Is EMI going away? Well it won’t actually decrease to zero but EMC engineering may well become a back- water in electronic engineering. In the first place there is the declaration of good work instead of the EMC test. Strike one. Then there’s the micro-minaturization. Strike two. And finally wireless. Strike three. Are we out? Is that the old ball game?

The test labs are in trouble because it costs a few thousand dollars a day minimum for an EMC test. For a few thousand a month you can have a junior engineer transcribe an EMC book and throw a few specifics about your widget into a brief report. If there’s trouble, you go to court and convince the lawyers you did what you’re required to do. You’re off the hook even if what you did do is inadequate.

Next is miniaturization. At first we all thought that the proximity of various circuits would cause a plethora of EMI problems. It does cause lots of internal interference problems but those are “design” problems. The other effect is shrinking circuits to make the runs shorter and the currents smaller. We’re talking micro amperes and nanometers. The fields are so small it is difficult to detect them even if you put a loop probe next to the chip. And the clads are themselves a few tenths of a centimeter long. Because the pulse rates are a few mega bauds the runs must be micro strips. The fields are pretty much confined to the dielectric near the conducters. The PCs are now the size of a matchbox which you set on the table and push the button on the automatic test gear. The problem is making sure the matchbox is turned on since there are no emanations.

There was a time, not so long ago, when the printer was a few meters from the central processing unit and data were sent back and forth on an RS-232 line. Even then it was seldom that you saw interference from the units, but the cables could be heard a block away. It was important how you arranged them and bundled them during the test. Now there is wireless (no cables) and even if you need to have cables there are ferrites (no cable fields).

Wireless means electromagnetic field communications, but because of the frequencies there is nearly no interference. Many of them use infrared. It doesn’t interfere with other circuits because of orthogonal digital coding. There is some RF in the GHz range but there isn’t much up there with which to interfere. Perhaps more importantly these electromagnetic emissions interact with dielectric and diatomic molecules, oxygen and nitrogen, for example, as well as water vapor, plastics, etc. Since these materials are nonmetallic, they have “large resistances” and so turn the E&M energy into heat. They are self-limiting, they propagate around a room and in some cases even around a building, but they are attenuated below the background fields before they can go more than a few tens of meters.

There will still be electro-mechanical equipment which must be large enough to fit with humans, i.e., a printer large enough to handle 8.5 by 11-inch paper and the motors powerful enough to move the paper and a pen. But the amount of EMC work involved with terminal equipment will not support much of a technical community.

It is possible but unlikely that a whole new genre of EMC problems will develop to rejuvenate the discipline. If so, it will probably be susceptibility with medical implants which need to be comparable in size with parts of the human body and may use microvolt and microampere signals which will be subjected to fields of tens of volts and will be very difficult to shield without interfering mechanically with functions in the body.