A few pages further into ENR George Kunkel makes the point that we should use the same analysis of shielding as the rest of the engineering community. Hear! Hear!

However, the situation is much worse than George painted it. In the first place most of the EMC problems are caused by terms that are not in the equations, effects that are not in the sophomore EE texts, incidental circuits that do not show up in the schematic and that are analytically intractable. Even if you can get the initial data needed they cannot be solved in closed “general” form. That is why they are omitted from the EE texts. To some extent computer modeling has allowed the “particular” solution of some EMC problems but it is difficult to get the required input data for the programs – Garbage in. Gospel out.

So what’s an EMC engineer to do? WAG is the answer of course. Oh, many an EMC guru finds some of the problems some of the time, but you can’t find all of the problems all of the time. In fact, you often can’t give a nice tidy engineering argument to justify your opinion drawn from 30 years of experience. Or is it three years of experience ten times? Or you can get a more nearly justifiable although less likely to be correct answer by using the ever- popular rules of thumb. Like the one that says you get the terminal voltage of a loop antenna by multiplying the field in volts per meter by the area of the loop (in meters squared). Did it ever occur to anyone that the result, whatever it is, cannot be the terminal voltage since its dimensions are volt meters. Oh well, just put it in dBs and we can ignore the dimensions.

We like to think our engineers think our area of expertise is black magic. They see units they don’t understand, answers gotten by using simple-minded equations with no derivation from first principles available to support them and which many of us don’t even understand. They see us using far-field equations at a meter for our antennas that are several meters in length and they know that it is definitely a near field situation, and they just shake their heads. I’ve got news for you. They are being polite when they tell you it’s black magic. They really think it’s bull. And they are more nearly right than most of us.

They put up with us because some government drone in the bowels of the Pentagon wants to show the manufacturers who’s boss. Or some flunky in the EU wants to make sure there is a trade barrier in place for equipment from out of town. To collect the money for the contract or to be allowed to sell their widget in Lower Slobovia they need to pass some silly tests or write some silly report or both. By the way, neither of them has the slightest bit to do with the proper performance of the equipment except that once in a while they hamper its efficient operation and they always increase its price. They put up with us because we administer the test and know the proper buzz words to put in the reports – dBs, ground planes, crosstalk and all that irrelevant stuff.

All of this notwithstanding, George is right. We can go a long way toward cleaning up our act from their point of view by using their units. For example, if the antenna factor were dumped and the effective length of the antenna in meters were used, then the lights would go on: a field of x volts per meter illuminating an antenna with an effective length of y meters produces xy volts. They would understand that and see that we are reasonable people.

One giant step forward would be to eliminate dBs. They have no practical use anymore and are an artifact of the days when Edison batteries and 600 ohm balanced lines were used in those new-fangled tele-o-phones. Are you sure you have to add because you don’t know how to multiply? The cost to the EMC community of dBs is orders of magnitudes larger than their benefits. In the first place we confuse our customers using exotic unnecessary notation in our reports and test reports. This is not good. In the second place we confuse ourselves too. Since the units disappear in the common usage of dBs we can’t immediately check our calculations and formulas using dimensional analysis, and that isn’t good either.

Changing from dBs to microvolts, microamperes, microvolts per meter and microamperes per meter would be nearly a no-brainer for us. You change nothing but the notation on the left side of your graphs so that 0dB becomes a microvolt, 10dB ten microvolts, 20db 100 microvolts and so on. Notice I didn’t put reference voltage in the dBs. That wasn’t a mistake. How many times have you heard that in conversations with the customers, and how many times have you seen it before in print?