COTS Issues Relating to Systems Integration
COTS Issues Relating to Systems Integration
Ken Webb, TÜV Product Service
Published: July 2006 - EMC Compliance Journal
It is some time since the EMC Journal discussed the risks of integrating COTS equipment into host systems based on the erroneous assumption that CE + CE = CE. It is worthwhile revisiting this topic in the light of experience gained in the interim of how system integrators are actually coping, particularly when they are adopting the Technical Construction File (TCF) route to compliance under the EMC Directive.
When there is a requirement to demonstrate compliance with the EMC Directive for a large system, it can be difficult to fully EMC test the system due to size constraints. Therefore, in many cases, partial EMC testing is performed in-situ, either at the final point of installation or at the manufacturer’s premises. Due to the limited information provided by partial EMC testing of this nature, the case for compliance with the EMC Directive rests heavily on a combination of the three key elements shown in Fig. 1.
Figure 1: Three key elements of EMC compliance
The information and data provided by these three elements is normally pulled together within a technical rationale or technical argument, forming part of an overall TCF. This article focuses on one of these key issues, the information contained on the supplier’s Declaration of Conformity (DoC) supporting the claim for compliance of COTS equipment.
Can system integrators rely on a supplier’s Declaration of Conformity?
When in situ EMC testing is the choice, system integrators will place greater reliance on the conformity of individual items of COTS equipment in order to overcome the limited data provided by in-situ testing. However, the use of COTS may pose compliance problems, particularly where the EMC standard(s) and test limits and levels applied are different to the final intended EM environment of the system.
Commercial EMC standards are organised into four main categories: basic standards, generic standards, product family standards, and product specific standards. The product specific, product family and generic standards contain limits and levels and are harmonised and published in the OJEU. The basic standards generally specify test methods and are, by definition, independent of any particular product. Basic standards are harmonised (for example, the EN 61000-4-X series) but are not published in the OJEU, the exception being the EN 61000-3-X series.
In TÜV Product Service’s experience, most suppliers of COTS equipment list only the applicable product specific, product family or generic standards, as shown in Fig. 2. There is nothing legally wrong with this approach, as reference to Regulation 42 (2) (f) of the UK Statutory Instrument 2005 No. 281 shows. This includes a requirement that the DoC ’shall state the numbers and titles of the applicable EMC standards applied by the manufacturer’.
Figure 2: Typical COTs supplier DoC
However, if the ’basic’ standards are not listed, this presents systems integrators with a problem, namely, which tests have been applied under these standards?
For example EN 55024 specifies 14 individual immunity tests grouped together on a port by port basis. In order to compare the EMC performance of COTS equipment with the intended EM environment of his equipment, the system integrator will need to know which of the 14 tests specified in EN 55024 have been applied.
In an ideal world system integrators would have access to their supplier’s EMC test reports, but experience shows that a request such as this is normally given short shrift by COTS equipment suppliers. However some companies are making a stand on this and are taking a firmer line with their suppliers. The following case studies represent two extremes adopted by system integrators.
Case Study No. 1
Company A started to get a feeling that all was not well when in situ EMC testing of its large systems revealed problems with the EMC performance of COTS equipment and invited a technical consultancy organisation to examine its processes. The consultant’s findings revealed a series of shortcomings, starting with purchasing.
Company A was taking a relaxed view of the compliance issue when appointing suppliers, failing to demand absolutely cast-iron documentation prior to placing an order. It accepted vague claims by suppliers that the COTS equipment was ’compliant with all relevant Directives’ without requesting details of the standards to which it had been tested, the precise test levels and limits applied, and the test results. Of course, suppliers are not obliged to divulge results, but the question has to be asked whether it is safe to do business with a company which refuses to cooperate with its customers.
Goods inward was not checking supporting documentation in relation to CE marking compliance, limitations on the use of the product, and requirements for measures such as screened cables, additional filters, etcetera.
Without this information, designers were fending for themselves when it came to system integration. Even if the EMC performance of COTS equipment is adequate, integration with respect to types of cable, routing/segregation of cables, and positioning and mounting of equipments in terms of segregation of sensitive elements from gross emitters is important.
Company A was discovering the effects of its negligence when it came to compliance testing, and the additional information it should have demanded from COTS equipment suppliers.
Case Study No. 2
When Company B purchases COTS equipment, its EMC performance has been rigorously analysed and documented. Each supplier to Company B is contracted to deliver an EMC profile including manufacturer/contact data, a description of the equipment, interface information, requirement for cable and cubicle screening, frequency utilisation, compliance with EMC standards, countermeasures and specific design guidelines, potential critical issues, and block diagrams. This process can go as far as validation of suppliers’ EMC test reports and evidence of EMC compliance. Furthermore, taking a strong line which must be applauded, Company B does not do business with suppliers which decline to provide all the information required.
The information supplied is fed into a matrix enabling risk to be assessed at host system level. This leads to an overview of sources and receptors of EMI, identification of locations where particular precautions are required, detailed assessment of these critical points by calculation, simulation or estimation to find countermeasures, definition of design and mounting rules and cable routing, and issues related to maintenance for ongoing EMC.
What can companies do to improve system integration with respect to EMC compliance? Firstly, they can open up lines of communication so that its people share information and experience. There is particularly a problem with some large, project-driven companies where project teams work in isolation, with no common practices. If one project team has adopted a rigorous, efficient, and successful COTS integration process, other projects teams probably won’t know about it.
Secondly, companies can raise awareness of the compliance issues from purchasing onwards and ensure that procedures are in place to assess the information provided by suppliers. Procedures may include, for example, specifying exactly what documentation must be produced - at the very least a DoC plus installation instructions - and ensuring that test standards, levels and limits are in line with the requirements of the target system.
Sometimes companies will come to the conclusion that they cannot get adequate information from suppliers, and know they never will, no matter how hard they try. In this case, a standards gap analysis is required, together with a ranking scheme and an assessment of the design of the COTS equipment in terms of how well or badly it is engineered and built. The gap analysis may suggest mitigating measures, which can be implemented to achieve a smoother road to compliance. If doubts remain the only option may be to test incoming COTS equipment.
Good processes mean lower long-term costs
Putting the right procedures in place right at the start will go a long way to ensuring that system integrators do not ’inherit’ their supplier’s compliance problems. The wrong approach to the procurement of COTS equipment increases the risk that a system will fail EMC testing and may need to be re-engineered and re-tested. In the worst-case scenario in-situ testing may not pick up a problem, and companies will then run the risk of attracting all the penalties of putting a non-compliant system on the marketplace.
The on-cost can be minimised by seeking professional assistance. Both Company A and Company B did;, the first because problems were apparent, the second to make absolutely sure it had put together a robust process for ensuring the required level of EMC compliance for COTS equipment. Professional consultancy is a cost-effective way of bringing core competence and experience to a project and, as the case studies have shown, can play a major part in helping system integrators develop a plan for managing the integration of COTS equipment, performing a standards gap analysis and, identifying process improvements where required.
