8th July 2003 - Recent developments in EMC and safety standards

John Woodgate, J.M. Woodgate & Assoc.

At the July meeting John Woodgate gave an informative and (as far as is possible with such a subject) entertaining introduction to the European Regulations on EMC and electrical safety that apply to professional audio equipment.

The relevant European Regulations are: the EMC Directive, which covers electromagnetic interference; the Low Voltage Directive, which covers electrical safety; and the RTTE Directive, which covers radio transmission and telecoms equipment. Directives are not laws, but instructions to the governments of Member States to implement them in national laws; a typical Directive may leave Brussels as a 20-page document, but by the time the DTI has turned it into UK law it is 65 pages which may or may not entirely resemble what Brussels intended!

The EMC Directive is concerned with both outgoing electromagnetic disturbances (emissions) and the effect on the equipment of incoming ones (immunity), in each case both through the ether and along cables. Emissions standards are intended to protect radio communication and set limits to the disturbances which other equipment must withstand; the first such standards for power distribution date from 1898. Immunity standards ensure the equipment continues to operate properly when close to a mobile phone, for instance; depending on the nature of the equipment and the way in which it might fail, there may be safety issues too.

The Low Voltage Directive applies to equipment powered by more than 50V AC, but ought also to apply to equipment with high energy batteries; however it has not been possible to formulate a definition of "high energy", because lawyers don't understand electricity, and there have been suggestions the LVD ought to apply to all power sources right down to 0V DC.

Equipment carrying the CE mark must conform to all relevant standards that have been "notified" in the Official Journal of the European Community (available on the Web at http://www.ojec.com/), but companies cannot change products overnight, so when a new EMC or safety standard is published there has to be a transition period before it becomes compulsory. This gives rise to several animals which inhabit the regulatory jungle, the strangest of which is the docopocoss or "date of cessation of presumption of conformity of the superseded standard", after which manufacturers must not ship products that conform only to the previous standard.

The manufacturer must provide a declaration that the equipment conforms to the technical requirements under applicable Directives, listing the standards (including dates, to identify which edition). It must be signed by someone with technical knowledge of the product and sufficient authority, usually the Technical Director.

All that the law requires is that the Declaration shall be true; there is no requirement to test the product, though that is often the only way to be sure it conforms. It is, for instance, for the signatory to judge whether a product that has been modified needs to be re-tested. In most countries, a product will only be investigated by the authorities if someone complains about it; however, there are some territories which test proactively, particularly Germany and Cardiff. If official tests show the product fails to conform the manufacturer is informed and asked to remedy the problem. Only if the manufacturer is unco-operative will any draconian action be taken.

The cost of compliance (testing and documentation) should be regarded as marketing, not development; if it seems expensive, compare it to the advertising budget!

The lecture concluded with technical details of the standards: EN 55103 (EMC), and IEC/EN 60065 and 60950 (electrical safety).

There followed a lively Q&A session, including:

Q: In the case of a condenser microphone with a valve amplifier, it must be impossible to stop mobile phone signals getting in at the capsule level unless it's completely enclosed, which defeats the purpose?
A: Valves are much less prone to rectification effects; FETs are 20-26dB better than bipolar transistors, and valves are better still. You can't use knitted mesh, but with etched foil it should be possible to cope with fields of 10V/m, though probably not 40V/m.
Q: What changes have taken place in 61000-3-2?
A: There was the "millennium amendment", a wide-ranging revision agreed at a meeting in Rome in December 2000 and edited by our speaker at the end of December, when the rest of the task group were on holiday. The revision process started in 2001 but has become bogged down and is about to time out, so various amendments which have been in "cold storage" (because you can't amend a standard at the same time as revising it) will appear during 2004; they include an improved specification of how high (e.g. 39th) harmonics, which put large (life-shortening) loads on capacitors and induce currents in phone lines, should be measured. There is also a new standard coming for equipment drawing more than 16A from a 3-phase supply; the 3rd, 9th, 15th, harmonics can add up in the neutral conductor, in theory reaching 2.8 times the current in the "live" conductors.
Q: Is the power industry doing anything about putting in thicker conductors?
A: BS7671 doesn't say anything explicit. The USA don't like 61000-3-2, and their system is different from European power, but they have had instances of neutral conductors burning up.
Q: What is a "screened cable port"; does it include XLR microphone leads?
A: It wasn't intended to; it was intended to apply to BNC connectors. You would have to specify the cable-mounted half very carefully in the case of an XLR.
Q: Is the IEC still able to get enough people to work on standards?
A: It is a problem in some areas, for instance the US power industry has opted out. But other countries such as Japan and Korea are putting more effort in. And we mustn't forget the AES: Standards Subcommittee 05, and particularly its Working Group SC-05-05, covers EMC-related issues.

John Grant