Dear Muriel Paul, Robert, Bob and Group You are all right about the standards requiring 80-100 cm etc.. But what is was saying was "what I think is essential", not what is normative. I got carried away by my opinions and experience and forgot what's normative. Sorry to Muriel, because what she wanted is possibly testing according to the standards and not about how I personally think the standard should be modified. The standard however is not so good for establishing the emission. The mains lead and the manipulations prescribed are too much a factor reducing reproducibility. Unless someone requires me to test according to the standard "á la lettre", I use 50 cm of mains lead. The LISN goes ON the conductive wall, which is approx 40 cm away. I can tell you , and promise all of you that I got much better reproducible results this way, avoiding any false curves in the mains wire. Collegues of mine struggled with 6-9 dB of difference depending on how the excess length of mains wire is folded (or not). Especially at higher frequencies (30 Mhz) where 1 meter of cord comes close(r) to a 1/4 of lambda, things really change. The shorter the better is what counts here. The wire does not just behave inductive anymore at this length. The same goes for the ground wire, but it's effects are smaller as it is shunted by the EUT's capacity to ground) In my opinion the test height (and limits) must be brought back reducing test hight to 10 cm as in EN 61000-4-6 using horizontal ground plane. This also increases reroducibility with large EUTS standing up. The CDN concept has proven reliability upto 200 Mhz, so it must do well upto 30 Mhz ! To be able to maintain distance of 50 cm to large EUT having "high"or distant power connections, a vertical groundplane should also be available. This is no problem in a shielded room in generally.
Some people say that one should just let the mains wire "hang-out" as it is, as in real life the mains wire extends also into the room the length it has. Ok but why cut it off at the LISN, in real life mains wire extends into the wall, far beyond any point we can place a LISN at and radiates freely ! Best would be to integrate the classic LISN within the CDN concept as found for EN61000-4-6 which is basically the same network as the 50 uH V-network in LISN, but standardized at 150 Ohms level being the average characteristic impedance of mains wires (and other wires)Pff what a phrase). Some forces are trying to converge the two, and want to modify the conducted emission level test to 150 Ohms. Much to investigate , especially the correlation between former and newer test results. Regards, Gert Gremmen, (Ing) ce-test, qualified testing =============================================== Web presence http://www.cetest.nl CE-shop http://www.cetest.nl/ce_shop.htm /-/ Compliance testing is our core business /-/ =============================================== >>-----Original Message----- >>From: Paul McCoy [mailto:pmc...@lsr.com] >>Sent: Thursday, February 24, 2000 2:48 PM >>To: Gert Gremmen >>Cc: Robert Macy; Muriel Bittencourt de Liz; Lista de EMC da IEEE >>Subject: Re: conducted emissions test setup >> >> >>Hello all, >> I will take a stab at Gert's question and offer a correction to the >>statement he made about the mains lead: >> The mains lead first; It should be at least 80 cm long >>because the LISN >>needs to be 80 cm from the closest portion of the EUT per the >>standard setup of >>CISPR 16-2 (which is replicated in 11, 14, 22, et. al.). If there is "one" >>specific length of mains cord associated with an EUT it should be used and >>lengths in excess of 1 meter should be bundled to 1 meter. If the >>EUT can come >>with various lengths of mains cord use 1 meter. Whether you use a ground >>connection to the EUT or not is determined by whether it is >>normally grounded or >>not (example: double insulated devices with "two prong" plugs). >>If the EUT is >>normally grounded then the ground wire needs to be a part of the >>main cord if >>that is usual or practical; or routed in close proximity to the >>mains cord (10 >>cm) if it is usually supplied separate (as in many large three >>phase devices). >>Whether or not the EUT is physically grounded or not it is 40 cm from the >>reference ground plane (floor or wall) for table top equipment, >>and on top of >>but electrically isolated from the reference ground plane for >>floor standing >>units. It needs to be at least 80 cm from any other conducting >>surface or ground >>plane for both cases. >> The second question, which sounds like a bad riddle "When is >>50 uH equal to >>6 uH?" I think is a case of apples and oranges. The 6 uH is not >>being compared >>to the 50 uH developing impedance in the internal branch of the >>LISN leading >>back to the power source, but to the 50 ohm characteristic >>impedance of the RF >>connection to the Spectrum Analyzer. At 1 MHz the reactance of a 6 - 7 uH >>inductor is around 50 ohms an it increases from there. >> >>Paul McCoy >> >>Gert Gremmen wrote: >> >>> Robert, Muriel and group, >>> >>> Therefore it is essential that the mains wire is short and limited >>> to 50 cm ( < 2 feet) or even less. The transformer model I can see >>> ( 3 longitudal straight "windings" , sharing the same magnetic >>field= def. >>> of transformer) >>> I don't see however, how 6 uH can have the same impedance as 50 uH ? >>> >>> Please explain in detail. >>> >>> Regards, >>> >>> Gert Gremmen, (Ing) >>> >>> ce-test, qualified testing >>> >>> =============================================== >>> Web presence http://www.cetest.nl >>> CE-shop http://www.cetest.nl/ce_shop.htm >>> /-/ Compliance testing is our core business /-/ >>> =============================================== >>> >>> >>-----Original Message----- >>> >>From: Robert Macy [mailto:m...@california.com] >>> >>Sent: Wednesday, February 23, 2000 10:50 PM >>> >>To: Gert Gremmen; Muriel Bittencourt de Liz; Lista de EMC da IEEE >>> >>Subject: Re: conducted emissions test setup >>> >> >>> >> >>> >>Gert, >>> >> >>> >>The connection (Wire A) the ground connection from the EUT to >>the ground >>> >>plane is indeed quite long and inductive and "does not exist" >>> >>However, the >>> >>three wires HOT, Neutral, and GND form a 3 winding >>transformer which has a >>> >>core impedance in the range of 7uH for a six foot length of >>> >>Belden AC cable. >>> >> >>> >>This "transformer" makes for some very interesting >>interactions at those >>> >>higher frequencies, since around 1 MHz the core's impedance is >>> >>equal to the >>> >>LISN. >>> >> >>> >> - Robert - >>> >> >>> >>-----Original Message----- >>> >>From: Gert Gremmen <cet...@cetest.nl> >>> >>To: Muriel Bittencourt de Liz <mur...@grucad.ufsc.br>; Lista de >>> >>EMC da IEEE >>> >><emc-p...@ieee.org> >>> >>Date: Wednesday, February 23, 2000 1:17 PM >>> >>Subject: RE: conducted emissions test setup >>> >> >>> >> >>> >>Hello Mureil, >>> >> >>> >>When it comes to conducted emission measurements: >>> >> >>> >> >>> >>The test consist of one loop: >>> >> >>> >>EUT --> capacitance to --> Ground plane --> LISN --> 50 Ohm >>--> 1of3 MAINS >>> >>Wires >>> >> >>> >> >>> >>The 50 Ohms will be replaced by the spectrum analyzer or >>> >>measuring receiver. >>> >> >>> >>CISPR prescribes that the ground connection A) of the EUT be >>> >>grounded to the >>> >>ground plane. >>> >> >>> >>All other grounds may be floating including mains supply to the LISN >>> >>The spectrum analyzer will of cause be grounded by means of the >>> >>signal cable, the only right way to do. >>> >> >>> >>The ground plane is one plate of the capacitor in the loop above. >>> >>The EUT is the other . In parallel to this you will find the >>wire A above >>> >>but it has self induction, and is virtually not there when >>> >>frequencies rise. >>> >>Make the ground plane too small: C gets smaller -> less >>current -> less >>> >>voltage on 50 ohms. >>> >>Conclusion : ground plane should be much larger then EUT so >>EUT determines >>> >>the capacitance. >>> >> >>> >>The metal is not relevant. Without ground plane you will find >>not enough >>> >>emission, your equipment will erroneously pass. >>> >> >>> >>You should connect mains supply to the LISN, so normally GND >>goes to the >>> >>ground plane automatically. Is not relevant for the test. >>> >>The ground plane will indeed be the final reference ground for all. >>> >>Your test equipment will be bounded to the ground plane by means >>> >>of the coaxial cable connected to the LISN >>> >>Yes your PC may have the same supply as the LISN (not after LISN) >>> >>Stay away however, from your EUT as far as possible to >>prevent coupling. >>> >>If having problems of mains pollution, add a good quality mains >>> >>filter boxed >>> >>in metal in series with mains wire to the LISN. Mount box to >>ground plane >>> >>thoroughly. >>> >>LISN may leak trough interference from mains. >>> >> >>> >>The loop above WILL receive all long wave and medium wave as well as >>> >>shortwave >>> >>radio emitters in europe, normally at levels below the limit. >>Sometimes >>> >>Murphy >>> >>decides to join you and your tests and then .... you will >>have to screen >>> >>everything or subtract the environment interference from test values. >>> >>(only pre compliance testing) >>> >> >>> >>Regards, >>> >> >>> >>Gert Gremmen, (Ing) >>> >> >>> >>ce-test, qualified testing >>> >> >>> >>=============================================== >>> >>Web presence http://www.cetest.nl >>> >>CE-shop http://www.cetest.nl/ce_shop.htm >>> >>/-/ Compliance testing is our core business /-/ >>> >>=============================================== >>> >> >>> >> >>> >>>>-----Original Message----- >>> >>>>From: owner-emc-p...@ieee.org >>[mailto:owner-emc-p...@ieee.org]On Behalf >>> >>>>Of Muriel Bittencourt de Liz >>> >>>>Sent: Wednesday, February 23, 2000 5:33 PM >>> >>>>To: Lista de EMC da IEEE >>> >>>>Subject: conducted emissions test setup >>> >>>> >>> >>>> >>> >>>> >>> >>>>Hello Group, >>> >>>> >>> >>>>Reading CISPR 16 and CISPR 22, I had some doubts about the >>correct setup >>> >>>>for doing conducted emissions testing. I'd like to know if >>someone can >>> >>>>put a light on this subject. >>> >>>> >>> >>>>My questions are: >>> >>>> >>> >>>>- The "ground plane" is really needed? If I don't use one, >>my results >>> >>>>will be wrong? Or without a ground plane I'll be testing >>for the "worse >>> >>>>conditions"?? >>> >>>> >>> >>>>- The ground plane can be made of any metal? Or there are specified >>> >>>>ones? >>> >>>> >>> >>>>- The ground plane must be bonded to the GND connector of >>my mains? (my >>> >>>>outlet is three pins: P, N and GND) >>> >>>> >>> >>>>- If I understood, the ground plane will be my "reference ground", >>> >>>>right? >>> >>>> >>> >>>>- My testing equipment (receiver+LISN) must be bonded to this same >>> >>>>ground plane? The ground plane is bonded to the ground >>connector of my >>> >>>>mains? >>> >>>> >>> >>>>- I use a PC to collect the test data. Can it be feeded by the same >>> >>>>circuit that feeds the EUT+LISN? The ground connector of >>the PC (the one >>> >>>>that's in the three pin plug) should be at the same ground than the >>> >>>>ground plane? >>> >>>> >>> >>>>Thanks in advance and Regards! >>> >>>> >>> >>>>Muriel >>> >>>> >>> >>>>-- >>> >>>>>>8-)8-)8-)8-)8-)8-)8-)8-)8-)8-)8-)8-)8-)8-)8-)8-)8-)8-)8-)8-)8-)8-)8-) >>> >>>>Muriel Bittencourt de Liz, M.Sc. - EMC Engineer >>> >>>>GRUCAD - Group for Conception & Analysis of Electromagnetic Devices >>> >>>>Santa Catarina Federal University - UFSC >>> >>>>PO Box: 476 ZIP: 88040-900 - Florianópolis - SC - BRAZIL >>> >>>>Phone: +55.48.331.9649 - Fax: +55.48.234.3790 >>> >>>>e-mail: mur...@grucad.ufsc.br >>> >>>>ICQ#: 9089332 >>> >>>> >>> >> >>> >> >> >>
<<attachment: Gert Gremmen.vcf>>
