Shielding and filtering are not independent of one another

14th August 2018
Posted By : Alex Lynn
Shielding and filtering are not independent of one another

In this article, Keith Armstrong, FIEE/IET, Senior MIEEE, ACGI, EurIng(Gp1), on behalf of EMC Standards, outlines some of the issues he has encountered after having been asked to fix EMC problems, and how designers can learn to ensure that filtering and shielding are compatible.

I have lost count of the number of times I have been asked to fix an EMC problem, only to find that a shielding box has been designed or purchased to provide XdB up to fMAX. Or a filter has been designed or purchased with a similar specification, but to reduce cost the filter has been mounted on a Printed Circuit Board (PCB) inside the box, with a cable from it entering or exiting the box through a plain connector.

Of course, we want to reduce the Bill of Materials (BOM) cost, but at the expense of having to respin the design at least once and losing sales through delaying the time-to-market? I don’t think so!

The cable entering or exiting the box through a plain connector is simply an ‘accidental Radio Frequency (RF) antenna’ that couples unwanted EM noise between the inside and the outside of the shielded box. This has little to do with the signals or power that the cable is meant to be carrying but has everything to do with the fact that all conductors, whether electrical, mechanical or whatever, never mind their net names, never mind whether they have green insulation with a yellow stripe, always interact with their EM environments.

When filters are tested by their manufacturers, they are first carefully installed in absolutely the best shielded enclosure that their manufacturers can afford. The upshot is that filtering and shielding have to be designed together, as a system. They simply cannot provide their shielding or filtering independently of each other.

The sad and stupid thing is that no filter manufacturer ever provides any information on what shielding is necessary for their filters to function as specified. So, I guess it is understandable that designers who haven’t yet learnt the hard way, the very costly and time consuming way, will assume that shield and filter specifications will be met by the shield and filters on their own.

SPICE and similar circuit simulators assume that electric and magnetic fields don’t exist, so they incorrectly tell us that a filter will achieve whatever its specification says, or circuit design achieves. Such over simplistic simulators completely ignore what will actually happen in real life unless appropriate shielding is integrated with the filter.

I can only hope that some designers will read this brief article, take the trouble to learn how to design shielding and filtering to work together to get the best out of both (takes a few hours, tops).

I hope they never have to repeat the horribly costly and time consuming learning curve that so very many designers have gone through before them. There’s simply no point!


You must be logged in to comment

Write a comment

No comments




Sign up to view our publications

Sign up

Sign up to view our downloads

Sign up

European Microwave Week 2018
23rd September 2018
Spain Ifema Feria De Madrid
Connected World Summit 2018
25th September 2018
United Kingdom Printworks, London
IoT Solutions World Congress 2018
16th October 2018
Spain Barcelona
Engineering Design Show 2018
17th October 2018
United Kingdom Ricoh Arena, Coventry
Maintec 2018
6th November 2018
United Kingdom NEC, Birmingham