|
JohnWoodgate's Column
CISPR 22/ EN55022: This, the emission standard for ITE equipment, is widely referenced in other product standards as if it were a Basic Standard. It is thus very unfortunate that the standard is the subject of a host of proposals for amendment, as well as being imminently issued in a new edition 4. From the public part of the IEC web site, http://www.iec./ch, we can find the following (some details omitted here):
CISPR 22 (1997-11) Ed. 3 Information technology equipment - Radio disturbance characteristics - Limits and methods of measurement
CISPR 22-am1 (2000-08) Ed. 3 Amendment 1
CISPR 22-am2 (2002-10) Ed. 3 Amendment 2
CISPR 22 Ed. 4.0 Information technology equipment - Radio disturbance characteristics - Limits and methods of measurement
CISPR 22 A2 f15 Ed. 3.0 PWI CIS/G/197/CDV Modification to CISPR 22: Measurement at 3 meter separation
CISPR 22 A2 f12 Ed. 3.0 PWI CIS/G/174/NP New networks for EMC testing of Local Area Network equipment
CISPR 22 f16 Ed. 4.0 CDM CIS/I/51/CD [Wrong title on the web page!] Examples of ITE with a radio transmission and/or reception function
CISPR 22 A1 f3 Ed. 4.0 CCDV CIS/I/71/CDV Modification of CISPR 22: Improvements to Annex C
CISPR 22 A1 f1 Ed. 4.0 CCDV CIS/I/65/CDV Modification of CISPR 22: Emission limits and method of measurement from 1 GHz to 6 GHz
CISPR 22 A1 f2 Ed. 4.0 CCDV CIS/I/66/CDV Modification of CISPR 22: Emission limits and method of measurement from 6 GHz to 18 GHz
CISPR 22 A2 f14 Ed. 4.0 CDM CIS/I/6/CD Modification of CISPR 22: Test set-up clarification
CISPR 22 A2 f9 Ed. 4.0 CDM CIS/I/44/CD Amendment to CISPR 22: Clarification of its application to telecommunication system on the method of disturbance measurement at ports used for PLC (Power Line Communication)
Quite how this is going to be resolved in CISPR/I is far from clear, and the question of what will happen in CENELEC TC 210 is even more important. Maybe it will be possible to combine the CDV documents into one amendment so that just one dow can be assigned. But until this is determined, there must be a lot of unwelcome uncertainty.
CISPR 20/EN 55020: Beware of editorial problems in Annex F! The paragraph above Figure F1 doesn’t make sense and the vertical scale of the graph in Figure F.3 should presumably be in millivolts, not volts! There are several other places in the standard where there is ‘room for improvement’, too. The same problems are present in two recent CISPR/A drafts at the first voting (CDV) stage, by which stage such matters should have been corrected.
These editorial problems are all too common, especially but far from exclusively in CISPR standards. In many cases, no English native speaker has been involved in preparing and editing the text. Most people are reluctant to do this sort of work, and managements are averse to funding it anyway. However, there ARE people (some are columnists!) who aren’t dismayed by the prospect of editing 50 pages of Euro-Engleesh. It needs to be borne in mind that industry gets the standards it deserves, and the waste of time (and quite possibly money) resulting from poorly-worded standards could pay for quite a lot of editing.
Switch-Mode Power Supplies in stand-by mode: It is well-known technically that operating an SMPS on light load causes very short current pulses to be generated in the switching circuit. What hasn’t been so well recognized is that these pulses have a spectrum that extends up to much higher frequencies that the broad pulses generated on heavy load. In fact, to a first approximation, the spectrum of a pulse consists of every harmonic of the repetition frequency, at equal amplitudes up to the frequency corresponding to twice the pulse duration. That frequency, or the harmonic nearest to it, is strongly attenuated, and above that frequency the harmonic amplitudes fall away, but often not monotonically - a sin x/x envelope is not unusual. So these short pulses may well cause the emissions, both conducted and radiated, to exceed by some margin the levels measured during conventional testing. There are proposals to introduce testing also in stand-by modes, and it is difficult to argue against that, because actual cases of interference have occurred.
Revision of IEC 61000-4-6; (r.f. conducted immunity) UK has voted against the FDIS, after much discussion, because there is much confusion over the use of decoupling networks.
CISPR receiver: A proposal to introduce a new detector into the CISPR receiver, for better assessment of disturbance to digital communications, has also attracted a negative vote, because of the cost of re-equipping, particularly for those companies and test-houses that have now unsupported Hewlett-Packard equipment.
Updating CISPR standards: The situation highlighted by the Cardiff prosecutions, that CISPR 14-1/EN 55014-1 doesn’t have any limits for emissions in the UHF television broadcasting bands, has focused attention on the other CISPR product-family standards, notably CISPR 13, 20 and (yet again!) 22, all of which have omissions that could mislead manufacturers.
Wise counsel would be to realise that emissions at ANY frequency from 9 kHz to 2 GHz could cause problems these days, and that will need to go up to 6 GHz in a few years. Beyond 6 GHz, emission ranges are expected to be too short to cause significant problems. So, even if the applicable standard doesn’t cover a particular frequency band or three, it would be a good thing to run some checks.
EMC fallacy: Some discussion has been prompted by an article by the redoubtable Ivor Catt in ‘Electronics World’, recounting anecdotes about EMC incidents. It is true that much EMC testing is done under artificial conditions which can rather easily be shown to deviate significantly from how the equipment under test is used in real life. However, there IS a feedback mechanism; if a test is not effective, there will be complaints of interference and they will lead, in due course, to an improved test and/or revised limits. The only problem with that is that the feedback mechanism doesn’t detect a too-stringent limit.
Perquisites of EMC standards work: CISPR and IEC TC77 and its sub-committees are scheduled to meet at a luxury resort on an island between Korea and Japan later this year. The current situation, with SARS still active and North Korea testing missiles, is making people curiously reluctant to be nominated to attend.
Banana skin: I’ve heard of an incredible and highly amusing (if it didn’t have serious overtones) EMC banana skin, but I can’t tell you about it (yet?) because there are big security issues involved. It is also possible that someone could thoughtlessly create an incident as a prank, if details were made public, and that might have dire consequences.
J. M. Woodgate B.Sc.(Eng.), C.Eng. , MIEE, MAES, FInstSCE Email: kernel@jmwa.demon.co.uk Web: http://www.jmwa.demon.co.uk/ ã J.M.Woodgate 2003
|