.
For signals that are are separated by a few hundred Hz or more, it is
often possible to use the "substitution method" to separate the signals.
To do this, you lower the resolution bandwidth of the analyzer
until the signals are distinct emissions on the display. You then
note the exact amplitude and frequency of the emission. The amplitude
will be at a lower level than the correct value due to the lowered
resolution bandwidth.
At that point, you disconnect the antenna from the analyzer input
and connect an RF generator to the input instead. Do this without
changing any of the analyzer settings. The RF signal should be
sinusoidal with no modulation.
Calibrate the RF generator output to the same frequency and amplitude
of the original signal at the reduced resolution bandwidth. Since you now
have a direct sinusoidal signal input, there will be no ambient to contend
with. Now increase the analyzer resolution bandwidth to the normal
measurement setting. The amplitude of the signal will increase by X dB.
Add this increase to the amplitude of the emissions signal from the
EUT that you measured with reduced video bandwidth. This is your
approximate raw reading of the emission.
There is error in that the real noise source may have significant
differences
from the test generator input such as bandwidth, modulation, etc. But
this method is used by registered test labs in high ambient areas and
is definitely a reasonable approximation for precompliance testing.
The downside is that the generator is not cheap.
Hope this helps.
Regards,
tony_fredriks...@netpower.com
----------
From: Moshe Valdman
To: emc-pstc
Subject: EMI testing in a noisy environment
List-Post: emc-pstc@listserv.ieee.org
Date: Thursday, May 16, 1996 4:34PM
Hi everybody,
I am sure many of you encountered the situation of trying to perform
EMI measurements with a spectrum analyzer or receiver in an
environment where "environmental" signals have very close frequencies
to the ones you are trying to measure. This could sometimes be another
unit of the same type of equipment which we are currently testing
which is running with covers open and cables unshielded in an adjacent
room.
Is there a way to separate signals which are about 100Hz apart before
they go into the receiver?
If this could be done I could do all my pre compliance in the lab and
get to the open field lab with a high level of confidence.
I would very much like to hear how others (who probably have much more
experience than me) do it. (There's no way I can get screened anechoic
room).
thanks
moshe valdman
mgr - Safety,EMC,reliability and components engineering
Scitex Israel
