Anybody else prefer discrete semiconductors over opamps?

JakeAC5253 · Post by **JakeAC5253** » 19 Dec 2011, 06:01

I cannot explain exactly why, but I think I prefer all variations of transistor/FET designs over opamp designs for guitar related dirtboxes. By now I have modded three different Big Muffs to achieve three different sounds and they all sound great, the modding was fairly effortless once you learn certain things and I was able to achieve my end tones. I've also been working on an opamp design or two for a lot longer than I would like to admit working on a single design to reproduce good sound, and it seems that the pursuit of good sound/perfect design tone chase needs to take a backseat to proper opamp design mechanics and minutia moreso than in discrete designs. I find opamp designs to be more limiting, and their end tones don't sound as juicy to me. It could be that I still have more to learn and whatnot, but I have met at least one other person who felt this way about DIY designs. Opamps do make excellent high impedance buffers though... What do you think?

mmolteratx · Post by **mmolteratx** » 19 Dec 2011, 07:36

I love op amps for buffering, active filtering and linear gain. I like distortion to come from JFETs or BJTs though. Most of my dirt designs mix the two, my boost design just uses op amps. Though I'll use op amps and clipping diodes when I'm going for that sound. I don't really have a use for anything but fuzz and tube sounds though.

mictester · Post by **mictester** » 19 Dec 2011, 12:38

It depends what you're trying to achieve. Op-amps are really good when you just need to change impedances, filter something, or derive a DC voltage from an envelope. The diodes in the feedback loop (the tubescreamer and the zillions of variants) have a place and actually if well done is quite musical. My "Really Smooth Overdrive" is an example of an effective use of op-amps. The sounds available from that circuit would be more difficult to achieve with transistors (though a Big Muff could come close).

Simple treble boosters and amplification stages tend to be easier to build with discrete transistors, but the op-amp will give you a greater range of gain (if that's what you need). Transistors and FETs tend to distort in a more benign fashion than the complex op-amp circuits (which are usually Class B output). Clipping caused by hitting the supply rails with op-amps tends to sound horrible.

I can't way I have any preference for one over the other - it just depends on what you're trying to achieve...

DrNomis · Post by **DrNomis** » 19 Dec 2011, 13:54

J-Fets can produce similar distortion harmonics to the harmonics generated by Valves/Tubes, you can think of a J-Fet as being the solid-state equivalent to Valves/Tubes, both types of devices are Voltage-Controlled, whereas a Bi-Polar Transistor is Current-Controlled, a single J-Fet stage on it's own doesn't really produce alot of gain, and what gain it does produce tends to vary a bit depending on how much current flows from the Source to the Drain (I'm using electron-flow here, negative to positive), as the Source-to-Drain current goes down the gain goes up, so a Fet tends to exhibit similar non-linearities to Valves/Tubes (predominantly Even-Order Harmonics due to asymmetrical waveform clipping).

Because the internal circuitry of Op-Amps tends to be symmetrical in design, the non-linearities are different, add to that the usual amounts of negative-feedback and you'll find that although the distortion figures are lower, the distortion harmonics are different (predominantly Odd Order Harmonics due to symmetrical waveform clipping).

I tend not to have any preference for Discrete Semiconductors or Op-Amps, I'm happy using either.....

drbob1 · Post by **drbob1** » 19 Dec 2011, 15:37

When I saw the title, I thought you were building your own op-amps from discrete components. Now that would be a manly variation on a tubescreamer! I've seen discrete component op-amps in mixing boards that weren't too big to fit in a pedal...

juanro · Post by **juanro** » 19 Dec 2011, 15:42

Have you seen this: https://www.diystompboxes.com/analogalc ... opamp.html ?

Sort of a discrete op-amp. While limited in it's posibilities, I have used it in the past and it worked and sounded fine.

Juanro

JakeAC5253 · Post by **JakeAC5253** » 20 Dec 2011, 05:33

DrNomis wrote:J-Fets can produce similar distortion harmonics to the harmonics generated by Valves/Tubes, you can think of a J-Fet as being the solid-state equivalent to Valves/Tubes, both types of devices are Voltage-Controlled, whereas a Bi-Polar Transistor is Current-Controlled, a single J-Fet stage on it's own doesn't really produce alot of gain, and what gain it does produce tends to vary a bit depending on how much current flows from the Source to the Drain (I'm using electron-flow here, negative to positive), as the Source-to-Drain current goes down the gain goes up, so a Fet tends to exhibit similar non-linearities to Valves/Tubes (predominantly Even-Order Harmonics due to asymmetrical waveform clipping).

Because the internal circuitry of Op-Amps tends to be symmetrical in design, the non-linearities are different, add to that the usual amounts of negative-feedback and you'll find that although the distortion figures are lower, the distortion harmonics are different (predominantly Odd Order Harmonics due to symmetrical waveform clipping).

I tend not to have any preference for Discrete Semiconductors or Op-Amps, I'm happy using either.....

Cool summary. The JFET's being voltage biased as opposed to current biased may be the reason I have not yet fully grasped them yet.

juanro wrote:Have you seen this: https://www.diystompboxes.com/analogalc ... opamp.html ?

Sort of a discrete op-amp. While limited in it's posibilities, I have used it in the past and it worked and sounded fine.

Juanro

Now that is cool. Sort of like a modified long tailed pair with a single ended output amp? Seems like it would make for a neat little project. Thanks for sharing.

Duckman · Post by **Duckman** » 20 Dec 2011, 06:18

Check this old thread and look for member UZILSD posted links. You'll may find interesting things.

https://www.freestompboxes.org/viewtopic.php?f=52&t=6095

RnFR · Post by **RnFR** » 20 Dec 2011, 10:36

oh yes, for distortion/fuzz tones I'm all about the discrete designs. but of course I'm a freak for a germanium transistor. they just seem to have a much smoother onset of clipping, not just a lack of high end. I guess diode clippers just never did it for me. even the bmp I built lately probably gets a lot of the distortion from the transistors due to the LEDs. op amps of course can't be beat for filters and such. like many have said, it's all about what you use them for- as well as personal taste.

culturejam · Post by **culturejam** » 20 Dec 2011, 15:06

I tend to like both, actually. For me, it's more dependent on the overall design and final tone than it is being transistor vs op amp.

And after all, an op amp is really just a small collection of transistors (along with resistors, diodes, and caps, of course). Granted, you aren't going to find any germanium op amps (at least I don't think so).

RnFR · Post by **RnFR** » 20 Dec 2011, 15:39

yes, but a single transistor(class A) amplifier distorts much differently than an op amp. I recently saw an article at ESP about getting op amps into class A, and just using the NPN transistors used within. I have no idea If this would actually make opamps distort nicer, though. it might be worth checking out.

Seiche · Post by **Seiche** » 20 Dec 2011, 15:43

i remember people getting a little upset when the honey bee turned out to not be an all disrete design but with an opamp at it's core instead (because they somehow understood björn would only do discrete designs or something). didn't anyone hear that?

culturejam · Post by **culturejam** » 20 Dec 2011, 15:49

RnFR wrote:yes, but a single transistor(class A) amplifier distorts much differently than an op amp.

Sure. I just don't think I'd be able to pick out, consistently, an op amp dirt box from a transistor-only dirt box from sound samples. In fact, I've been quite wrong in the past when trying to guess the dirt source from sound alone. But maybe that's just my own ignorance?

RnFR · Post by **RnFR** » 20 Dec 2011, 16:12

maybe?

I don't know about others, but I don't think it's particularly hard to pick out one from the other. not that I haven't been fooled in the past. well, at least in higher gain designs. with really low gain stuff, I think it's a bit harder. I do listen to guitars at work all the time, though. not everybody has that luxury. I could describe them, but I'm afraid Id get into those vague, tired descriptive words that everybody loves so much.

culturejam · Post by **culturejam** » 20 Dec 2011, 16:24

Are you saying it's easier or harder to tell with high gain?

I'm not sure about low gain stuff, because I use my amp for that 99% of the time. I just know that I've heard a 4049 high gainer sound an awful lot like a JFET high-gainer.

And I once tested a JFET Vulcan next to an op amp Muff and myself and others present agreed that the tone was remarkably similar.

diagrammatiks · Post by **diagrammatiks** » 20 Dec 2011, 18:54

all semiconductors sound the same.

harumph.

RnFR · Post by **RnFR** » 21 Dec 2011, 00:46

culturejam wrote:Are you saying it's easier or harder to tell with high gain?

I'm not sure about low gain stuff, because I use my amp for that 99% of the time. I just know that I've heard a 4049 high gainer sound an awful lot like a JFET high-gainer.

And I once tested a JFET Vulcan next to an op amp Muff and myself and others present agreed that the tone was remarkably similar.

no, i'm saying it's easier to tell with high gain. as far as the 4049 goes, i wouldn't know. i haven't really messed with them much. i should!

bhill · Post by **bhill** » 21 Dec 2011, 01:08

Is this going to turn into one of those "Poly vs Nitro" holy wars?

RnFR · Post by **RnFR** » 21 Dec 2011, 01:25

oh boy i hope so. we can never get enough of those.

JakeAC5253 · Post by **JakeAC5253** » 21 Dec 2011, 01:35

bhill wrote:Is this going to turn into one of those "Poly vs Nitro" holy wars?

Preamp tube covers vs. no covers