A Premier on Electromagnetic Interference (EMI)

The June 25, 2009 edition of Electronic Design magazine has a great four page primer on electromagnetic interference (EMI)- The Dark Force Of Evil In Electronics: Electromagnetic Interference.

The basics of EMI for non-engineers:
EMI is unwanted signals coming from (and occasionally getting into) electronic equipment. The most common type of EMI is noise from high-speed digital lines, i.e. lines that communicate a series of 1s and 0s. If the transition between 1s and 0s happen very fast and/or the wire carrying that signal is very long, the wire will radiate signals that can interfere with other electronic equipment. Signals normally travel down a wire at about 180ps (that’s 180 trillionths of a second) per inch. So if your wire is one inch long and a transition from 0 to 1 takes 1,000ps, you can treat the wire as a single element. If your wire is 10 inches long, however, one end of the wire can have completed its 1,000ps-long transition before the other end has seen even the beginning of this transition. A wire with a different signal on each side can act as an antenna transmitting unwanted signals. (This is just one possible cause of EMI.) A lot of electrical engineering is making sure high speed signals get from one place to another without being distorted or causing interference.

I like how the article begins with what sounds like engineers’ excuses when their product fails EMI testing:

Is there an electronic product or circuit that’s not susceptible to electromagnetic interference (EMI)? For that matter, are any devices EMI-free? Simply put, no.

Come on, nobody's perfect, hey?

On a more serious note, I like how it mentions that static discharges can cause EMI in a product (rather than the usual case of a product just emitting unwanted signals):

Don’t forget electrostatic discharge (ESD) as an interfering source. ESD, of course, is the momentary current flow that occurs when a high voltage between two points is dissipated. Lightning is the most powerful example, but any static discharge can produce EMI and damage unprotected circuits.

I experienced this the hard way last year on one of my projects.

A premier like the one in Electronic Design should be required reading for EE students.