As Robert mentioned, this subject has been discussed in the last few years. A couple of areas not touched upon by recent replies has to do with the type of antenna and mutual coupling factors.
A well designed antenna will be balanced, i.e., it will be geometry independent. Looking at the factors for our 3110b, they look very nearly the same for horizontal and vertical polarization as well as at 3 meters and 10 meters. This suggests we should save the money for this model antenna and have 1 factor verified (10 meter horizontal) instead of having 4 factors verified. A broadband antenna may not have mutual coupling at 3 meters, but a tuned dipole will. Of course there is usually not enough space in a chamber to use tuned dipoles. We have had better luck on the OATS with dipoles than with broadband antennas -- more accurate and more repeatable, but longer execution time. An interesting point if you use tuned dipoles at your OATS, it has been shown by a study (by NIST?) that for all tuned dipoles of the "FCC reference" variety (EMCO 3121C?) that the variation from the factors in the standard (ANSI C63.5) were on the order of +/- .5 dB (for a universe of ~50 antennas?). One could make an argument to use the factors in the standard instead of using the factors provided by a cal lab. The dipoles provide you with an accurate OATS NSA. Then its time to correlate the results by dipole to results by broadband antenna, then go on to the chamber. Is it worth it? Depends on the answers you are getting and how much time you have. Just my $.02 worth, Don Umbdenstock Sensormatic > ---------- > From: Robert Bonsen[SMTP:[email protected]] > Reply To: Robert Bonsen > Sent: Wednesday, March 15, 2000 1:47 PM > To: [email protected]; [email protected]; > [email protected] > Subject: Re: Antenna factors to be used for NSA measurement > > > This is a discussion which starts up at least once every year. There are > some issues with measuring NSA which are not very well understood > including > antenna factors. > > First of all, how you test and which factors to use should not be a matter > of opinion. Unfortunately, the standards do not completely specify the > requirements on the antenna factors which enhances confusion and > introduces > room for data manipulation. Having said that, there are some simple > guidelines which are based on physics, literature and practical > experience. > > 1. There is a difference between measuring EUTs and performing site > attenuation measurements. NSA measurements are solely used for verifying > the performance of the site. It uses a simple antenna substitution > technique, in which the site under test is verified against the > performance > of the site on which the antenna factors are calibrated. Hence, the better > the cal site on which the antennas are calibrated resembles a "perfect > OATS" (i.e. an OATS with an infinitely large ground plane), the better > quality the NSA measurement will be. As far as measurement method goes, a > dual antenna factor calibration suffices if the same antennas will be used > in the NSA measurement, there is no accuracy advantage when using a > three-antenna method in this case. > > 2. Measurement distance. There are substantial difference in the antenna > factors (and site attenuation) values at various range distances. In order > to perform a correct NSA measurement, use the same range distance for the > AF calibration as for the NSA measurement. The same goes for the EUT > measurements, always use the same range distance. > > 3. Geometry. Although some OATSes and a few semi-anechoic chambers will > pass NSA with using horizontal antenna factors only, this is not a correct > measurement method. Since we're comparing the site-under-test to the AF > measurement OATS, any change in the setup which results in a change in the > physics of the AF or NSA measurement will introduce an artificial > systematic error into the measurement. Both literature and experience has > shown that AF measured at different geometries, with different > polarizations and/or different source antenna heights), produce different > antenna factors. Variations of up to 3 dB and sometimes more are to be > expected for the low frequency regions. These variations will result in > added errors in any site attenuation measurement in which the correct > corresponding antenna factor is not used. In a lot of cases, this will > bring a perfectly ok chamber or OATS out due to the artificial measurement > error. > > 4. Accuracy. Antenna factors provided by the manufacturer or measured by a > cal lab are typically not of sufficient accuracy (with a few exceptions). > Accuracy in the AF measurement is extremely important since the NSA > measurement does not provide for a lot of uncertainty margin for the AF. > Typically, AFs are measured with a 2 dB uncertainty at best, even when > higher accuracies are claimed. Cable layout, padding, equipment, etc. are > extremely important. Also, AFs are typically only measured at one height > at > one polarization, which is not sufficient for accurate chamber cals. > > I recently was asked to calibrate a chamber using factors which were > claimed to be better than +/- 1dB, but the frequency steps were so large > that a small bump in the AF was completely missed, resulting in a more > than > +/-2dB uncertainty. Had the lab paid attention to setup and the > peculiarities of this antenna, the AF measurement would have been fine, > and > we would have saved considerable time and money. Since most labs have a > substantial margin for antenna factor error (up to 3 dB) in their > uncertainty budget for EUT measurements, accuracy and geometry does not > become much of an issue. However, in NSA measurements we're looking at > substantially smaller margins and all of the aforementioned issues become > important. > > On a final note, free-space factors are not an alternative. Simple physics > dictates the presence of the ground plane (which is not present in > free-space measured AF) adds a variable to the equation which in turn may > add systematic errors to your measurement. And this has been demonstrated > in literature. Free-space factors are for free-space measurement > facilities > (fully anechoic chambers). Using free-space factors (or single geometry > factors) can only be allowed if the additional uncertainty is included in > the error budget of the NSA measurements. And this is not the case with > the > current standards (ANSI C63.4-1992, CISPR22). > > Hope this helps. For a little more detail, check my web site. > > Regards, > -Robert > > Robert Bonsen > Principal Consultant > Orion Scientific > email: [email protected] > URL: http://www.orionscientific.com > phone: (512) 347 7393; FAX: (512) 328 9240 > > > ------------------------------------------- > This message is from the IEEE EMC Society Product Safety > Technical Committee emc-pstc discussion list. > > To cancel your subscription, send mail to: > [email protected] > with the single line: > unsubscribe emc-pstc > > For help, send mail to the list administrators: > Jim Bacher: [email protected] > Michael Garretson: [email protected] > > For policy questions, send mail to: > Richard Nute: [email protected] > > ------------------------------------------- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. To cancel your subscription, send mail to: [email protected] with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Jim Bacher: [email protected] Michael Garretson: [email protected] For policy questions, send mail to: Richard Nute: [email protected]
