Thank you Robert for a well detailed response. Even though this issue seems to come up every year, more clarification is the outcome. Your response clarifies issues that were not as fully addressed (or at least not fully understood by me) in previous threads. I'll call this "Chapter Three" of my ever ongoing "NSA studies".
Don Umbdenstock . > ---------- > From: Robert Bonsen[SMTP:rbon...@orionscientific.com] > Reply To: Robert Bonsen > Sent: Wednesday, March 15, 2000 7:02 PM > To: emc-p...@majordomo.ieee.org > Subject: RE: Antenna factors to be used for NSA measurement > > > > > >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. > > > IMHO this is not completely accurate. There is not such thing as a > geometry > independent antenna factor, bad antenna balance/design only has an > aggrevating influence on a physical inevitability. Check Zhong Chen's and > Michael Foegelle's article in the 1998 IEEE EMC Conference proceedings, > entitled "A numerical investigation of ground plane effects on biconical > antenna factor". In that paper they prove that even for an ideal antenna, > with a perfect balance, the physics/geometry of the setup (due to the > presence of the ground plane which alters the incident plane wave as well > as coupling between the antenna and its image under the ground plane) > inherently result in differences in antenna factors between polarizations, > antenna heights and test distances. Actual antenna factor measurements as > well as basic antenna physics back up this theory. By the way, Zhong and > Michael work for ETS, manufacturer of the 3110B. > > Measured and predicted data on 3110Bs and other antennas (BiLog, other > biconicals) show that for 2m transmit antenna height, which is pretty much > the "standard" Tx antenna height cal labs test at, the vertical and > horizontal factors are usually very close. However, once you lower the > antenna transmit height to 1m (vertical and horizontal) and 1.5m (vertical > only), substantial differences exist not only between antenna factors at > different polarizations but also between the factors at different antenna > heights for the same polarization. These differences can be several dBs, > which is introduced as a measurement error in your NSA measurement if you > use a factor which is not measured at the same height you're measuring > your > NSA at. This effect is more noticeable at shorter range length. Hence, 10m > factors tend to be closer together than 3m factors. However, chamber and > OATS performance at 10m range length is more critical than at 3m so this > tends to even out. > > For chamber NSA measurements, the aforementioned variations in AF with > height and polarization prove to be sufficient to bring a chamber out in a > lot of cases, or make its performance look worse than it really is. This > becomes a money issue when chamber manufacturers sign up for > better-than-4-dB performance. For OATSes, there is a substantial > performance margin so antenna factor error, although it has a negative > effect on measurement accuracy, will not bring the OATS out of spec. I've > been involved in OATS calibrations in which the performance margin was not > sufficient to bring the OATS in, whereas when proper antenna cal factors > were used the OATS passed well within spec. > > FYI, I used to work for ETS as a chamber design engineer before becoming > an > independent consultant. As such I've done quite a few antenna calibrations > on 3110Bs and other bicons and combination bicon-logperiodic to be used > for > NSA calibrations, geometry specific for different test distances. > Experience shows that at the low end, up to about 200 MHz, it can be > almost > mandatory to have geometry specific antenna factors because of the > potential substantial measurement error introduced by using "wrong" > antenna > factors. Because of higher directivity, antenna factor variations are > reduced at higher frequencies (where log-periodics are used), and a single > antenna factor typically suffices. > > Unfortunately, not all organizations and experts were aware of this issue, > or ignored it. That's why ANSI C63.5-1998 is written the way it is. The > problem of antenna factor variations with different geometries is ignored > because for EUT measurements it does not pose an immediate problem. > However > it is an issue with normalized site attenunation measurements. More work > is > currently being done in this area and the issues are being addressed. > > -Robert > > > > Robert Bonsen > Principal Consultant > Orion Scientific > email: rbon...@orionscientific.com > 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: > majord...@ieee.org > with the single line: > unsubscribe emc-pstc > > For help, send mail to the list administrators: > Jim Bacher: jim_bac...@mail.monarch.com > Michael Garretson: pstc_ad...@garretson.org > > For policy questions, send mail to: > Richard Nute: ri...@ieee.org > > ------------------------------------------- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Jim Bacher: jim_bac...@mail.monarch.com Michael Garretson: pstc_ad...@garretson.org For policy questions, send mail to: Richard Nute: ri...@ieee.org
