 |
EMC in the New Millennium - Peter J. Kerry |
EMC is an important part of the ever-changing
world in which we now find ourselves. Society is increasingly dependent on radio,
telecommunications and electronic products, but their use is only possible if the
necessary EMC criteria are met. However the criteria of the last century will not
necessarily be those required for the one that we have just entered.
EMC is no longer a backwater of engineering where we can afford to leave the problems to
the ‘experts,’ it is an area where every engineer needs to become involved.
Undoubtedly EMC has made significant strides forward during the past decade, but further
progress is necessary if we are ensure that we continue to enjoy a relatively
interference-free environment in the years to come. This paper looks at the six major
challenges that need to be tackled if we are to be successful in controlling EMC in this
new millennium.
1. Introduction
Historically Electromagnetic Compatibility (EMC) has tended to be in the backwaters of the engineering profession, with any problems being resolved by the installation engineer or the servicing department. However, over the past decade EMC has made significant strides forward, with much of this progress being fuelled by the implementation of the European EMC directive. During this period there has also been rapid development in technology and its application; which poses new challenges for EMC engineering.
If we are to continue to use the radio services we want, together the new technology products that are being produced, then the EMC challenges that this poses need to be mastered. The engineering profession is well placed to play a key role in overcoming these challenges; so as we enter the new millennium, it is timely to reflect on the present situation, to identify the challenges, and to develop strategies to deal with them.
EMC is not a new subject and indeed the International Special Committee for Radio Interference (CISPR) was established as long ago as 1934. The objective of this group was to protect radio services from interference. This is a difficult and challenging task, with the work set to continue into the foreseeable future. Indeed despite the costs involved, the present high level of participation continues to grow, with over 200 delegates at the 1999 meeting. This interest in CISPR’s activities indicates the importance that industry and the radio regulatory authorities give to EMC. The difficulty of the task is probably best reflected by the following quote from the report on the inaugural CISPR meeting.
"The work with interference limits progressed only slowly. At the first CISPR meeting the lowest protected field strength was set at 60dBµV/m with a modulation index of 80%. The lowest protection ratio was to be 40dB. Various limits were proposed but none even came close to the (then required) 20dBµV/m."
The challenge of meeting the desired limits is still with us today, the only difference is that we cannot afford the luxury of going slowly.
Undoubtedly the most significant milestone in the history of EMC is the European EMC Directive. Expressed simply, the European EMC Directive requires that a product placed on the market be certified to have adequate immunity and acceptable levels of emission, i.e. it should be compatible with the environment in which it is intended to operate. Whilst this paper focuses on the radio aspects of EMC, it should be noted that the Directive equally applies to compatibility with the power environment. The Directive provided several routes to achieving compliance, but in general industry considered the simplest option was to certify it as being compliant with agreed international standards. This in turn caused a rapid growth in demand for EMC standards both at the international and European level. Rather interestingly, this rapid proliferation of standards has caused the European Union to review the situation in which we now find ourselves. This has resulted in the European electrical standardisation organisation (CENELEC), establishing a Strategic Review Panel to review the situation to ascertain whether there might be a way of reducing the number of EMC standards and/or simplifying the means of compliance with the directive.
A PC or TV? Why are the EMC standards different?
Present EMC standards are classified in a three level hierarchy:
Basic Standards
As the title implies, Basic Standards provide basic details for the method of testing particular phenomena. Hence the Basic Standards form the foundation of the standards hierarchy. The Generic Standards then apply the Basic Standards to a particular EMC environment by defining the limits to be used. The Product Standards then are at the top of this hierarchy by providing details of how limits should be applied during the testing of a particular product, or family of products. In an ideal world these three types of standards would stack one on top of the other like building bricks. However, many of the Product Standards have a long history predating the basic and generic standards, and it is this that has given rise to the inconsistencies between the various standards. Yet it is upon this non-coherent base of standards that the EMC Directive is required to operate.
The legal requirement for EMC certification has caused each industry product group to require its own EMC standards. Producing a standard for each of these industry areas was never going to be practical, given the limited time scale available. In order to fill any vacuums, the generic emission and immunity standards were developed. These now exist - two for the domestic environment and two for the industrial environment making a total of four Generic Standards. This new environmentally based approach to EMC standards is a radical departure from the traditional product-based standards. Indeed it begs the question do we need EMC product standards in the future?
EMC is about ‘compatibility.’ Consequently it embraces both unwanted emissions from equipment, and the need for adequate immunity of the equipment. Whilst everybody agrees with the need to control interference by controlling unwanted emissions, the inclusion of EMC immunity requirements in the Directive is a new idea from previous requirements. This is still a controversial subject with many believing that this is a quality issue and is decided by the customer. However, in a world where there is a seemingly unending growth in radio services, the consumer expects that products will continue to operate in the presence of mobile phones and base station transmitters etc. In many areas immunity is seen as a safety issue, with manufacturers testing their products well beyond the requirements of the EMC directive. There are well balanced arguments for and against EMC immunity requirements, and no doubt the debate will continue. In many respects it is academic, as the EU has reviewed the operation of the EMC Directive, and has decided to retain the requirement for EMC immunity.
In relative terms the speed of development of EMC standards over the past decade has been very rapid and only history will tell whether we have been successful. The present standards pose many questions; one might well ask how the limits were arrived at, as often they seem to be historical values rather than something more logical. The other area that might cause concern is the fact that the testing requirements generally stop at 1GHz, apparently neglecting to protect radio services in the remainder of the frequency range. The one thing that is certain is that the present standards have scope for improvement.
We are living in a continually changing world, and the EMC environment is no exception to this. It will not be sufficient to rely on present standards that were derived from some past premise, unless we are able to periodically revalidate that premise. If we are to continue to protect the use of the new and evolving radio services, then the EMC standards must take into account these changes. Equally radio system designers need to be aware of the existing EMC environment and that any required improvements to the environment will take time to realise.
You will no doubt already have gathered that the challenges we face are far from being simple; however, if the scenario is studied, it will be seen that many of the various factors can be grouped together. My analysis is that there are six major challenges that engineers need to tackle if we are to be successful in controlling EMC in the new century. These may be described as follows:
The following sections discuss the issues that these challenges involve.
Amongst the many electrical products in a typical home, there will be a television, a computer and a fridge. These products are tested to three very different EMC standards, although from an EMC perspective they are operating in a single environment. This anomaly has arisen through standards being developed on a product basis, this enabled low cost specialised test methods to be developed for specific product groups. However without any overall management to ensure consistencies between the various product standards, anomalies were bound to happen. Whilst it was possible to live with such differences, technology convergence has highlighted the discrepancies. In some instances manufacturers are testing to several standards in order to be certain that they are compliant with all the relevant standards. This is clearly an unreasonable cost burden, and some rationalisation is needed.
Technology convergence is most clearly seen with the advent of multimedia products. The television and the computer are set to become similar if not identical products. Equally there is nothing to prevent enterprising manufacturers from integrating these into the refrigerator or any other domestic appliance. Product convergence is with us now, hence there is an urgent need to address this in EMC standardisation. To some extent this has already occurred with the development of the Generic Standards, but more work is needed if this is to become a reality.
Where suitable Product Standards do not exist, then the Generic Standards become applicable. This is an area where industry trade associations (ITAs) could have a significant role in ensuring that EMC standards remain relevant to products in their domain. One solution would be for the ITAs to develop application guides or codes of practice for the application of the Generic Standards. These can be developed much more quickly than the international committees operate, hence ensuring that standards continue to remain relevant to the products being produced. This could be used as a radical approach that could lead to the eventual reduction in the number of EMC standards required.
The frequency range covered by present EMC standards is often far from clear. The scope of a standard often states that it covers the range from 0 – 400 GHz. However upon closer inspection it is found that there are no limits to meet and no testing is required above 1 GHz. This situation arises from the fact that the scope was written to address a legal problem, whereas the test methods were written to address a practical problem. It is possible to live with this discrepancy, but the 1GHz limit no is longer technologically acceptable. Indeed all responsible manufacturers recognise the need to test up to 2GHz to ensure protection of the mobile phone services.
There is however a need to address the question of how high in frequency should EMC testing be required to go? It is not envisaged that all products will need testing up to 18GHz, particularly as testing EMC performance in the microwave frequency range is a costly and time-consuming business. However with the speed of microprocessors increasing without any apparent limit, the current testing limit of 1GHz is not an acceptable option. How to limit unnecessary testing whilst continuing to protect radio services is a problem that requires a pragmatic solution. It is noted that the approach in the United States is to test up to ten times the clock frequency of the microprocessor. This approach appears to have proved itself empirically, so it may prove to be a possible solution.
An engineering challenge - faster computers means higher frequencies to be controlled.
We are entering an age where radio and television services are going digital, and also the sources of interference are frequently digital, yet EMC test methods remain rooted in analogue technology and the limits are based on the protection of analogue radio services. The major problem with the digital age is that interference often manifests itself differently. Similarly digital technology has sparked an unprecedented growth in communications. This has resulted in more radio services requiring protection from ever more potentially interfering sources.
Of particular concern is that digital radio services have different tolerances to broad band and narrow band interference than do analogue radio services. It is difficult to foresee how this will reflect itself in terms of interference limits and test methods, but it is an area that requires urgent investigation. Consider the following four scenarios:
a) The introduction of the personal computer brought with it the widespread use of clock signals, which were quickly identified as a potential source of interference. The current EMC standards for PCs have tight emission limits (e.g. CISPR 22) to avoid such interference. Innovation spawned the dithered clock oscillators to simplify meeting the limits by spreading the energy over a broader bandwidth. Regrettably studies have shown that digital radio services are more susceptible to broadband interference.
b) An interfering signal that is an annoying whistle for an analogue radio service, may remain totally unheard by a digital radio service. Consequently this form of spectrum pollution may go unnoticed.
c) Interference to a digital television picture is most likely to result in a frozen picture, and not manifest itself as the traditional blobs across the screen. The viewer may then deduce that the set is faulty and not report it as an interference complaint.
d) Cellular radio phones will search for a vacant channel, and hence automatically avoid those with interference. Consequently interference to a cellular base station may only cause a reduction of system capacity and not manifest itself as a problem except at peak times.
The simplest solution for all the above scenarios might be to reformulate the regulations to effectively ban the dithered clock, and forget the remaining scenarios, hoping that they will not result in interference complaints. This is obviously not a solution that is acceptable to the engineer, but alternative solutions will require reviewing the fundamental concepts of EMC. As the first step of resolving this problem there is a need to re-evaluate what we define as interference. Whilst the quality of the radio spectrum will continue to be related to the number of interference complaints, some additional suitable measures of EMC success will need to be developed. Without these, it will be difficult to agree on any new EMC limits. However this is not solely an EMC issue, as any decisions made will impact all aspects of electrical engineering; it is no longer possible to design a new system, of any sort, without considering if it will be regarded as having suitable EMC characteristics. Resolving the digital EMC problem is going to be a major challenge to all engineers in the coming decade.
Aside from the debate over the limits for digital services there is the more general concern over EMC limits. Traditionally EMC limits have been derived as the result of a compromise consensus between the interested parties. The present EMC limits, together with the test methods, have been arrived at separately for each product group. Hence there are discrepancies between the limits applied to different products used in the same EMC environment. This anomaly has been discussed in Section 5. Yet despite the lack of coherence between EMC standards, the number of interference complaints remain at a low level. Whilst there are many factors that might account for this, one has to face the question; are current limits realistic? This is not an easy question to answer.
One of the difficulties in answering the limits question is that present EMC standards fall into the three-level hierarchy discussed in Section 3. Without an agreed reference point there is no possibility of achieving coherent EMC standards. Whilst the Generic Standards are often used as reference EMC standards, what is actually required is a database of defined protection levels for radio services that can be used as the basis for deriving future EMC limits. It should be noted that CISPR has recently established a new subcommittee (CISPR/H) to produce this database. This is no small undertaking, and it will be some time before it feeds through to new and revised EMC standards.
The other aspect of EMC limits is that of consumer awareness. This aspect has never been fully discussed when deriving EMC standards. This particularly impacts domestic audio-visual equipment. At present the standards are based on an average television or radio. These fall far below what a hi-fi enthusiast might be expecting. Equally the cost associated with designing products to comply with standards could eliminate the low cost products aimed at the bottom end of the market. Generally the consumer is prepared to accept a lower performance if he/she considers the product to be good value for money. The other side of consumer awareness is that consumers should not be able to expect to use all products in all circumstances and expect the compatibility to remain satisfactory. It is not practical to specify EMC performance to meet the ‘worst worst case’ scenario. Most consumers are reasonable in this regard provided the fundamentals are explained to them. The question remains, how to improve consumer awareness? It is important that industry provide the answer before becoming pressurised by the complaints from dissatisfied consumers. It is pleasing that some areas of industry are tackling this challenge, with ideas such as grading products, and supplying them with high quality connection leads. Eventually such solutions will work their way into the EMC standards for the benefit of all.
The final factor driving the need for revision of EMC limits is the increase in the amount of electronic equipment. Limits have traditionally taken into account the probability of interference when setting limits. With the present vast increase in electronic equipment there is a need to reassess this factor to see whether it will impact on the necessary limits. In particular limits were designed to apply to products with a protection distance taken into account however, the growth of networks is tending to make this an irrelevant factor.
Whether EMC limits need to change or not, there are sufficient changes in the electrical environment to require the present limits to be reassessed. Failure to do this could adversely impact the future use of the radio spectrum, and be the cause of interference between electrical products.
Up to now, the limitations and deficiencies of the existing standards have been addressed. Regrettably the EMC of networks and installations is a missing link in the current mass of EMC standards. EMC standards primarily address the issue of manufacturing and marketing products and ensuring that the products are compliant. However, in many circumstances these products are connected to other products with cables that may be unscreened or unbalanced. Naturally this will seriously deteriorate the designed EMC performance. To further exacerbate this problem, many existing networks are being used for the transmission of data at speeds far in excess of their original designed bandwidth. There is a need to address these issues to ensure that medium and high frequency radio services are adequately protected for the future. This is not solely a radio problem, as EMC problems also manifest themselves as cross-talk between pairs in the same cable, and coupling to cables in adjacent ducts. In a recent radio interference case, a cable television (CATV) system was found to be causing radio interference. Upon investigation it was found that this was due to the CATV system picking up the interference from a buried telecommunications cable in a parallel duct. The current move to open up telecommunications in some countries by ‘unbundling the local loop’ is likely to compound the EMC problems, as it will become increasingly difficult to ascertain who is responsible for causing the interference. To further exacerbate the network EMC situation, wired home networks are becoming available. Many of these systems use the existing power or telephone wiring, which without co-ordination will lead to EMC problems. Competition and choice in service delivery benefits consumers, but there needs to be a balance struck between this, and the required level of protection for radio users.
Unless all those with network interests are prepared to co-operate in resolving the common EMC issues there are serious problems ahead. Having said this, it has to be appreciated that it is somewhat difficult to achieve the co-ordination of such a broad range of concerns. Hence it is understandable why this problem has not been tackled fully. There is a possibility of producing network EMC standards, and indeed in many areas such work is proceeding. It is becoming common practice for people to purchase pre-wired business accommodation, and pre-wired homes will soon become commonplace. To address such requirements there is a definite need for standards, however standards alone cannot solve the problem. Even the best EMC standards can be rendered useless with poor design, or when poor installations are used. There is a need to fill this void with guidelines for the design and installation of systems and networks. It would be in the interests of trade associations to prepare such guidelines, but this is essentially an education issue that is dealt with in the next section.
In addition to the EMC standards for equipment manufacturers, and guide lines for the installers, there is also a need for enforcement standards for the regulatory authorities. Most countries have their own national regulations for resolving radio interference. Whilst these are usually effective in resolving interference complaints, generally they do not extend to inter-network EMC issues. We are living in a global market place, where national regulations are regarded as being technical barriers to trade, so the national regulation route is not the preferred solution. This points to the need for agreed in situ measurement enforcement standards. It is fair to say that within the EMC community there are split views on this. On one side there are those opposed to enforcement standards as they could be used for policing installations etc. On the other side there are those concerned that without such standards, national authorities could impose their own limits.
The computer network connections could pose the biggest EMC challenge.
EMC education is seen as the major challenge for the future. The need for this exists at all levels. Regrettably EMC is frequently regarded as a ‘specialist area’ or even a ‘black art.’ However EMC might better be regarded as an environmental issue that has not yet gained the public awareness given to other environmental issues.
At the user level, there is a need for the general public to have a basic understanding of EMC. To some extent this already exists, and people have learnt to live with problems such as their cordless phone not working near the fluorescent light. However as housing density rises, the potential for interference between flats and houses increases. Dealing with such problems is often as much a social issue as a technical one. Solving such problems could be easier if there were a better understanding of the factors involved. The public perception of the word ‘digital’ is that of quality and interference free products and services. Yet this is not the reality, and an improved awareness of factors such as ‘freezing television pictures’ is needed. There would be a substantial reduction in the number of interference complaints if the public understood that television aerial systems have a limited lifetime due to exposure to the extremes of weather and atmospheric pollution. Similarly boiler thermostats operate in a harsh environment and often need replacing long before the boiler is ready for replacement. The present public understanding of EMC is limited, and linking interference to an old aerial or thermostat is not immediately obvious. It should be noted that whilst there are many variable factors contributing to the deterioration of these devices, typical life expectancies are in the order of 5 to 10 years. The challenge of putting such information across to the public is difficult, and is probably better done by the retail and servicing industry.
One of the major areas where consumer awareness is lacking is that of mobile phones transmitting periodically when they are not being used. Whilst it is normally acceptable for an aircraft pilot to require all mobile phones to be switched off whilst in flight, stopping them from being used in petrol stations, cars, hospitals etc. is much more difficult to enforce. To the average man in the street, not using a mobile phone means leaving it in his pocket unless it rings - this does not solve the potential EMC problems! It has been a long time since mobile phones have been confined to vehicles. Consequently there is a need for all products to be immune to these radio transmitters, or where this is not possible then some active means of restricting their use in sensitive areas needs to be found.
The retail and servicing industries are in a unique position between the manufacturer and the customer. Hence they are ideally placed to inform the public how to avoid and if necessary deal with EMC problems. With the gradual demise of the specialist retailer, this poses a challenge. Recently the Radiocommunications Agency has issued an information booklet (RA323) aimed at the television industry to assist them with the handling of interference complaints. One of the key points of the booklet was to address the importance of the quality of the television receiver’s aerial system, and that a period of around seven years was the expected life of a typical aerial system. It is in the industry’s interests to take note of this, as this is a neglected and potentially financially attractive part of the television market.
Whilst in the previous example the Radiocommunications Agency took the initiative to produce the booklet, in the general situation, it is in the interests of the trade associations to produce such guidelines. If trade associations are to fulfil their role, then guidance on the best use and installation practices needs to address EMC issues as well as the more traditional safety and performance ones. The EMC Directive is not completely clear as to its application to systems and installations. This is not helped by the standards being aimed primarily at products. Codes of Practice and Industry Guidelines could establish the best practice for individual industry groups, and go a long way to ensuring that the basic requirements of the Directive are complied with. It is noted that in some areas, trade association membership is in decline. One of the reasons given for this decline is that they are not seen as being relevant, yet the European Directives are one area where the support of a trade association is beneficial to industry. This is particularly the case with the EMC Directive.
The final area where EMC education is needed is that of the professional engineer. EMC is regrettably seen as a specialist area, hence it is often not included at the project design stage. How much time and money does industry waste on re-engineering products to meet the required EMC limits? Until EMC is a significant part of engineering education this waste of resources is set to continue. It is heartening to learn that parts of industry are learning (albeit the hard way) and the level of testing failures is reducing. Of particular note is that there is an international group working on EMC emissions at the chip and board level. One of the motivating factors was that lowering emissions would increase their market penetration. It is good when EMC performance is seen as an aid to marketing that the topic will take its rightful place in the echelons of engineering.
From many different perspectives, EMC has become a key subject that poses many interesting challenges. With the drive towards the computer society, e-commerce and the expansion of radio services to support this, it is essential that the EMC issues are correctly resolved. It is no longer a backwater of engineering where the problems can be left to ‘experts.’ It is an area where every engineer needs to become involved.
EMC has come through many changes, and is destined to go through further changes in the near future, yet there is a danger that in the heat of today’s production line pressures, this need to look forward is disregarded. Already I hear only too often the line, ‘Yes EMC is important to my company, but I have to do any EMC activities in my spare time.’ If you have been agreeing with me so far, then perhaps it is not you that should be reading this article, but I am certain that there is somebody close to you that should read this too.
 |
Peter J. Kerry has been with the Radiocommunications Agency and its predecessors for some 30 years. During this period he has been involved with a wide range of radio systems. For the past decade, he has been responsible for EMC standardisation, and is currently the chairman of the international EMC standardisation committee (CISPR). This article is based on his keynote address to the 1999 IEE York EMC Conference, and is intended to stimulate the debate on improving the management of EMC issues. The Radiocommunications Agency is an executive agency of the UK's Department of Trade and Industry and is responsible for the radio spectrum in the UK. |
This article also appeared in the April 200 edition of "Electronics & Communication Engineering Journal".
April 2000 |
|
