> On Sun, 27 Jul 2003 09:43:34 -0700, you wrote: >=20 > >The easiest way is to do the process using one of the many safety/EMI > >testing labs. > > > >You can go with the big ones such as TUV (http://www.tuv.com) and UL > >(http://www.ul-europe.com/en/global/) or a smaller one (we are using > >http://www.emccomponent.com/). > > > >These guys live this stuff daily and knows everything about the > >testing and certification process. > > > >The complexity of the process depends on the nature of your product, > >a pace maker for example will be more costly than PIC Programmer ;-). > > > >Call their sales people and ask about the process and the cost. >=20 > And when you have got up off the floor, remember that for certain > types of product, the cost can be reduced to close to nothing if you > are confident that your product meets the requirements of the various > applicable standards. e.g. if your product is a 4MHz PIC on a well > laid out PCB, well decoupled, with minimal scope for radiating (unduly > long wires etc.) then testing for radiated emissions is likely to be a > waste of time - you'll be lucky to be able to detect it far less > measure it. Even for self-certification you have to be able to=20 prove that the product complies. You must maintain,=20 for a period of ten years, documentary evidence of the=20 basis of the certification.=20 The best proof is a protocol from the various tests=20 and measurements, applicable for the product. Even a 4 MHz PIC can radiate way over maximum allowed=20 limits, with a bad layout. The EMC standards does not actually say what the=20 documentary evidence must be. If you previously have=20 done a similar product, have tested this, and found=20 that it is well within the limits for compliance, you=20 can perhaps use the test reports from this product,=20 and state the similarities between the two products. >=20 > Remember that test houses are in the business of selling their > services, not telling you how little testing you can get by with, >=20 Also remember that if you find that one test house is=20 not working with you, you can always change to another=20 test house. They know this as well. For a normal industrial product, with not too many=20 different types of inputs/outputs (the conducted=20 induced RF 150kHz - 80MHz, can take a long time with=20 lots of cables), we normally use 1 to 2 days at a test=20 house, at a cost of about 2000$-4000$. Regards / Ruben =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D Ruben J=F6nsson AB Liros Elektronik Box 9124, 200 39 Malm=F6, Sweden TEL INT +46 40142078 FAX INT +46 40947388 ruben@pp.sbbs.se =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics