MIL-STD 461E does not usually ask we be particularly accurate, merely sufficient.
QUOTE 4.3.1 Measurement tolerances. Unless otherwise stated for a particular measurement, the tolerance shall be as follows: a. Distance: 15% b. Frequency: 12% c. Amplitude, measurement receiver: 12 dB d. Amplitude, measurement system (includes measurement receivers, transducers, cables, and so forth): 13 dB e. Time (waveforms): 15% f. Resistors: 15% e. Capacitors: 120% END QUOTE Also note: QUOTE 4.3.10.4.2 Modulation of susceptibility signals. Susceptibility test signals for CS114 and RS103 shall be pulse modulated (on/off ratio of 40 dB minimum) at a 1 kHz rate with a 50% duty cycle. END QUOTE On the civilian side, some years ago I was at a test lab I won't identify, and asked if we really had 80 percent modulation. Sure, we did; there was one volt of audio going into the SG external moduation jack and the generator was set for 80 percent. After a little griping they let me look on the SA. THEN they hauled out the 'scope. They were into both positive and negative peak clipping. Some things you HAVE to check. (PS: We still didn't pass.) Cortland KA5S > [Original Message] > From: Ken Javor <[email protected]> > To: Untitled <[email protected]> > Date: 8/11/2010 1:03:09 PM > Subject: Re: Calibration supplier for signal generator with pulse modulation > > This gets back to Mr. Walton's differentiation between EMI testing and > calibration laboratories. > > Clearly modulation, measured with an o'scope, can be measured to the > tolerances available from the o'scope. That's plenty good enough. It > strains credulity that there is one brand of scope that would give sloppier > results making it easier to pass a test. > > The essence of what Mr. Walton and I are saying is well-expressed by the > rules for using significant digits: You don't get four digit accuracy as the > result of computations using two digit accuracy inputs. > > EMI limits are quite arbitrary; trying to meet them with accuracy exceeding > the limit placement process is unnecessary. As stated earlier, the only > reason for MU control is on the basic facility chambers/OATS so that there > is some repeatability from facility to facility, and so that it isn't easier > to pass at one facility than another. > > Anything beyond that is superfluous and an unnecessary expense. > > Regarding the example of the cable at 10 GHz. The MIL-STD-461 measurement > system integrity check does indeed check that the cable is properly > accounted for in terms of attenuation. Mr. Demirci is correct that the > signal generator can be expected to be a better match to the cable than is > the actual antenna used. However, to the limits of accuracy required for > EMI testing, the antenna factor calibration takes care of the mismatch > between antenna and cable, since the antenna is calibrated in a 50 Ohm > system (albeit likely using pads between antenna and cable for calibration). > If further accuracy is desired, most manufacturers provide vswr > characteristics for their antennas. > > Fear isn't the issue. The issue is a misapplication of a process where it > isn't required, and the patina of precision and accuracy this dishonestly > bestows on an inherently inaccurate and imprecise discipline. > > Ken Javor > > Phone: (256) 650-5261 > > > > From: Deniz Demirci <[email protected]> > > Date: Wed, 11 Aug 2010 08:35:06 -0700 > > To: Untitled <[email protected]> > > Conversation: Calibration supplier for signal generator with pulse modulation > > Subject: RE: Calibration supplier for signal generator with pulse modulation > > > > Could you define how precise? > > > > What is your acceptance criteria for the modulation dept? > > Let's say for basic 80 % AM modulation. Is your tolerance 0.1 % or > > between 60 % to 90 % is good enough. > > There is more than a dB peak difference in the signal for 60 % AM and 80 > > % AM modulations. You won't realize the difference with a field probe in > > radiated immunity tests with modulated signal. (Simple math; > > 20*Log(1.6/1.8) = -1.02 dB) It seems not very insignificant to me > > > > Another case; > > If you are using 6 meter cable for radiated emission at 10 GHz, your > > cable characterization uncertainty is more than 1.5 dB even with a quite > > high grade cable (Experimental measurements). Check the cable vswr > > figures in their specs measured in ideal conditions. Even a simple cable > > is a significant uncertainty contributor. > > Your measurement antenna is not exactly 50 Ohm at 10 GHz also. Site > > imperfection is another story. Those figures are not accounted for in > > the MIL-STD-461 RE102 verification. Everything seems to be very good > > when you terminate the measurement cable with a signal generator (50 > > ohm) output. I don't agree with this <quote> an ultimate > > "proof-of-the-pudding," </quote> > > > > You can use all your engineering skills for the tests but you have to > > quantify your risk using type B measurement uncertainty analysis and do > > the Type A if you can afford to see where you stand. > > > > As Mr. Gremmen said " MU is not difficult" > > No need to be afraid, > > > > OOO (Own opinions only) > > > > Best regards, > > > > Deniz Demirci > > National Technical Systems (NTS Canada) > > Phone: 403-568-6605 ext 244 > > fax: 403-568-6970 > > email:[email protected] > > web: http://www.ntscorp.com/about/locations - This message is from the IEEE Product Safety Engineering Society emc-pstc discussion list. To post a message to the list, send your e-mail to <[email protected]> All emc-pstc postings are archived and searchable on the web at: http://www.ieeecommunities.org/emc-pstc Graphics (in well-used formats), large files, etc. can be posted to that URL. Website: http://www.ieee-pses.org/ Instructions: http://listserv.ieee.org/request/user-guide.html List rules: http://www.ieee-pses.org/listrules.html For help, send mail to the list administrators: Scott Douglas <[email protected]> Mike Cantwell <[email protected]> For policy questions, send mail to: Jim Bacher: <[email protected]> David Heald: <[email protected]>
