Jensen Transformers was founded in 1974 by Deane Jensen as part of his quest for excellent studio sound. He believed the transformers of the era were of a substandard quality, and our field has benefitted from his research and dedication ever since. When Deane tragically passed away in 1989 he left his business to Bill Whitlock, who has maintained the level of quality ever since. This year, after decades of using their products, Peter Janis — the owner of Radial Engineering — purchased Jensen Transformers (see our sidebar interview). It was great to chat with Bill about Jensen and how transformers actually work.
Missiles and Op Amps
I basically moved to California in 1970 after growing up in Florida. I had graduated from school back in Florida, and I had some jobs. One of them, after I worked there for a year and got my top secret clearance, I found out that the reason they were so obsessed with the quality was that we were making the trigger mechanisms for H-bombs, so I got out of that for moral reasons. I did learn a lot about electronic components and how they fail. At the laboratory I worked in, we torture-tested resistors and capacitors and parts with outrageous extremes. Then we analyzed how and why they failed. That taught me a lot. Then I worked almost a year for a company, a division of Schlumberger, the French oil exploration company. I worked on airborne telemetry systems, which involved a lot of very high-speed discrete op amps, voltage controlled oscillators and things. I was fortunate enough to work under a brilliant analog designer who taught me a lot. I built a lot of his breadboards and we analyzed designs. I felt like I was very privileged to learn a lot about the secrets of discrete op amps. Then I saw an ad in the newspaper that RCA was hiring technicians for jobs in missile tracking stations. This was in 1968, when the draft was on, and I was not too excited about the prospect of going to Vietnam. I figured this would get me a deferment at the very least. They gave me a test that I essentially aced, and they said I could just choose where I'd like to be. They offered me positions at tracking stations on little islands in the South Atlantic, South Pacific, or work aboard one of their two missile tracking ships. The ship job sounded particularly fascinating, so I took a job on board a missile tracking ship. It was this 520-foot ship that was a converted World War II troop carrier. They had instrumented this with huge radar dishes. There was no mistaking what this boat was about. It was painted white, had an 80 foot L-band radar dish and a 60 foot C-band radar dish, a telemetry dish, and it worked largely in the Pacific, largely tied up at Pearl Harbor. Every time they were test firing missiles out of the air force base, they would send us out a few miles from the intended test splash position, and then we would acquire tracking on these missiles when they'd break the horizon and track them all the way to splash. Then we'd send secret reports back to the Air Force on how the thing deployed, the chaff clouds, the multiple re-entry vehicles, and all that good stuff. I worked in the navigation division and learned more than I ever wanted to know, actually. It was totally state of the art. We had to do drills on getting them removed from the equipment and thrown overboard in less than five minutes in case we were ever to be boarded by someone. This was back in the days when the U-2 pilot [Francis Gary Powers] had been captured. They warned us that we'd potentially be taken prisoner. We did actually go on spy missions once a year. We ended up off the coast of Kamchatka spying on the Russians; collecting data on their missiles. I was not really in the military, but everybody thought we were. We had status on military bases. They treated us like officers and whatnot. It was fun. I learned a lot there too.
On Deane Jensen's passing
It was his custom every Saturday morning to go out for a walk. We'd be working at home, and around 10 or 11 o'clock, he'd go out for a walk through the neighborhood. He wanted to do it alone, because he said that it helped him to get his thoughts together. He went out for a walk one Saturday morning, and he didn't come back. That afternoon I started calling around, wondering where he'd gone. I called a couple of friends I knew and asked if anybody'd seen Deane. I never was able to locate him. Monday morning, I called over to the Jensen offices and asked if he'd showed up for work. They never called me back, but Eric Denton came to my door and said that they'd found Deane dead in his office. He'd drunk a bottle of wine, written a suicide note, and shot himself in the head. His suicide note kind of explained part of it. Apparently he had managed to get himself really in debt with Reichenbach Engineering who was building the transformers for him at that point. He did all the designing. He was very fussy about how Faraday shields were built and terminated. He was asking for a lot of stuff that was not routine. Apparently, after Ed Reichenbach, the founder, had died in 1986, the lady that took over the business didn't get along with Deane. She was always giving him a lot of grief over having things done a certain way when it would cause extra work. There was constant friction. Apparently Deane had allowed this debt to pile up, money that he owed them. Again, this was corroborated by the rest of the staff at Jensen. She started making demands, where if he didn't pay something right now, she'd cut him off. He previously fired an employee, Jimmy Bauer, who had physically attacked Deane. He just fired him, and apparently Jimmy went out and filed a million dollar lawsuit against Deane for intentional infliction of emotional distress, or some such. Between those two, that was what he wrote in his suicide note. My supplier is blackmailing me and my best ex-Christian friend has betrayed me completely as a friend. I just can't take it anymore.
Spatializer Audio Labs
They actually went public, and I still own a bunch of stock that's now worthless, but that's another story. I think it's actually still listed on the NASDAQ exchange under SPAZ, They got some hot-shot manager who was over-ambitious. Rather than stick to the business they knew, they tried to diversify. They bought this company that was a real loser. It dragged them right now. They're essentially gone. The shell of the company got bought by DTS about ten years ago who seem to be sitting on it. Anyways, when I was given the stock as bonus stock, it was worth $5 a share, but now it's trading on NASDAQ for one or two cents a share. I kind of want to just lock the door on it. That was a learning experience.
Common Mode Rejection
This is an argument I had with the IEC back in 2000. The IEC put out a call for comment back in 1999, recognizing that their test for common-mode rejection bore no correlation to real-world results of noise rejection. I looked at the old test, and I said, "Duh. I can tell you why!" They were testing the receivers with perfect laboratory sources. In the real world, signal sources are not perfectly balanced. I don't want to go into a whole lecture here, but you can show that sources of balanced signals, if you look at the impedances of those two output pins with respect to ground, they're very commonly mismatched by 10 ohms, sometimes a lot more. For a lot of engineers, the aha moment comes when you redraw a balanced interface as a Wheatstone bridge and show that the source impedances of the signal output and the common-mode input impedances of the balanced input form a bridge. If the ratios of those impedances are perfectly matched, none of the ground voltage difference between the two boxes gets converted to a differential signal on the line. If you think about it, there's only two ways to stop that conversion from common-mode noise from getting into the signal. One would be to make the source impedances zero, which is impractical. It just can't be done. Op amps won't stay stable driving line. The other way is if I can somehow make the receivers' common-mode impedances infinite. Then it would be impossible to convert any of the noise to signal. That's the key of why a transformer does what it does. As common-mode input impedances approach infinity, they're nothing but very small capacitances.
"Transformers Make Audio Warm"
That's an end of the business that we decided years ago to stay away from. What people have become attached to is actually the sounds of imperfections in those old transformer designs. Either the core material was pretty awful... a lot of times it was steel, and not even very well heat-treated. The distortion characteristics of a transformer come primarily from the core material. That's why we always choose very high nickel content material, which is intrinsically orders of magnitude better than steel. The only trade secret we keep at Jensen is that we have something of our own secret sauce when it comes to heat-treating this material to eliminate the hysteresis distortion, which is the distortion that happens at low signal levels. People have become very attached to those old harmonic distortions in what we would consider these awful transformers, like some of the old UTCs. Triads were good, but of course the Jensens are better. We shy away from that. One way to think about it is that if your goal is transparency, it's easy to measure how close you're getting to perfection. But if your goal is coloration, what blend do you choose? How do I know when I'm there? There's no standardized measure of what level of secondary harmonic distortion will achieve this "warmth." We've decided to just categorically stay away from that side of the business. Right now I'm working on what I think is an entirely new approach to emulating transformer distortion. If you really understand the kind of mysterious transformations that go on inside the transformer... it's a passive device, but there's a lot going on in there in terms of equalization and non-linearities. I think I can simulate 99% of that with some active circuitry. That's something I'm going to be working on in the next six months of so. I'll run it up the flagpole and see who salutes. It's kind of an old transformer emulator with lots of little trims on it to make it sound different ways.
Physics in the Walls
Physics. You have to know that it's everywhere. In fact, a paper that Jamie Fox, who supplied all the electrical wiring for the Bing Concert Hall up at Stanford University. It kind of became a testbed for a paper that we wrote in 2011. I'll just summarize it in a couple sentences, but an elephant in the room since 50 years ago, and a question that always bothered me that I never got a good answer for, was where does the voltage come from that drives ground loops? Nobody ever had a good answer. Then one day I started thinking about it. Good old Faraday's Law that says if there's a fluctuating magnetic field somewhere and a conductor in that field, there's going to be an induced voltage. It turns out that the voltage that drives ground loops comes from right inside the walls... the wires in the conduit, the AC wire. It's kind of something that nobody's ever looked at. It was kind of a theory, and I did a little tabletop experiment to convince myself that it was real. In 2011 we decided to do a large-scale experiment. Jamie Fox had been a student when I did a lecture up at Cal Poly, and he went on to become an engineer up in the Bay Area. Since he was fresh out of college, I asked him to handle the theoretical side of the paper. I let him work all the Maxwell's equations and predict what I should be seeing in my experiments, and then I did the experiments. Anyway, we were very pleased when theory and measurement agreed within about five percent. We published this paper that showed that what you did with the premises wiring in the wall can make a huge difference with your noise problems. In fact, something as simple as... the way electricians normally install wiring in conduit is to get the right colors of wires and pull them all in en masse. It turns out that if you take the pairs of white and black wire for each circuit and twist them before you pull them into the conduit, you can reduce the system noise by a factor of up to a thousand. Just with one simple little trick.