I recently dropped in on a meeting of my local (Seattle) chapter of the IEEE Electromagnetic Compatibility (EMC) Society. This is not my usual hangout, but I can highly recommend it to digital people like myself as a way to learn more about EMC in general, and also as a cheap way to collect lots of free advice.
Anyway, after a charming lecture by Bill Ritenour on the need for electrostatic shielding at gasoline pumps, we turned our attention to the problem of how best to teach EMC concepts to persons with a pure-digital background. As a result of this discussion, and much thought afterwards, I am now finally able to put my finger on the basic reasons why many digital engineers have such a difficult time dealing with EMC problems. Contrary to the opinion of some in the analog world, it is not because they are dumb (far from it). Neither is it because they didn't study hard enough in school (most did). It actually has little to do with the individual engineer at all. The underlying, root cause of many of our present-day difficulties with EMC is a matter of attitude: digital engineers don't believe in EMC. This unfortunate situation has been brought about by a confluence of circumstances. Our educational institutions, our vendors of instrumentation, integrated circuits, and simulation tools, and the lackluster performance of some of those in engineering management all share part of the blame.
Without intending any harm, our institutions, vendors, and managers have propagated five great misconceptions which prevent many, if not most, new digital designers from understanding EMC at any level, and, in fact, from even believing in its existence. To the new digital engineer fresh out of school, EMC is, at best, a myth.
The better you understand these five great misconceptions, the better equipped you will be to understand the point of view of many digital engineers, and to help them overcome the EMC difficulties they will inevitably face.
1. Digital engineers don't believe current flows in loops
Look at a digital schematic. Consider the logic nets that carry digital signals from gate to gate. We all know these signals are propagated in the form of electron currents, and that these currents always flow in loops--yet, on the schematic, the paths for returning signal currents are not shown.
Many digital engineers believe the return paths are irrelevant. After all, they reason, if the logic drivers act as voltage sources, and the inputs act as voltage receivers, why worry about the current? This great misconception is reinforced by manufacturers of oscilloscopes and logic analyzers who primarily market voltage-mode probes. If we had good current-sensing probes with a pinpoint proximity-activated head small enough to see the current flowing on an individual BGA ball, the flow of current would suddenly become a "reality" for many engineers rather than a merely theoretical concept.
If you are going to work with a digital engineer on a common-mode cable radiation problem, for example, first establish whether the engineer really understands that current really does flow in a loop.
2. Digital engineers don't believe in the H-field
I attribute this misconception to our educational system, with its disproportionate focus on electric field effects, as opposed to magnetic. This is a relic of the tube era, which was characterized by very high-impedance circuits. For example, the plate circuit of a tube might have an impedance of 100,000 ohms, much higher than the impedance of free space, (377 ohms). Therefore, most of the near-field energy surrounding the plate circuit would be in the electric field mode, and most cross-coupling and parasitic coupling problems would involve electric-field, or capacitive, effects.
Today's high-speed digital systems have low-impedance circuits, near fifty ohms, much lower than the 377-ohm impedance of free space. Most of the near-field energy surrounding a digital circuit is in the magnetic-field mode, not electric. Therefore, most crosstalk, ground bounce, and interference problems in high-speed digital systems involve loops of current, magnetic fields, and inductance.
In the EMC world, it is common knowledge that the near-field energy surrounding a digital board is mostly magnetic. Digital people don't know about that.
3. Digital Engineers don't believe gates are differential amplifiers
On a typical product datasheet the input voltage sensitivity is rated in units of absolute volts. It is not clearly stated that the gate responds only to the difference between the voltage on the input pin and whatever voltage happens to be present on its designated reference pin. Nor are we clear about which is the designated reference pin. (For TTL it's the most negative power rail; for ECL it's the most positive.)
This ambiguity leads many engineers to think that a gate can sense "absolute zero" volts, as if it had a magic wire leading out of the chip to the center of the earth where it could pick up a "true" ground reference potential. As a consequence, they fail to comprehend the difficulties that arise when the ground voltages at two points in a system are unequal.
This is a case where digital specsmanship has not served us very well. Of course, no vendor wants to admit that their chips are susceptible to ground shifts, so we can't expect them to talk much about it. On the other hand, we all need to understand that those system architectures which permit large ground shifts between chips are likely to malfunction in addition to generating massive amounts of EMI and falling subject to ESD and other immunity problems. This is serious stuff.
You will find that most inexperienced digital designers have spent little time thinking about the existence of different ground potentials in their systems, the effect that would have on performance, or the mechanisms that create ground shifts.
4. Digital engineers don't believe in electromagnetic waves
Despite the many obvious examples of electromagnetic fields at work (like microwave popcorn and television), many digital engineers do not believe these effects happen inside digital systems. The root of this belief is that waves do not exist in Spice. We have trained a whole generation of circuit designers to believe that the world of Spice-based software simulation is a manifestation of real circuits operating under real conditions. We have done a poor job training them to understand its limitations. In the mind of a digital designer, fresh from school, Spice doesn't do E&M fields, ergo, they must not exist (or they don't matter).
I don't mean to knock simulation. It certainly has its place. In general, simulation can work wonders if you know what it is you are modeling. If, on the other hand, you are working with something like EMC the benefits of simulation are oversold. For EMC, where the whole problem is that we rarely know which effect matters the most, simulation doesn't work. In the immortal words of Samuel Clemens, "The calamity that comes is never the one we had prepared ourselves for."
5. Digital Engineers don't believe an understanding of EMC will advance their careers
This is a management problem. It's easy to see how it comes about.
Imagine that Joe is a top-notch product designer and digital guru. He has just demonstrated his mastery of EMC by getting his latest product to pass FCC and EC regulations on the first scan. He's a genius!
What happens next is highly predictable. Joe's design career is over. He will never design another processor at that company. He will instead be asked to debug Fred's EMC problems, and then Bob's, and then every other piece of junk that comes down the pike. He's effectively banished to the test range, to repeating over and over his EMC experience, while others reap the rewards of having gotten their sloppy processor boards to "function".
In today's business world, the typical digital engineer is rewarded for mere digital functionality rather than total readiness for manufacturing.
I'd like to think that we can turn this situation around. I'd like to think we can count on our EMC professionals, our signal integrity experts, and all the smart researchers at our universities to help undo these five great misconceptions, and to help us make some real inroads into the EMC difficulties that will face us in the coming decades. I'd like to think that if we did, it would make a big difference for the future of the computer industry.
At the very least, I hope we can get some more digital folks to show up at local EMC Society meetings. It's worth the trip.