This question popped up on one of the email lists that I’m a member of.
“Calling all labs – In your experience how many products pass the Unintentional Emissions test first time?”
I thought I’d share my response:
Speaking from my own experience. Over the last four years of running a consultancy, pre-compliance and low cost test EMC laboratory I would (very roughly) estimate that around:
- 50% of products pass their desired radiated emissions limits without any modification
- 33% or less pass all of the applicable tests first time without modification
The major caveats and notes here are that
- These figures are for customers products where the EMC performance is not known before testing. We do a lot of work helping people solve existing EMC problems but we are not counting this in these figures.
- Most of my customers are smaller businesses that can’t afford to employ an engineer to just look after compliance. That job role is either split amongst several people or the engineer in question has to look after quality, manufacturing, sustaining, thermal, system, and everything else. Speaking as someone who has designed many products and systems in the past, trying to design for functionality whilst simultaneously considering best EMC performance is HARD.
- The products that pass first time generally fall into one of three categories
- Products that we have design reviewed before the design was finalised
- Retests of products that have already been through our lab once
- Products that are very simple in nature
- Our hit-rate at being able to solve our customers problems is around 90-95%
- The “ones that got away” where we were unable to help deliver a compliant include
- No action taken: Products where it was deemed by the manufacturer not economically feasible to modify the product (e.g. product going end of life)
- No further communications from the manufacturer so we don’t get to find out what happened next (no news is good news, right?)
I would echo the sentiments of others on this thread regarding the need to design in compliance from the start.
One of the problems with the field of compliance is that it is too often “learned through experience in industry” and not explicitly taught. When it is taught at academic level it is often a surface treatment with a theoretical look at shielding or maybe crosstalk with no other practical context or background.
The split between industry and academia is one of the possible causes. Yes, there are exceptions to this but they primarily remain exceptions. I had discussions with a local university about some guest lectures on compliance and the theme of the response was “it doesn’t really fit into any of our modules” and “we can’t have it as an optional lecture as none of the students will attend”.
The number of times I hear “oh, thanks for that. No one has every explained it that clearly before” is worrying!
That’s my rant, hope it was useful