Fun with FETs

May 27, 2018
 by Paul McGowan

I remember the first time I heard the sound of a FET (Field Effect Transistor) in comparison to a BJT (Bipolar Junction Transistor). It was quite a revelation. It happened not with individual devices but with an integrated circuit op-amp.

Most op-amps “back in the day” were BJT based and that’s what we were used to listening to. Our first op-amp was the iconic 709C designed by wild man engineer Bob Widlar of the Widlar Salute fame pictured here.

The 709C was a great sounding op-amp if you knew how to manage it (probably as great as Bob Widlar if you knew how to manage him). The next best sounding op-amp didn’t come along until the classic 5532.

Within a few years of the heyday of op-amps a new type appeared. These were touted as steps above traditional BJT based designs because they had a FET at their front end. FETs were better for many applications because they are essentially voltage amplifiers—like vacuum tubes are—and BJTs are current amplifiers. The first one I ever tried in our phono stage was a classic, one we still use today in servo applications. It was the TL082, a design that has outlived most of its competitors.

Fortunately for us, the TL082 is a drop-in part for just about any circuit. If you are using an 8-pin socket you could quickly and easily slip in a TL082, a NE5532, a 301, or a 741 without changing a thing and simply listen to the differences they present (the 709C needed some tender love and care to optimize it. Steps not needed or useful on these others).

Knowing the TL082 had a FET input meant there was a chance it might mirror some of the sonic qualities of a tube. And when we tried it, we were right. Compared to the 709, or 5532, the FET op-amp was smooth as silk. Warm, easy on the top end, not as etched as the others. Wow. At first, we thought it might be a miracle chip, something we could just start using in our designs and raise the level of performance by multiple notches.

But time is cruel to less than great performers. After living with the part for a few days we began to miss the details of the music, the upper harmonics seemed rolled off, the life had been sucked dry and in its place was a warm substitute. But warmth cannot replace details lost.

Which lead us to an entirely new path I’ll tell you about tomorrow.

Subscribe to Paul's Posts

10 comments on “Fun with FETs”

  1. I always find this type of post interesting, because they allow me to discern, or assimilate if it is the case, what others know more than me.

    Some audiophiles hate OP-amps, because more than one of the purist writers of the “specialized” audio magazines, it occurred to them to say that these devices should not be used in Hi-End audio, this, without explaining why .

    I have had the opportunity to replace the old LF-353 with the not so old: TL-072, with flattering results.

  2. Isn’t an op-amp an amp based on integrated chips instead of a discrete circuit built upon separate semiconductors and transistors etc?

    Haven’t such op-amps always been the clearly inferior or budget choice and anyone locating himself somewhere in high end avoids them? Why then is “op-amp” such a matter?

    1. Op-amp describes a functional block of circuitry that can be discrete or an integrated circuit. Not to be too much of a spoiler, but tomorrow’s post describes how we went from chips to discrete and why.

  3. One very important part of that story that is that it is easy to be initially seduced by one aspect of an equipment’s sonic character. Early on, as an audiophile “pup”, it sometimes took months of listening to realize the shortcomings that most always come with it. It was a sometimes financially painful experience. The balance between detail and musicality is a fickle thing.

  4. I strongly doubt that the development of op-amps was driven by audiophile of even highfidelity requirements. Probably miniaturization and reduction of manufacturing costs were the driving factors. Thus why blame op-amps? Hardcore audiophiles especially in Japan remained most loyal to vacuum tube designs. Portable transistor radios and cassette tape decks and low-fi components were the primary target mass market segments for op-amps. I guess only a minority of engineers designing stereo gear were obliged to follow audiophile aspects seeing the low Hifi standards and poor speaker designs of those days. And sadly enough the promises of digital audio for establishing a much higher level of sound quality were not fulfilled at all initially. Can we expect a real quantum leap in sound quality by newer technologies? Just look at this finding: https://www.monoandstereo.com/2018/05/as-crow-flies-and-surprise.html#more Neither quadrophonic sound nor home theatre concepts could kept the marketing promises. Thus I am keen looking forward to hear the quantum leap promised by the IRS killer.

  5. One technician/ designer told me in reference to op-amp chips (when used in the audio circuit) put it to me succinctly… “noise.” ‘Discrete’.. he told me was the way to go when possible. Me not having the knowledge to know what to think, could only notice when he removed the op-amps in an output stage of a DAC he modified for me… and had replaced them with Lundahl transformers… found nothing to argue about. Prior to that mod, what was op-amp noise, I had thought was simply an artifact of “digital noise” of the DAC. It was gone… and, now musical. That was when I became aware of the op-amp vs discrete debate..(and transformers.) Though, I must say that I found the high-end Muses 01 op-amps (using FET’s) were a good compromise in one preamp when substituting discrete was not possible.

  6. Great story, Paul. I remember building a Hafler preamp kit. It used a simple circuit and rolling opamps from the neighborhood Radio Shack was great fun!

  7. There was a coven of audiophiles in National Semiconductor including silicon designers and an FAE (Field Applications Engineer). The FAE caused trouble by clandestinely searching wafer production in Tucson for months to find a defect free wafer, and then processed it into matched pair RETs (ring emitter transistor). This is essentially an integrated circuit paralleling dozens of small transistors for higher speed, higher current, lower noise tradeoff than can be obtained by a monolithic transistor. Slicing the wafer into pairs meant perfectly matched parameters over process variables and temperature. This was the ideal opamp front end.

    When he took this idea physically realized to management, they shut it down because audiophiles are too small a market, and the cost of scanning wafers was too high. They cut a deal where the internal audiophiles got to keep all the precious RETs as long as they didn’t let knowledge of their existence leak out. I believe this group is responsible for the “Overture” series, which include the best sounding chip amps.

Leave a Reply

© 2023 PS Audio, Inc.

linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram