Gert; Certainly the coupling to a cable is dependent upon the frequency of the signal and the cable's length. In the case of a product whose cables may be 30-40 meters long, coupling may theoretically exist at frequencies down to ~2.5MHz. There are a few considerations:
1) the current radiated immunity standard (IEC 801-3:1984) tests down to 27MHz (quarter wavelength = ~2.8m) 2) the actual cable in Mr. Fielder's situation may be a quarter wavelength at ~2.5MHz 3) EN61000-4-3 tests radiated down to 80MHz (quarter wave= ~1m), with lower frequencies, as you mentioned, being tested using direct coupling (EN61000-4-6) 4) IMO, cables should be exposed to the uniform field area (1.5m) and then sufficient to exit the room 5) IEC 801-4/EN61000-4-4 require testing of cables that may exceed 3m in length (although this length during test is not specified) 6) as you say, the cable LENGTH is rarely varied (per 55022) to maximize emissions; 2.5 m is the practical limit on emission from a cable at 30MHz, with the real field variation coming from cable orientation/coupling 7) the reality of testing!!- while it may not make engineering sense to do emission testing with a cable longer than 2.5 meters, IMHO it doesn't make sense to have a separate cable made for each different test (801-3, 1000-4-3, 1000-4-4, 55022/11) It is in consideration of all of the above that I recommend a minimum 3m length for cables which may be significantly longer in production. As for the cable clamp in EFT, in a situation where a system includes 2 'black-boxes' connected by a cable >3m, the clamp has to be placed first at one end, and then at the other, which is why I generalized my statement earlier. Regards; Bob Martin Sr. Technical Manager ITS - Northeast (978)263-2662 [email protected] The opinions expressed are my own and not necessarily those of my employer. ---------- From: Ing. Gert Gremmen To: Robert F. Martin ITS/QS-Box; [email protected]; Francis L. Fiedler Subject: Re: Cable lengths needed List-Post: [email protected] Date: Thursday, April 16, 1998 3:58PM Hello Robert, Martin, As during emission testing and immunity testing cables act like antennas, we should treat them like antennas. The efficiency of a wire antenna is a sine function of its frequency and length. Therefore at a quarter wavelength optimum emission/receiption takes place. For emission testing cables longer then 2m50 meters (lowest F= 30 Mhz :: 10 meters wavelength) don't make sense. In immunity testing the same reasoning is valid, but due to the limited uniform field area size only one meter or less of it is effectively participating in picking up signal. In TEM and GTEM cells this length is even shorter. EN 55022 requires you to vary the length of the cables to maximise emissions. This is most often not practically realised. Instead max length cables (3 meters ore more) are used. For this reason the IEC decided in their new standard (IEC 1000-4-3) to start immunity testing at 80 Mhz instead of 30. The frequency below 80 Mhz. is extended to 150 Khz and is current injected into the cables using coupling/decoupling devices. The philosophy is that at these lower frequencies the contribution of cables is much higher as the enclosure energy pickup.(IEC 1000-4-6) For EFT and the clamp, the end of the cable should be the end near the apparatus under test. I believe that is what you meant to write. For mains voltage the EFT pulses are galvanically coupled into the phase wires. Regards, Gert Gremmen Ing. BTW is it standard or norm ; standardization or standardisation or normalizing I am not nativety speaking/writing this language (as you may have remarked) so please group, can someone shine some light on these linguistical topics. ? == Ce-test, Qualified testing == Consultants in EMC, Electrical safety and Telecommunication Compliance tests for European standards and ce-marking Member of NEC/IEC voting committee for EMC. Our Web presence: http://www.cetest.nl List of current harmonized standards http://www.cetest.nl/emc-harm.htm 15 great tips for the EMC-designer http://www.cetest.nl/features01.htm -----Oorspronkelijk bericht----- Van: Robert F. Martin ITS/QS-Box <[email protected]> Aan: [email protected] <[email protected]>; Francis L. Fiedler <[email protected]> Datum: donderdag 16 april 1998 18:21 Onderwerp: RE: Cable lengths needed >It has been my philosophy is such cases, that a 3 meter length should be >provided. > >During radiated emissions and immunity tests, a minimum of 1 meter >should be exposed. In the case of immunity, the 1 meter should be in the >'uniform' field area (even if you are not requiring the field uniformity >test). The 3 meters then allows the cable to exit the enclosure. > >In the case of EFT, the standard specifies tests of cables >3 meters, >and requires that the 1 meter coupling clamp be placed near the end of >the cable. Using a 1 meter cable would be difficult because of the size >of the clamp, and may also affect the results because of the potential >for improved 'suppression' of the coupled transient via the support >equipment. > >Bob Martin >Sr. Technical Manager >ITS - Northeast >(978)263-2662 >[email protected] > >The opinions expressed are my own and not necessarily those of my >employer. > ---------- >From: Francis L. Fiedler >To: [email protected] >Subject: Cable lengths needed >Date: Wednesday, April 15, 1998 6:24PM > >Greetings, > >Our equipment operates with cables that vary from two feet to two >hundred feet in length. The length is dependent upon the customer's >needs. Some cables are provided by the customer for their applications. >We test our equipment to industrial environment using EN50081 and 50082. > > >What length of cable do you think should be used for test purposes? The >length of wire subjected to the test signal is usually one meter with >additional 0.3 meters for positioning or decoupling purposes. Long >length of wires can result in use of large areas and or bundling. Many >test instructions specify not to bundle. > > >All of your comments will be greatly appreciated. If you prefer to >answer directly to me, your comments will be welcomed. > > >Respectfully, > > >Francis Fiedler >
