I get to work with a lot of great stuff in the world of electricity and power electronics. From the very beginning, I’ve taken a hands-on path to these innovations and discoveries with an unrelenting curiosity. Along these lines, I’ve worked on a lot of equipment.
Early on, my work was in repair and modification. For the last couple of decades, it’s been on the engineering side. But while I’ve seen and taken part in a lot of innovation, one area of power electronics needs to catch up with the speeds offered by today’s state-of-the-arts devices—magnetics materials.
One of my first big scores happened the summer after eighth grade. My peers had odd jobs picking up construction debris, mowing lawns, and delivering newspapers. I found a gig mowing the lawn at “The Toolstore,” which was down the street on the main avenue running through town.
I had an old Cooper Cyclovac that I found in the trash, cleaned out the carburetor, and changed out the plug, and it ran like a top. Mowing at the Toolstore was risky, though. It was on the main drag, in a suburb adjacent to Chicago. There were lots of bottles, strings, strung out cassette tapes, twine, packing straps, and just plain trash. I could never find it all in my initial walk around the property.
I took a few chunks out of the blade, knocked the crankshaft a little out of balance, and had endless strands of debris wound around the shaft. But it paid well, and I spent most of it in the store buying odds and ends that I needed—the good stuff. No junk tools!
An Early Project
A few blades into the summer, I was mowing along when a very angry customer all but rolled me over. He had a large air compressor on a trailer—an 80-gallon tank, a big motor, and a three-cylinder radial pump. I couldn’t hear over the valve slap from the remaining 5:1 compression in my antique Brigs and Stratton engine, but the body language and events indicated that the air compressor was bought there—and it didn’t work.
By then, I knew a few things. For example, I could bend conduit pipe pretty well. I also knew the ampacities of most wire and basic codes. I finished my mowing and approached slowly. The customer who bought the compressor felt like he’d been had, and he had the pipes to deliver that notion at about 140 dB standing a half-inch away from the store owner. So I asked if I could have a look.
I climbed up on the trailer and noted the 120-V ac, 20-A plug. I asked what his service was. He said that he’d made an adaptor to get that plug into the dryer outlet in his shop. I asked if the dryer outlet was 220 V. He responded that he didn’t know. It was just there. He added that the breaker kept tripping, the reset button on the motor would always pop out, and the compressor was junk.
I asked if I could pull the side plate off of the motor and look at the connections. By now I had the store owner’s attention. I talked through it, wanting to fix it instead of getting into the dispute. “Yes, the motor is strapped for 120-V service. See this wiring diagram? If we just move these conductors like it says, we can set it up for 220-V operation,” I explained.
The customer inquired how we would know if it worked. I suggested that he take it back to his shop with the plug and adaptor, plug it in, and give it a try. “Don’t even unload it. Just see if it runs. If it does, bring the adaptor back here and we’ll put the 220-V plug on the cord,” I said.
He came back quite content 20 minutes later with a ragged adaptor made out of 16AWG speaker wire doubled up and seemingly stripped with a dull chainsaw. I carefully dressed the ends of the machine cord with strippers, twisted them up, and terminated them to the lugs on the plug. I tightened everything down, checked it all for strain or pull-out, and sent him on his way after suggesting that the adaptor be kept in a safe place like a dumpster.
The owner of the Toolstore pulled me aside and told me, “You don’t mow the lawn no more. I need you fixing stuff and assembling stuff in here.” This was the beginning of a great summer. I rebuilt countless tools—electric, pneumatic, portable, stationary. I assembled all the stuff in the showroom and organized it all. Great stuff.
This was one of many stops along a long path that led to power and energy engineering. Today, I take that same inquisitive, active, curious approach to most problems, whether it’s a self-commutated N phase brushless dc (BLDC) motor drive or a multiphase buck converter for a voltage regulator module (VRM).
Recently, I’ve become involved with the gallium-nitride (GaN) devices at International Rectifier (see “GaN And The Smart Grid Continue To Shape Power’s Future—Hopefully”). There have been numerous articles to the effect of “Where is it? We’re waiting.” Much like the results that came from the Haynes Shockley experiment, these things will take time to proliferate.
When you turn on your transistor radio or perhaps your 7.1 home theater system, you will note that there aren’t any of the “point contact” transistors in those circuits that Bardeen and Brattain were convinced would be so prolific. They were wrong. It took a while to iron out the details of the bipolar junction transistor as we know it today, and ultimately Schockley’s notion of diffused base and emitter regions prevailed. Looking back, it certainly was an important technological breakthrough.
Along similar lines, International Rectifier’s GaN on silicon (Si) will climb a similar curve. It has many advantages over similar technologies, but the primary advantage lies in the design for manufacturability. GaN on Si requires no special lines, wafers, boules, seed crystals, or processing. It uses a standard silicon wafer where the GaN material is grown on the surface. Details are protected and reserved for IR’s lead technologists and engineers, but GaN on Si is the diffused base and emitter of our time.
Out of this, IR has a new device with much faster switching times, much lower RDS(on) per unit die area, and much lower parasitic capacitances. This will only get better. But there’s a minor problem. If I can build an offline dc-dc converter that runs much faster, I’ll need magnetic components that can handle these higher switching frequencies with fairly low losses in the transformers, inductors, and reactors.
One of my daydreams is that somebody develops a nanoscale string of iron molecules, perhaps with a few fatter dielectric molecules hanging off of the sides to keep the adjacent strands from touching. We could then wind this nanoscale fiber through some turns of wire and have a nearly ideal magnetic core material with very high permeability in the ungapped state, a nearly ideal permeability tensor that stayed reactive many decades beyond the frequencies of interest, and absolutely minimal eddy current losses. I suspect BSAT would be quite high as well.
If I had this fun stuff, the first thing I’d do is build up a high-output micro-PFC (power factor correction) stage, perhaps in the shell of a Hubble 5-15 plug, perhaps with a micro-isolated dc-dc stage, only I can’t do that with present magnetic materials. The soft ferrites we have today are good, but not nanoscale good. This needed breakthrough will exploit the size reductions from high-speed switching with GaN devices.
The second thing I’d do with third- or fourth-generation GaN devices would be to build a matching transformer to run my 5/8 wave vertical amateur radio antenna, catch some decent sunspots, and work a few stations on 10 m with a class D amplifier!
I’d run the switching frequency from a phase-locked loop (PLL) at the carrier frequency, knock the harmonics off the output with a low-pass LC filter/integrator/matching network, and let the integrator round out the class D modulation and carrier to AM or A3E emissions on the carrier. The more integration, the more carrier suppression.
Is that practical? Certainly not. But wouldn’t it be neat to talk to the guy with the old-school class AB1 plate modulated tube amplifier and the 200-W audio amplifier modulating the plates of his finals with a 99% efficient class D RF output stage? Based on what I’ve seen, we could have the switches to transfer power at 28 MHz in a few years. We may also have the gate drivers to drive them by that point.
Lastly, if we could flash some of this nanoscale material, perhaps modified for better retentivity, we could make permanent magnets that were at or near the ideal. This would make the 1.5T N52 material look feeble, alleviating all that economic pressure from the scarcities of N52.
I’m not seeing the breakthroughs in the magnetics world, though. It’s like they stopped when they could pull the nickel out of powdered iron materials to offer “Xflux.” I’m sure I’ll get pounded on a rebuttal from somebody in magnetics, but walking around APEC the last few years, I don’t see anything on nanoscale magnetic materials. No samples, no literature, not even a sales pitch.
This game changer is needed. Just like the curious kid who once questioned a few things that were unseen with an air compressor, I now question the magnetics industry. I can’t pull the cover plate and rewire this metaphorically, but I’d gladly design structures with any preliminary nanoscale materials, instrument them, and try them out.