6
\$\begingroup\$

I have often seen circuits for guitar pedals that use an emitter-follower configuration as the input and output buffer (see example below). If an op-amp can provide a greater input impedance / lower output impedance, why are they not used instead?

Is it simply the case that a single transistor circuit is cheaper to mass produce? Some people (audiophiles) say that transistors sound better, claiming that op-amps have a more "clinical" less warm sound. I am quite skeptical about this since there is no obvious difference in the frequency response at audible levels.

"Tube Screamer" input stage

If I'm not trying to mass produce a pedal and I'm looking to build something of quality, is there any reason not to use an op-amp as a buffer? In your experience, do op-amps have any intrinsically negative effect on the sound?

\$\endgroup\$
6
  • 6
    \$\begingroup\$ Well, "the simplest thing that works" is a good advice. Audiophiles say a lot of BS, you're right to be skeptical. Not many opamps can be powered with a single supply, so using them in a battery powered device somewhat limits your choices (or complicates things by introducing a "fake" dual supply, but then +-4.5V is not enough for some opamps)... So, my advice: stay with transistors \$\endgroup\$
    – Undertalk
    Commented Oct 18, 2018 at 13:09
  • 1
    \$\begingroup\$ Any objective measurable reasons are out of the window when you talk about audiophile stuff. Trannys are indeed cheaper, but sometimes you also don't need/want high impedance, since you rather would like a current audio signal instead of a voltage for noise reasons. \$\endgroup\$
    – PlasmaHH
    Commented Oct 18, 2018 at 13:09
  • 1
    \$\begingroup\$ An emitter follower is going to have odd order distortion, if you want to appease the audio****, a JFET will have 2nd order distortion. BJTs tend to be fairly electrically robust, compared to a FETs. \$\endgroup\$
    – sstobbe
    Commented Oct 18, 2018 at 13:32
  • 2
    \$\begingroup\$ Yes, when it comes to audiophile stuff, it's not so much the accuracy and precision that most are after, it's actually the distortion products that make the sound "unique". Some prefer Pepsi, others Coca-Cola. Some companies l think do things a certain way in order to preserve their "sound brand" more than anything. \$\endgroup\$
    – user201365
    Commented Oct 18, 2018 at 13:36
  • 1
    \$\begingroup\$ Guitar amps and effects boxes aren't about simple amplification. They are part of the instrument, and form the sound. If you want clean amplification, that's one thing. If you want to influence the sound, that's another thing entirely. As a guitar player, you are creating a particular sound. Imagine it as the difference between a painter (pictures) who intentionally sets the colors he paints to create a visual impression against the guy building a camera to capture an accurate picture of the nuances that the artist created. \$\endgroup\$
    – JRE
    Commented Oct 18, 2018 at 14:03

2 Answers 2

1
\$\begingroup\$

From a pure engineering point of view an opamp circuit can be made such that it will have less distortion than a simple emitter follower as shown in your schematic. I can build both (in real life and/or in a circuit simulator) and prove that with hard numbers.

That is not to say that the simple emitter follower is worse or better (for audio) than an opamp based solution. What I can measure/simulate are only certain aspects.

Some prefer Tube amplifiers which most of us know add distortions. So the most distortion free solution from an engineering point of view might not be the best sounding one.

Regarding your simple emitter follower: as commented, it is a well known, simple to use, robust, proven, low cost solution. There is no reason to use an opamp. The lower output impedance isn't always needed. And low enough is low enough. There is no need to go lower than "low enough" now is there?

I am quite skeptical about this

and you should be, the emitter follower solution does the job, no need to make things more complex than needed.

\$\endgroup\$
3
  • 1
    \$\begingroup\$ Quite a few of the cheap "jellybean" op-amps from the olden days (70's and 80's) have pretty severe crossover distortion. It doesn't affect how they work for non-audio, but would make soft passages sound crappy. There are a few parts that are good. Finding the right op-amp, and engaging in the running battle with a purchasing department that can't understand why you won't specify a "perfectly good" replacement, is a lot more work than just using an emitter follower (which, if you think about it, isn't using many more parts). \$\endgroup\$
    – TimWescott
    Commented Oct 18, 2018 at 15:17
  • \$\begingroup\$ @Bimpelrekkie in what sense is a transistor more robust than an op-amp, could you please clarify? The high input impedance of the op-amp is quite desirable with guitar pedals. The emitter-follower shown only delivers around 200-300k of input impedance (assuming a beta of around 100). \$\endgroup\$
    – Chris-Al
    Commented Oct 19, 2018 at 8:33
  • \$\begingroup\$ An opamp solution is more complex so it has more points of failure. A one-transistor solutions is simple, has less points of failure and is therefore more robust. only delivers around 200-300k of input impedance And why would that not be enough, why would you need a higher input impedance? Your first step is always to determine what (input impedance) you need, only then can you say I need more. \$\endgroup\$ Commented Oct 19, 2018 at 13:59
1
\$\begingroup\$

All the comments about "well, the emitter follower adds a subtle sound color" are rather moot, because most of these pedals go on to deliberately create MASSIVE distortion. Also, I note that in my Behringer UM300, the input emitter follower buffer remains in the circuit even when the effect is switched out, and there's a second emitter follower output buffer that is always in the signal path too.

Bottom line: I seriously doubt anyone could hear the difference between a precision op amp and a 1-transistor buffer in this application!

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.