Build or Buy EMC Testing?  

   

It seems strange at a time when there appears to be many EMC test services around and with extensive capability that a company would invest many millions of Euro's into building their own test facility, especially when they are not even a manufacturer, as in the case of Lufthansa Technik.  However, the reason for having your own test facilities can extend beyond the basic cost and that is what we thought we would consider in this short article.

   

Build Your Own Test Facility

 

The first thing that puts most people off is of course cost.  As a minimum we estimate 300,000 - 500,000 for a basic EMC compliance facility, but this is just the basic outlay, then you have to staff and operate the facility, purchase and maintain up to date technical standards, the capital equipment has to be regularly calibrated and again maintained.  So why would a company make such an investment?

  

One reason a company might have their own facility would be that they have a lot of new product development and innovation of existing products requiring a large number of EMC tests to be conducted per year.  If you have between over say product releases per year then maybe the cost of testing can be recouped after 5 - 7 or so years?  However, in our experience the main reason companies bring testing in-house is time; time-to-market, time for retest if something fails, plus the gain of knowledge from being able to try small changes and test their impact.

   

In the automotive supply world the VM's also often require regular product and process validation testing and this will include EMC (some commercial companies also do this as a matter of course).  Again this can increase the utility of an internal test facility, often one of the main reasons for not having an in-house capability.  With multi-national companies, additional revenue for a single business unit may also be obtained by offering the EMC test services to other business units in the group.  Occasionally, these EMC services are offered outside the group as contract services to other partners in the business, but this is rarer than might be expected as many with internal testing would rather maintain their privacy and do not want potential conflicts with internal and external demands on their EMC test resource. 

  

The learning gained from being able to test products in-house, even if only for single EMC phenomena, is difficult to cost but invaluable for the gain of insights into product design for EMC and ultimately in assisting the infamous "right first time" approach to EMC.  Twenty years ago an RF spectrum analyser would have been a rare and expensive piece of test equipment, but today most development laboratories have at least this as their own EMC test kit, if nothing else.  The value of some form of internal EMC testing is clearly recognised, even if no cost-benefit-analysis can be provided to justify it.

    

Buy in Test Services

    

There is no shortage of EMC test facilities in Europe and in most developed countries elsewhere.  These test service providers offer many benefits over internal facilities, even above initial capital outlay, for performing EMC testing.  For one they keep abreast of the most up-to-date testing methods, standards and equipment as part of their business.  They tend to have some of the best and fastest equipment, hence actual test times may be shorter than a basic in-house facility could provide.  They also often have complimentary services, such as vibration or multi-geographic compliance capability, that may in fact not be possible with in-house facilities.  Often one of the most important features external services provide is easy traceability to national and international standards and an independent (of the producer) compliance record.

   

It appears cost should not be an issue in favour of external service providers, prices at EMC test houses have been relatively stable at around 1500 - 2000 per day for some time, typical test plans may take 3 to 5 days to complete at a modern facility.  The main negative is booking time and obtaining test slots.  With all the best product planning in the world, advance booking for a products EMC compliance testing is a risky business.  Although some services have shortened availability, a 4 week time from booking to testing is not unusual, even longer for specialised testing or in-demand service providers.  This doesn't at first glance seem too long, but what if something fails, the time to re-work the product and retest can add massively to the development time, and often EMC testing is left until the last minute, despite it being one of the few legally required aspects of an electrical products design requirements.

  

The answer to the scheduling problems and risk of failing EMC at the last hurdle is to minimise the risk.  There are several easy and cost effective ways to do this.  The most obvious is to perform EMC testing on prototypes or first off samples, just to gain some indication of the total perfromance of a product.  It is surprising how little usually changes between prototypes and production units, its rarely as much as the designer will probably claim.  Another way to mitigate failure is to use existing EMC proven circuits and PCB layouts in new designs; a good interface circuit and/or power supply circuit will usually work with several different designs, don't re-invent the wheel in every new design.  Lastly talk to your service providers especially about failures or even new products, they might be able to suggest tests that are likely to be the most critical to your new circuit in obtaining EMC compliance, the chances are they may have seen other companies with similar circuits and know what to expect, you won't be their first customer with a Bluetooth and USB interface for example, even if never in a your application space.

  

Summary

  

There are always good reasons for both sides of this argument, I have worked in companies on both sides of this and although I always prefer in-house facilities I also know that these are rarely commercially justifiable, they are often a leap of faith by the management.  I have also worked with many good EMC test service providers and know that the charges they make are no indication of their quality.  The most important item I would suggest looking for in a test service provider is a stable workforce, if their test engineers have been there for some time they will know their business and at the end of the day, as ever, it is the people you have doing the testing that will make all the difference, whether in house or from an external service provider.

 

    


www.AutoEMC.net 2008                                                           TOP OF PAGE                                                                 HOME