EMI/RFI Evaluation

[rlsharpe]

I have to perform an evaluation on the electromagnetic compatibility of some new components. Typically, the components have been tested to a standard and I can just compare that standard with our specs/requirements. In this case no testing was done and I am at a loss. I'm hoping someone can help me with how to "prove" the components are acceptable for use. I know they are I just don't know how to document it. components are all from Viconics: R820 SCR power controls; R810 power switching modules solid state relays; R850 step controller with integrated vernier stage; and C1025 electronic thermostat for modulating electric heat applications. this is a relatively new requirement and we are all struggling with how to do it. Thanks for your help.

 

[Noway2]

EMI/RFI can be difficult to test. You will probably need to isolate the device under test both from the (building) electrical system and have a "clean" environment. The last place I worked had a copper-mesh room that was a big Faraday cage for this purpose. The power was isolated from the building by a LISN - line impedance and stabilization network. They then used various antennas and a high end spectrum analyzer to make the measurements.

The test standards often times indicate, to some degree, how to perform the test, or at least what must be tested. If you haven't already, start by getting a set of the standards as well as any standards referenced.

If you are really at a loss, you could enlist some help from a nationally recognized testing labs, but you will pay big bucks for it.

 

[VEBill]

E3 testing to prove compliance (or fail) is extremely simple. You box up the units, ship them to a reputable E3 testing house, receive the report, and pay the invoice. That part is "easy."

Dealing with the inevitable failures takes months...

 

[rlsharpe]

Thanks but we don't want to get into actually having to test the components ourselves(or pay to have it done). Our hope is to be able to pencil-whip it away. We'll buy something else that has had compliance testing done if we have to; unfortunately in this case that's not an option. As i tried to describe most components will say they comply with IEC # or FCC Part 15 or something similar. We can then tie those standards back to our spec and call it good; not gonna happen in this case. We are a power plant and all we are trying to do is replace some failed components with the latest model. I have seen one project were they were able to say something along the lines of 'typically low power devices,' etc, and were able to justify it. i don't understand enough about the components i'm dealing with to be able to say any of that.

 

[VEBill]

Depending on the contractual circumstances, if you're lucky there's an evil Program Management approach that is good weapon. You price up the by-the-book full test program - probably low end of six figures in this case. You submit this price (including your mark-up), and the associated six month schedule, to the customer and ask for their authorization to proceed. The letter ends with another much cheaper option of just going ahead and trying it with perhaps a few hours using a scope to make sure the signals are clean.

Obviously this approach ("democracy is the theory that people know what they want and deserve to get it good and hard") only works in certain contractual circumstances.

But the schedule delay is what it is. The customer might argue who will pay the bill, but if the delay is a non-starter then it's a bluff.

Good luck.