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, [email protected] ---------- From: Moshe Valdman To: emc-pstc Subject: EMI testing in a noisy environment List-Post: [email protected] 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
