CE certification

[ricardo]

In da 90's dem new-fangled EC directives caused a lot of fear among manufacturers.  But in fact even then, IIRC, you could self-certify.

The 2 main issues were the LVD (BTW Low Voltage means mains) and RFI/EMI.

You DEFINITELY want to test to the LVD (CE60065).  As Abbey says, its all sensible stuff and easy to do.  In Oz, any electrician has the necessary gear to do it all.  The important thing is to keep the 'File' which is the evidence that you've done such & such a test at such & such a time.

Though not specified, if you are a manufacturer, you want your QC records to show you test a proportion of your production regularly too.  At Wharfedale/LEAK, all production was tested for important parts of BS 415.

That's cos in the UK and probably most of the EEC, the Technical Manager (ie you), goes to prison if someone is hurt and there is only wonky evidence of testing.

I wrote an internal house note about how to design stuff so it could be easily modified to meet both the US UL type certification (emphasis on catching fire) and the EC type (emphasis on electrocution).  IMHO, CE60065 made things easier for makers.

The EEC RFI/EMI stuff was in fact far less stringent.  eg if your CD player stopped playing when zapped by RFI, your 'File' just needs to record the fact and the User Instructions just needs to tell him to press START again.  So self-certify away. 

If your CD player goes up in flames when zapped, it probably needs more work  ;D

There aren't any RFI standards to meet for important stuff eg microphones.  In fact some famous name maker's stuff have terrible RFI/EMI immunity.

YMMV if you're doing life-support stuff. 

For emissions, the US FTC regulations were far more stringent and probably need 3rd party testing if you've got computer stuff inside.

 

[ruffrecords]

It is impractical to invest in formal lab testing for one off, but one or many should be built using proper human safety guidelines.

When in doubt build it right.

JR

Interesting discussion. How about the case where a custom device consists of two parts? One a separate  power supply that could supply many custom variations of the 'main' product and the second being the customised main product such as a mixer for example. Presumably the power supply could be tested in its own right to meet the LVD and EMC. The 'main' part probably does not need the LVD at all if the supply voltages are low enough and there are no up-converters inside it. If it is purely analogue then the emissions part of the EMC should be trivial. The only concern would be to document how its performance might be affected when subject to interference.

Cheers

Ian