Nelson Pass is a brilliant designer of audio amplifiers and has become a sort of legend in the audio community.
He holds several patents related to audio circuits, founded companies such as Threshhold, Pass Labs and First Watt, and is active in the DIY audio community.
Recently Paul Norton of Linear Systems had the opportunity to talk with Pass about his background, JFETs vs. Transistors, N-Channel vs. P-Channel, matching and more.
Norton: How did you get into audio?
Pass: I grew up in the 60’s listening to jazz and rock all night on SF bay area AM radio. I also liked to build stuff – model planes, slot cars, Jacob’s ladders – and loudspeakers. I’ve been doing amplifiers for 50+ years, starting with kits, then repair and modifying commercial products. At UC Davis I met other people seriously working on amplifiers, and this led to a career starting with ESS the loudspeaker company in 1972. There I spent two years helping development of Heil transformer products, and then I left to form Threshold Corporation with Rene Besne, also from ESS.
At Threshold I developed the 800A amplifier, which was the first dynamically biased Class A amplifier, followed by the 400a and 4000 and other products, which were widely copied. In 1978 we released the Stasis series of amplifiers, still manufactured. In 1991 I left to form Pass Labs and created the Aleph series of single-ended Class A amplifiers, followed in 1995 by the “super-symmetry” design which is still the basis for our products after 25 years. In 1998, I founded a second company, First Watt, which makes small Class A amplifiers having unique and interesting designs, made in limited quantities. Along the way I also did designs for Adcom (apparently the GFA555 remains the highest selling amplifier ever made), Mobile Fidelity’s cutter-head system, Nakamchi, Soundstream and a several others.
I have always had an interest in the “DIY” part of audio, and since 1972 I have published about 90 articles for the magazine Audio Amateur, Audio Magazine and others (Most of these can be seen at the www.firstwatt.com website) as well as participating daily online at www.diyaudio.com.
This same history is also an evolution in design philosophy. My first ten years focused on making amplifiers having low distortion without so much feedback, and so I used Class A circuits. The next ten years was working with means of lowering distortion with only local feedback, and then five years exploring single-ended Class A FET amplifiers.
After that 25 years developing and refining the super-symmetric circuits that define the amplifiers made by Pass Labs, and also exploring the subtleties of harmonic amplitude and phase related to the perception of the sound of an amplifier. At this point I mostly work with FETs, with a special preference for JFETs, which includes Static Induction Transistors (SITs). Throughout, I have emphasized the sound, reliability, and simplicity.
Norton: How did you become associated with Burning Amp Festival?
Pass: I knew several audio DIYers in the Bay area through the website www.diyaudio.com and in 2007 was contacted to participate in the first BAF. I brought a large truckload of audio parts to the Fort Mason firehouse, was given a Santa cap, and distributed the stuff to attending DIYers. That was the start, and I have participated since.
Norton: Thoughts on JFETs vs Transistors
Pass: Of course, JFETs are Transistors, but I understand you mean non-JFETs. I have always looked for the most performance with the greatest simplicity, and JFETs deliver this. JFETs are the most favored by that same criteria, although it is very difficult to obtain large JFETs power Transistors. In the arena of less power JFETs have very low noise, very high input impedance, simplicity in biasing, and a square-law gain characteristic that mimics that of tubes, making them very desirable for audio applications. There have been high power JFET Transistors, “Static Induction Transistors” (SIT) invented in Japan and made famous for audio use as “VFETs” in the ‘70’s by Sony and Yamaha, and manufactured until recently by Tokin. They mimic Triode tube characteristics in solid state, and get up to as large as 3 Kilowatt dissipation, 1200 volts and 200 amps. Serious parts…There are now high power JFETs made from the newer Silicon Carbide (SiC) and Gallium Nitride (GaN) material, and there are even some SIT versions existing or in development.
Norton: N-Channel vs P-channel. Advantages?
Pass: Not particularly. Most of the time I use N channel because they are cheaper and more available.
Norton: Singles vs dual, advantages?
Pass: Duals are matched parts, and that is useful, although I test all JFETs, matching for Vp or Idss anyway.
Norton: Thoughts on Quad JFETs?
Pass: Paralleled, they give more gain, more current, more dissipation, and less noise, all very welcome.
Norton: JFET matching tips?
Pass: Consistent conditions are important, including temperature, voltages, currents and test duration. All these affect the result. I first go through a population testing for a rough value, for example sorting them into bins by 1 mA values for Idss. Then I go back again to each bin and further sort them in 0.1 mA increments. Two tests adds to the labor, but it results in much less human error – it is not easy to test thousands of transistors, accurately dropping each into 1 of 100 bins…
If the need for accuracy is not high, and the Idss is high, say >20 mA, then I test by pinch-off voltage, Vp with the voltage at 10 volts or so and using a 1 megohm Source resistor. For a given JFET type, this tracks Idss very well and you don’t get the effects of chip heating during the test.
Norton: How did you meet Siegfried Linkwitz?
Pass: I met Siegfried the first time at BAF. There were many coincidences between us – he moved to Santa Rosa about the time I left for UC Davis, lived close to where I had lived, and his kids went to the same schools. His first high-end loudspeaker purchases were early ESS models (before Heil) which were just being released when I joined the company. Later, I moved to Sea Ranch, where Siegfried had a house, and we ended up on opposite ends of the same beach. When he was there we would get together to listen to stuff and sit on his back deck in the sun. We got along very well, and I ended up getting involved in development for the analog version of the LX-mini and LX-mini +2 /S versions.
Norton: Current/future projects?
Pass: I have a long list. Everything that is not a loudspeaker involves some kind of JFET. All I need are more quantities and lower prices… For example, I am resurrecting the Beast with a Thousand JFETs (see the article on www.firstwatt.com) which uses thousands of these parts. I need lots of LSK170B and LSJ74B, and they don’t have to be particularly low noise, as they are used in the output stage of a power amplifier…
Norton: How and why you became associated with Linear Systems?
Pass: I met the boyz from Linear Systems at the Burning Amp Festival years ago, and have seen them in supportive attendance ever since. Apart from that, they are the only replacements for the long discontinued complementary JFETs that were made by Toshiba. These particular JFETs, the 2SK170 and 2SK74 and the LS versions LSK170 and LSJ74 are crucial to the products from my three companies that manufacture audio equipment and the DIY audio designs I publish.