There seems to be a problem in the halls of Government – well, several in EMC which may be related to the fact that all the people in power in DC are lawyers. Oh, there were problems when there were a few engineers to help out, but now it borders on silly.

The powers that be are going to allow, no, encourage power companies to use their wiring to connect hackers to the high-speed Internet. Now, to a lawyer and a taxi driver this seems like a good idea, but in Germany they tried it and it doesn’t work. Since our servants in DC think everything can be solved by adjudication, why bother finding out what happens when someone tries something? The fact that the radio noise will obliterate AM radio and make it impossible to receive shortwave broadcasting up to about 15MHz doesn’t matter since people are using the Internet to communicate.

I am at a loss to figure what’s going on with government requirements on EMC labs. A manufacturer no longer needs to have any EMC measurements made on his digital communications jammer. He just makes a declaration: “Well, I declare.” Can you believe that his widget won’t make excessive noise? If you are foolish enough to want to make measurements and silly enough to want to be in out of the rain when you make them, you must be sure that the measurements are within .001dB accurate. You must, because you have a roof over your “open field site,” make four sets of measurements of “site attenuation,” average them and compare them to some Never-never land curve.

Of course, you’re fine if you come within 4dB of this curve. Presumably since the site attenuation is actually a measurement of a bogus EUT (the transmitting antenna), your measurements on a customer’s EUT will be 4dB different from what they should be. This assumes the curve corresponds to something that goes on in the real world.

But worse, say your lab is 4dB low and mine is 4dB high. I measure 2.5V/m, you measure 1V/m. That’s close enough, right? I mean who cares if they hook their TV up to 120V or 300V when they plug it into the wall.

I guess it makes sense to accept a declaration of conformance. It might be just as good to have an experienced EMC engineer make a guess about the radiated emissions from an EUT as to rely on measurements made on a site with a 4db excursion from “ideal site attenuation.”

At one lab recently I was in a 10M semianechoic chamber where the local EMC guru had placed a couple of pieces of sheet metal a few decimeters on a side at judicious places in the room to “bring the room into conformance with the ideal.” He said the room construction crew had suggested he do it. At first I was convinced that this was against the rules of both the specification writers and God, but after a few beers it occurred to me that although it probably was against the rules, it probably was the best way to make emissions measurements.

Suppose you had to set up your measurement gear using combinations of sheet metal and billboards covered with tiles, then bring your site attenuation into conformity with the ideal curve. Say you had to tweak it to within 1/2dB high or low. What would happen? Well, it would cost you a few thousand dollers to do it, but the EMC community would get better answers. If your lab is 1/2dB low and mine is 1/2dB high, then I measure 1.06V/m and you measure 1V/m, about a 6% error. Now we’re talking. Not only that but we might save at least half of the $200,000 sites that don’t conform to the standards. But what do our servants on the Potomac care about money – our money?

I also can’t figure what’s going on with cell phone specifications. Although every day they interfere with such life-threatening equipment as garage door openers, cell users may be allowed by the FCC to make calls on commercial airliners to relieve the boredom on long flights. I for one will not be bored if they’re prattling on a flight I’m riding. It will be white knuckles all the way.