More than 40 years ago we would sit for hours in a shielded enclosure in front of a test sample with an Empire Devices NF 112 on a cart beside us. Everything was turned on and hooked up, but we had to keep checking because nothing happened. We had headphones so we could hear any signals or noise and kept our eyes glued to the meter but the needle never moved. We tuned slowly through the EMI receiver’s 9GHz of frequency coverage because of the meter’s mechanical integration time constant. We dutifully changed antennas and each of the four tuning heads, waiting the required 20 minutes for the head to warm up before we continued the scan. In those days no one I knew ever heard any interference above the limits. In fact, without a low-noise preamplifier no one heard anything but the soft hiss of the receiver’s noise floor.

We wasted thousands of hours and tens of thousands of dollars and learned nothing we didn’t know beforehand and reported nothing in the test report but two readings of the noise floor in each octave. This was particularly stupid since the testing was all done to MIL-E-6181D which required a test plan and you could make a case for exempting everything above 1GHz. Also, after the military saw the data (much of which was above specified limits) they would routinely issue a waiver. We have learned a lot since then. Right? Nope – we are still performing unnecessary measurements on equipment we know will pass the test.

Sitting next to my laptop is a TV remote, a plastic box half the size of a pencil box festooned with buttons of various shapes and colors with nearly no words on it but lots of runes. Sorry, I don’t read runes so I need to wait till my grandson comes around to turn on the TV. The thing has no input lines, no output lines (it communicates with the TV using infrared LED emissions), no control lines and no power lines. It operates on two AA batteries for at least six months so currents on the necessarily short conductors must be minimal (after all, they have to fit in the box). I’d be willing to bet that the thing was set on a table in a 15-meter chamber and tested for emissions a la CISPR-22, and to everyone’s great surprise it passed. In fact, the EMI meter printed out a graphical record of its noise floor. Since it would be used in Minneapolis in January by a guy in leather slippers shuffling across a nylon carpet, it was tested by design engineers using a piezoelectric charcoal lighter to make sure you couldn’t zap it into Never-Neverland. This also made it unnecessary to test it to the requirements of IEC-61000 but they did.

Some similar devices (calculators, notebooks without RF input, etc.) have battery chargers. Oh no, not a power cord! Now it HAS to be tested. Hogwash. First of all, the thing is charged once in a blue moon, and, second, if it does interfere, you simply plug it in on the other side of the room. There should be a written exemption in the specks for hand-held, battery-powered equipment that either does not connect electronically with the outside world or uses ultrasound or infrared to communicate. Time and money are wasted to perform EMC tests on them.

Of course, one can make a case for testing cell phones, blackberries, dingleberries and iWHATEVERS. Most of them use RF to talk with base stations or other hand-held equipment and so are listening for RF noise and may react unpleasantly to it. In addition, they may holler “Alli-alli-outs-in-free” and mess up some system in the vicinity, so it makes marginal sense to run tests.