Muriel, If an ambient (interfering) signal is strong enough, you will be unable to make accurate readings of emissions on or near its frequency. "Subtracting" the ambient's trace merely moves the place where it is displayed so that it LOOKS like it's gone. It does not cancel out the received signal. The plot looks better -- but you still can't measure near the ambient.
Another problem with strong ambients is, from my experience, that mixing products inside your equipment may give rise to what look like product emissions. This is easily detected, but eliminating this problem calls for a site (at least an area) specific _system_ design when setting up the equipment site, and this is rarely, of ever done. Determining if an emission is "real" may be done by switching in an attenuator (perhaps 10 dB) and seeing if the "emission" being measured decreases by 10 dB; if it is a product, it will go down by 20 or more dB and you can ignore it. However, if it's strong enough to be a concern,it can cover up a real emission. If mixing products are a problem, it is sometimes possible to eliminate them by filtering the input to reduce their sources. A good high-pass filter will work wonders above 120MHz or so if you are bothered by strong FM broadcast signals. (In order to make this a non-issue from the beginning,one must know what the strong signals are before putting the system together, and insure that the analyzer or receiver, preamplifier and antenna, and so on, are designed and configured to be able to handle the strong ambients without overload or mixing. Although it is preached everywhere, a low-gain amplifier at the antenna end of a site is a good way to begin,but rarely done. Most folks, it seems, opt for inexpensive coax and a high-gain, broadband preamplifier, which invites such problems. However, that is properly another thread.) To truly cancel out an ambient -- this, from an article that appeared long ago -- one can set up an antenna some distance away, and combine its output with that from the measurement site in such a way that the ambient -- but not the emissions being measured -- are equal in amplitude and 180 degrees out of phase. This does work, and is used by the military and ham operators. I began to build a setup of this type at my former employer, but moved on before I got it done. You may easily experiment with this by using two antennas, connected by approximately equal length feedlines to a "T" fitting at your analyzer input. One is the normal measurement antenna; the other is a similar antenna on a tripod, so it may be moved around. By moving the antenna and reorienting it, amplitude and phase may be adjusted so that an ambient is cancelled. However, if the cancellation antenna is reasonably close to the site, it affects accuracy of measurements. It should be at least ten times further away from the EUT than the measurement antenna. You can see that kind of experimental setup is impractical for normal use! A real cancellation controller would have phase and level controls, probably pre-marked for particular sources at each site where one was used. Good hunting! Cortland Richmond ====================== Original Message Follows ==================== Muriel Bittencourt de Liz wrote: Hello all, I'm doing my measurements of conducted emissions using a Spectrum Analyser, but my site isn't shielded, i.e. some interferences ( radio stations ) appear at the screen of the SA. So, i'm doing this: i record the signal with the product tested at "off" and when i turn the product "on", i subtract the signal of "on" minus the signal of "off". this means that i subtract the "ambient noise" from the noise being generated by the product itself. What i really want to know is: is this procedure correct? am i doing a nonsense thing? --------- This message is coming from the emc-pstc discussion list. To cancel your subscription, send mail to majord...@ieee.org with the single line: "unsubscribe emc-pstc" (without the quotes). For help, send mail to ed.pr...@cubic.com, j...@gwmail.monarch.com, ri...@sdd.hp.com, or roger.volgst...@compaq.com (the list administrators).