John Ulrick's Story
We were supposed to be hard at work designing weapons systems for America - that's Arnie Nudell and me. Arnie and I worked at Litton Guidance & Control Systems in Woodland Hills, near L.A. We did build a weapons system, a laser rangefinder that measures the distance from an airplane to the ground for accurate bombing. We were also audiophiles, if there was such a term at the time. The year was 1967. I am an electronic engineer, Arnie a physicist. At Litton, my job was to design the electronics for the laser rangefinder and Arnie's was to design the world's first Q-switched pulsed ruby laser. This was the laser that supplied the pulse of light that went from the airplane to the "target" and bounced back, used to calculate the distance from the airplane to the target. Part of that system involved a mirror to point the laser beam at the target. I designed a servo-control system, which points the mirror and its switching amplifier. During lab testing of the servo-control system, Arnie noticed that the servo was controlling the mirror at 20Hz, a magic number to any audio person. Arnie had a good idea - why don't we make a servo-controlled woofer? We went across the hall to the Litton engineering library and found a book called Acoustical Engineering by Harry F. Olsen - the first bible of acoustics. The book had a page on Servo Controlled Speakers but no analysis, just a block diagram. We worked together to design the system. Arnie calculated the radiation resistance for the slopes and I designed the servo system and the power amplifier to drive the woofer. Arnie called his friend, Gene Czerwinski, to make a special 18-inch woofer with a second winding to generate the feedback signal. At the time Gene was an engineer at Bendix and making speakers in his garage. Gene later left Bendix to found Cerwin Vega. At this point one could wonder, who is making all these weapons systems? The woofer worked. There were two more aerospace engineers working at nearby Atomics International, Ron and Bob, making electrostatic panels. (The pentagon always said you must value the technology spin off from military research!) What a great combination, electrostatic panels that don't have much bass, coupled with an 18-inch servo-controlled woofer. The system was built and set up in Arnie's living room. Using a Sherwood amplifier to drive the electrostatic panels, the tweaking began. The amount of servo feedback changed. The crossover points moved and moved again.
Arnie and I decided to start a company. You could tell by the name there was no professional help in naming the company. The name was "Nutek" for Nudell Ulrick technology. We began selling a system called the "Servo-Static I" through a local dealer. The local dealer was about a mile from Litton and owned by Dr. Jim Henderson a physicist who you will hear more about later. All of this was still part time with a new guy involved, Cary Christie, a very talented industrial designer making the cabinetry in his garage. Arnie and I put out feelers to raise money and were finally backed by a rep firm that sold me electronic parts at Litton. No sophistication here either; we didn't even have a business plan because we didn't know what one was. A corporation was formed with Arnie and I holding equal interest. Fortunately, when we went to incorporate, the name Nutek was taken. The name Infinity was the third choice on the list, but available. The company name was Infinity, the year 1968. Again Arnie and I demonstrated our management prowess in choosing who was going to be the first president; we used the coin toss method. I was the president the first year and then we alternated for the first several years. What confusion! In August 1968 myself, Arnie and Cary and a couple of other people started full time in a 2,000 sq. ft. building in Chatsworth. Infinity was started. At the start, Infinity had three speakers; the Servo-Static I ($1800), the 2000, a bookshelf with electrostatic tweeters ($276) and the 1000, a 2 way bookshelf speaker.
As time allowed, I started working on a digital switching amplifier based on my switching amplifier experience at Litton. Later my friend Dr. Jim Henderson, now a physicist at Hughes, joined in on the work. With Jim analyzing and me circuit designing, work progressed. Designing the power switching section with the parts available in the mid 70s was a real trick. I used power transistors from TRW, which were very expensive. On a good day I could go through $200 worth of them. About a year before the winter CES show in Chicago we naively committed to showing a digital switching amplifier. As the show grew closer, the Infinity gang heard Bose was going to show a switching amplifier. Not wanting to end up second, we pressed on to get a working amplifier ready for the show. For the last three months of October, November and December several people worked straight through, no days off. We spent Christmas morning with our families and met at Infinity in the early afternoon on Christmas day and had the Rose Bowl game on in the lab on New Year's Day.
The first day of the show was about Jan 7th. I had airline tickets to fly to Chicago a couple of days before the show opened, but the amplifier wasn't working. Finally, the night before the show, the amplifier was working, but the displays were making a ticking sound. Finally we just disconnected them, listened to the amp for about 15 minutes in the sound room and I caught the last flight that would make the show. I was picked up at the airport at 7 a.m. and rushed to the show. With my coat draped over it, I carried the amplifier past the line of press people waiting outside our room to see an amplifier they thought was in that room. We set it up and it worked! The doors opened at 9 a.m. and we showed the first digital switching amplifier ever. It was 125 watts per channel and ran flawlessly for the four days of the show. Bose didn't show up with a switching amp-Thanks!
Following the show, the amplifier was produced and sold as the SWAMP I, a 250 watt per channel amplifier. The name SWAMP came from one of Jim Henderson's computer programs with the file name SWAMP. Infinity sold a few hundred of them. Arnie and I sold our interest in Infinity and I left in 1979. Arnie stayed on as President till about 1990 when he left to start Genesis. Well, Infinity's part of Harman now, they sell through those big chain stores and the new gang even wears ties to work.
More fun. I had the opportunity to head up a new hi tech company, set up to develop a video multiplexer that allowed the transmission of two TV programs over a single satellite channel with a bunch of guys I loved working with including Jack with 125+ patents, and Dr. Jim. Big mistake! By the time we completed the research, got all kinds of patents and licensed the thing to a big broadcast equipment company, MPEG had become the standard. After some time running a computer company and an avionics company to make sure that wasn't what I wanted to do, I started Spectron and began work on a new generation of digital/switching amplifiers. I have always believed in them. It took three years of very long hours. My 600 hp fuel injected drag boat's deck has been used as a table and my Lotus formula Atlantic race car is just waiting to go racing. On the design side, one thing that sure has been a big help this time is all the computer aided math, PCB layout, drafting and simulation; MathCAD, AutoCAD, all that stuff.
It seems to have worked out. Enough people really like the sound, and unlike the Infinity amp, they are very reliable. (The addition of four diodes fixes the Infinity amp - wish I knew that then.) And, they are small and light. Even back at Infinity I envisioned a one-kilowatt amplifier. Conceptually, digital amplifiers are simple. You just keep sending pulses of energy alternately from the plus rail and then the minus rail into the speaker and have a feedback circuit to control all these bundles of energy coming from the plus and minus rails. Controlling the voltage at the speaker is like guiding a hockey puck on the ice. First you push from the left and then from the right and you are always making corrections as you guide the puck along the ice. The digital amp gets interesting because you do this 500,000 times a second so the digital circuits to control corrections must be very fast.
Conceptually, the digital amps are the way to go. They use virtually no analog circuitry (except the input buffer). This eliminates nearly all the analog gain stages, which are inherently slow, each adding a small amount of distortion and delay. It's the slowness of analog circuitry that diminishes the benefits of loop feedback, resulting in an undesirable effect on the sound. In Spectron's digital amp, the audio is converted into a digital signal at the input. From that point on it's processed in high-speed digital logic (TTL). The output section involves two FET switches, an inductor and a couple of capacitors, which are not in the signal path. If the switches were perfect, with zero turn ON and OFF times, no ON resistance and the logic had no delays, the amp would be perfect. Of course, none of the parts are perfect, but they are getting closer every year.
Well, if all goes well, next time you hear from me, about all I'll have to say is the amps are working great, they just keep going out the door and yeah, I've got time to do lunch!