EMC Engineering: What Does It Mean?

I've mentioned before (and stated on my resume, which is linked in the sidebar) that my primary area of professional experience is that of Electromagnetic Compatibility (EMC) engineering. At present I am (following a major plant-closure layoff where I used to work) part-timing as a lab hardware technician, but of course wanting to maintain my electrical engineering and EMC brain-connections during this rather unusual period of my life. Moreover, the whole reason I ended up specializing in EMC was because of how interesting it is -- and I expect more people would find it interesting if they actually knew about it. Thus, this post!

Anyway, the main objectives of EMC engineering are:

1) To prevent electrical/electronic devices (and their subsystems) from interfering with proper functioning of other devices (or subsystems within the same device), and

2) To design, build, and/or modify devices such that they are not unduly vulnerable to incoming interference (or self-interference at the subsystem level)

What amazes me is that I did not even know about EMC as a discipline* until I was already out of school and working as an (unspecialized, at the time) junior electrical engineer. After all, pretty much every device that enters the market has to pass an EMC test. Making sure a device will meet EMC requirements is a whole other layer of engineering beyond just having the device do what it is supposed to do from a functional standpoint.

That said, most of us living in cultures of electronic ubiquity likely have first-hand experience with EMC issues. My father was an avid amateur-radio hobbyist when I was little, and sometimes the signals associated with that equipment would end up coming through the television and random sets of speakers throughout the house. Sometimes it would just be audio interference, but (in the case of the television) visual interference was also observed on multiple occasions.

At the time I was mainly irritated at having my Zelda dungeon-crawls interrupted by bursts of scramble and garble, but part of me was also fascinated. Somehow, even though these connections were not overtly visible, all the electronic devices in my environment were interacting with one another on their own terms. Which is a notion I've had to revisit many times in the process of troubleshooting (and working to prevent in the first place) EMC issues at work.

The very nature of electromagnetism means that the only way to have a truly 100% interference-free environment is for no devices to be functioning at all -- which of course defeats the purpose of having them to begin with. Thus, EMC engineering must be performed along the lines of both optimization and compromise. I.e., in designing or modifying a device (or system of devices) you want to be able to get maximum performance while minimizing the chance of problematic emissions or vulnerability.

Some examples of situations entailing application of EMC principles are as follows:

- Many digital circuits/devices perform better when the "edge" of the square wave comprising, say, the clock signal is more defined. However, since square waves are generated via Fourier series implementation, you can end up hurting EMC performance if you simply attempt to make the edge of your signal as "sharp" as possible.

Waveforms with more gradual "edges" (such as sine waves) are much more EMC-friendly in this regard than square waves; the lack of steep/discontinuous slopes means fewer harmonic components at problematic amplitudes. There are some devices on the market that actually can operate with a sinusoidal clock, however, this is certainly not always going to be the case. Therefore, a design optimized for EMC and data integrity will often need to establish some compromise between a "nice-looking edge" and a signal whose harmonic components are within reasonable limits for EMC performance. This might entail anything from changing a resistor value (to alter a time constant) to changing the manner in which the data is transmitted (e.g., implementing differential conductor pairs).

- Say you have an electronic device housed in a metal enclosure or chassis. If the housing is intended to provide shielding as well as protection from dust and mechanical injury, etc., then you are going to have to consider such factors as "how large can I make the ventilation holes without compromising shield integrity?" You are also going to have to make sure and specify what parts of the enclosure and any associated fasteners should be left free of paint, because in some cases the circuit ground needs to be able to make good contact with the enclosure.

- A new circuit board is being created. Not only does the right number of layers to accommodate all the relevant signals need to be determined, but also the arrangement of those layers. E.g., where should the ground planes be located in relation to the signal planes?

- You install DSL internet in your apartment and afterward hear an obnoxious buzzing whenever you go to use the phone. Usually internet providers will include filters with your modem package, but if not, you can probably find them at a local electronics store or order them online.

Of course there are many other examples I could list here, but this is meant to be an "introduction to the subject" post so I will stop with those. The point of all this, though, is essentially to illustrate that while some aspects of EMC engineering definitely require a lot of deep technical analysis, there is also a very practical, everyday level on which EMC is relevant to pretty much everyone who is likely to be reading this (meaning, anyone who uses and/or lives in an area where multiple electronic devices are expected to peacefully coexist).

---

* Some may consider EMC to be a sub-discipline/offshoot of RF engineering, and it definitely overlaps somewhat with signal integrity as well.