Report on the EMCTLA Workshop

on the 12th February 2001

By Malcom Burchall C.Eng FiEE, EMC and Power Electronics Consultant

This workshop was held to discuss various EMC matters of direct concern to the EMC Test Labs Association (EMCTLA).  It was felt that some of these matters could be of use to those of our readers who will be particularly concerned where they show a tightening or extension of regulations.

1.  The Expression of uncertainty in EMC testing

This session was mainly given to a presentation of the draft of the United Kingdom Accreditation Service (UKAS) document LAB 34 by Alex Penfold of UKAS.

This document lays out in detail how to calculate the uncertainty of a given EMC measurement taking into account the accuracy of the various instruments and components in the coupling path.  In this respect it is a very valuable document which is intended to replace the previous document NIS81 currently being used by UKAS accredited test houses.

The worrying feature of this is that it can result in a test house only being able to give a qualified pass to a piece of equipment even if it has emissions below the specified limit but not below the lower tolerance limit.

This is something that has also been concerning CISPR in recent years and goes against the statistical concept of the 80/80 rule embodied in CISPR22 (BSEN550022).

It has to be remembered that the limits quoted in the original CISPR standards were pragmatic limits based on the amount of interference perceived by radio and television users and took into account the typical uncertainties of the existing test houses.  There is a danger that in introducing the concept of a qualified pass by the UKAS test houses we will be seeing an arbitrary tightening of the test limits by as much as 6dB.

The Draft document LAB 34 can be found on the EMCTLA Web site www.emctla.co.uk.

2.  Worked examples of Uncertainty Budgets

Phil Carter of Acemark Europe Ltd gave a very detailed set of calculations of typical test house uncertainties.  Any reader who is interested in the fine details of this is recommended to read the paper itself available on the EMCTLA web site.

3.  ISO/IEC 17025

This was another presentation by Alex Penfold detailing the changes which ISO/IEC 17025 had made from the previous UKAS document M10 used for the accreditation of test houses.  Test houses meeting ISO/IEC 17025 are deemed to have met ISO9000.

This standard will be used for the re-accreditation of all the UK test houses starting from April 2001 and completing in June 2002.

4.  Assessment of Competent Bodies

UKAS have been delegated by the DTI to carry out assessments (not accreditation) of all the UK competent bodies in accordance with URN99/627 and the appropriate parts of EN45004:1995 and SI 1992 No 2372 to be completed by October 2001.

5.  CISPR 22 Telecom Port Tests

Tim Williams of Elmac Services detailed the changes and test methods introduced in the 3rd edition of CISPR 22 for the “Measurement of disturbances of telecommunication ports”.  Compliance has to be shown by 1st August 2001 so this is of immediate interest.

The definition of “telecommunication ports” can be thought of as any port connected to a long cable.

A number of test methods are given in Annex C using either coupling networks or 150 ohm coupling terminations.  Alternatively current methods can be used.

There are problems with the test methods partly due to the large variety of cable types but also due to the difficulty of stabilising the common mode impedance at 150 ohms and some solutions were presented for these.

6.  US requirement for EMC and Telecommunications  The EU-US MRA

This paper should have been given by Dave Imeson Compliance Europe Ltd but there was insufficient time.  However the most important part of the printed paper was the requirement for EU test houses to be specifically accredited for tests to FCC Part 15/18.  There is also the possibility of using the newly-created “Telecommunications Certification Bodies” who are empowered to certify products.

7.  EN61000-3-2 and EN61000-3-3

This session was devoted to an update of the harmonics and flicker standards situation by Phil Carter.

Now that amendment A14 of EN6100-3-2 is in force (as of the 1st January 2001) all the problems associated with the previous edition should start to vanish.

The major problem for test houses is that the software controlling the commonly used test sets will have to be modified to meet the new standard.

As soon as possible the EMCTLA intend to carry out a repeat of the inter-laboratory comparison which was used to highlight the deficiencies in the old standard.  It is hoped that this will show good correlation between different test houses which will hopefully put this harmonics problem to bed for the last time.

The modification of the IEC standard is under way but it will be some time before we see any results from this.

The situation with the voltage fluctuations and flicker standard is that there are ambiguities in the present standard with measuring inrush current surges which vary either with the point of wave switch timing or vary with a previous history of magnetisation.  It would appear that most of the test houses are working with an unwritten agreement that this standard need not yet be applied even though it became mandatory on the 1st January 2001.  A new version of the standard is due to be published very soon which removes these ambiguities at which time there will be a lot of catching up to do.

It seems that a lot of manufacturers and test houses still do not understand the ramifications of this “via the back door” introduction of inrush current limits disguised as voltage deviations.  The problem with a lot of test houses is that they have had to rely totally on complicated test instruments to measure flicker (which is extremely hard to measure with simple instruments) and very few of these instruments actually measure the first half-cycle true r.m.s. voltage deviation.  In fact many of these test sets specifically exclude the first short time of operation and wait for a stable condition to be achieved.  There seems little evidence that this is being addressed.

It is more than likely that some equipment modifications will be called for as test houses start to fail non-compliant equipment.

In addition to the above standards there are their larger brothers already issued or coming along for equipment rated above 16A or inherently unable to meet the <16A standards.

8.  Testing above 1 GHz.

This presentation was given by Brian Jones, EMC consultant, and highlighted the problems of testing above 1 GHz.  These requirements are starting to appear in new or amended product standards.

As equipment frequencies continue to rise it is becoming increasingly necessary to carry out these very high frequency emission and immunity tests and the test equipment and methods available are very limited.

For equipment covered by EN55011 (I.S.M.) substitution methods will have to be used up to the end of 2001 when amendment A1 is due to be published.  For EN55013 equipment substitution or measurements by a measuring set are accepted.  A lot of the details for this are contained in CISPR16.  The most important document for most manufacturers is CISPR22 where an amendment covering the range 1-18GHz is currently under discussion.

One important new piece of test equipment on the market is a 1-7GHz calibrated noise source (CNR) which can be used in substitution testing or site calibration.

A new part 3 of standard EN 50147 is also currently being worked on to provide for emission measurements in fully anechoic rooms as well as immunity measurements.

Malcolm can be contacted on tel: +44 (0)1799 540784 or Emal: malcolm@burchall.co.uk