OP275 Opamp - Is It Really That Special?

[merlinb]

 However, I'm fairly confident that I could reliably pick up the different chips if comparing, say, a pair of OPA827s against a TL072 in a high fidelity playback system.  The OPA827 sounds subjectively cleaner and clearer, and this is supported by Groner's tests.

Yeaaahhh... I'm gonna have to go ahead an disagree with you there. You're saying you can 'hear' sub-0.01% distortion even amongst the 0.5 to 3% produced by any loudspeaker??
(I guess if you subbed a TL07x into a circuit that drives a few hundred ohms it would suffer, but that's hardly a fair test)

[Paul Marossy]

Minimally educated speculation.  As long as things stay linear, I'm doubt I can hear the difference with my damaged old ears, but when things become non-linear, i.e. distortion, there may be differences in the harmonic content generated, i.e. harmonics/overtones, that I can hear.  Since different op-amps have different spec.s and respond to resistance, capacitance, transconductance, bias, etc., i.e. Circuit Design differently, (isn't' that why there are so many different ones) there may be something going on that is audible.  The different circuit topography of the op-amps may distort differently, respond to overload differently, take longer to recover, etc.

I have noticed that a higher level of even ordered harmonics seems to add midrange for example.  I think I have a heard a swirl to the 4558 distortion that isn't present in some of the higher quality op-amps.  I'm probably not a good test subject though. 

I have long speculated that if people really do hear differences between various opamps, that it is due to harmonics. Now that you mention it, I have noticed that there does seem to be a correlation between even order harmonics and "more" mids.

But back to my original reason for posting this, I am wondering why some circuits respond extremely well to your playing dynamics and others simply don't. One pedal maker out there claims it's because of the OP275, but I am not sure that I buy that. 

I have noticed that very high gain pedals do seem to "flatline" as Allan Holdsworth puts it, meaning that they clip the signal so much that it's kind of like a compressor, play hard or play soft it sounds more or less the same. I dislike that very much in anything that I use. So much so that I just won't use it. But then some very touch sensitive circuits can have pretty high gain, still respond very dynamically to your playing and also sound "open" at the same time. It seems to me that it's in the circuit design, but it's hard to say how just how much of that is contributed by the opamp itself. I would tend to say that it is a very small ingredient. I think it also has a lot to do with how you actually play the guitar, too.

In my testing, I have found that all opamps will clip about the same and it's really all about the circuit that they are in. It seems to me that an awful lot of it has to do with how you treat the mids and highs while keeping the bass from being too mushy sounding. Get that balace right and you get something that sounds very dynamic.

[12Bass]

 However, I'm fairly confident that I could reliably pick up the different chips if comparing, say, a pair of OPA827s against a TL072 in a high fidelity playback system.  The OPA827 sounds subjectively cleaner and clearer, and this is supported by Groner's tests.

Yeaaahhh... I'm gonna have to go ahead an disagree with you there. You're saying you can 'hear' sub-0.01% distortion even amongst the 0.5 to 3% produced by any loudspeaker??
(I guess if you subbed a TL07x into a circuit that drives a few hundred ohms it would suffer, but that's hardly a fair test)

Have auditioned with speakers as well as AKG K702.  The differences are easily audible, IMO, and I would be happy to undergo an A/B/X test.

Curiously, I've tried over a dozen different op amp types, and sometimes really wanted to prefer the "latest greatest" and found it came up lacking.  So, this challenges claims about expectation bias.  And, again, most of my listening has been done outside of pedal applications, where the rest of the system is relatively revealing.  Going through a distorted tube amp and a 12" guitar speaker is much less likely to reveal the sort of nuances I'm talking about.

[familyortiz]

The OP275 has a noise spec that is 1/3 that of the TL072. Just sayin...

[merlinb]

The OP275 has a noise spec that is 1/3 that of the TL072. Just sayin...

That's voltage noise spec. However, the OP275 has 150 times more current noise, and that's what really matters in most stompbox circuits; in most cases this will make the OP275 noisier than the TL072. It's a moot point though, as opamp noise is usually insignificant compared with the noise from other sources in a stompobox.

[searoad]

"touch sensitive"....hmm
I don't know if everyone likes to hear the picking sound, a very clear one. Once I used a chip like that in a hi-distortion pedal and got a very sensitive dynamic picking, but the picking sound is too loud then I had nothing to do just to gave it up.

[familyortiz]

good point merlin. Too subtle to tell overall, but one thing I was thinking about... I was playing with some buffer designs and varying the feedback resistor values found in other deisgns, and using OP284s, I was getting just a bit of 2MHz noise that thickened the trace by about 2x in p-p noise. I was wondering if any designers are designing in just a bit of oscillation like this, controllably? 

[R.G.]

... I was getting just a bit of 2MHz noise that thickened the trace by about 2x in p-p noise. I was wondering if any designers are designing in just a bit of oscillation like this, controllably? 

   
First you have to be *capable* of designing in just a bit of 2MHz oscillation.

It is more likely that your ground clip lead is more than about 2" long, forming a pickup loop than effects "designers" blending in a little 2MHz noise on purpose. It is quite difficult to get clear, solid, no-noise grounding if you're not both doing good hum rejection grounding and good RF grounding at the same time if your scope has a wide bandwidth. It's an acquired skill.

[familyortiz]

No, a feedback value of 10k from your original Hum Free ABY design was giving the oscillations, for the OP284. When I went to a simple zero ohm feedback config, the oscillations went away, so the gnd lead thing, although always something to be careful about, was not the culprit.
Of course the controlled oscillation thing is a crazy idea, but on the scope, the noise definitely thickened the trace for whatever it's worth, soundwise.

[Ben N]

FWIW, I replaced an OP275 with a TL072 in a DOD FX10 boost, and it seems to me that I got a reduction in noise. That is all.

[slacker]

Seen as how Paul is talking about the Black Cat OD-1 and that uses only opamp clipping to get it's distortion it makes sense that different opamps could sound different. Different opamps will clip at different levels and how they clip could sound different.

[R.G.]

No, a feedback value of 10k from your original Hum Free ABY design was giving the oscillations, for the OP284. When I went to a simple zero ohm feedback config, the oscillations went away, so the gnd lead thing, although always something to be careful about, was not the culprit.
Of course the controlled oscillation thing is a crazy idea, but on the scope, the noise definitely thickened the trace for whatever it's worth, soundwise.

Any time you deal with signals that have "mega" or "giga" after them, things get much stranger. Yep, some opamps get into minor parasitic oscillations depending on the feedback configuration, especially around unity gain. But I can assure you on this one that I did not design in a little 2MHz oscillation. It's likely that the particular opamp was having some issues with layout at unity gain. A small (10pf-100pf) cap across the feedback resistor would probably swamp out the strays on the input and squash this.

I have to watch what I understand from reading. When I see the word "design", I think it implies a knowledge of the principles and conscious intent to achieve a well stated objective, as well as the skills to get there. There are some people who design effects, but vastly more who have easter-egged in parts til they get a sound they like, and consider Ohm's law to be too much trouble to learn. So when I read "design" I took that to mean that whomever did the hypothetical effect in question (a) wanted to have the MHz oscillation and (b) understood how to do it reliably.

A well-prepared designer would probably not leave such an oscillation in, because it portends bad results if a component drifts or someone changes a power supply or wire location. A tiny oscillation outside the audio band - well, in the audio band, too, I guess - means that something has gone gain-phase crazy a little bit. It needs only being released from whatever is keeping it tiny to eat the whole operation of the circuit. Unless there is something specific inside the circuit to keep it in bounds, it's an un-skilled design decision.

That does not mean that such a thing would not ever be done. The "bias" on magnetic tape recorders is almost exactly this, an untrasonic oscillator which is mixed with the audio signal as put on the tape to dither the signal around the tap hysteresis. The ultrasonic subcarrier signals on FM radio broadcasts are another example of a high frequency signal riding on top of the audio. But there is a defined, pre-stated purpose for them, and means to limit their extent, as well as ensuring that they happen ever time and in the same way.

So I considered the idea of a designer knowingly putting in a MHZ signal, and reflected that doing this said more about the designer than the design.

And, as you note, changing part values (10K->zero) or opamp made it go away. Parasitics.

I believe you that the trace thickened, and that changing the feedback made it go away. I was just pointing out other common possibilities I've seen.

As to what it's worth soundwise - wow! I wish I could hear 2MHz. Maybe it could dither out crossover distortion in the same way that the tape bias oscillator dithers out hysteresis, but if it did, there would be a means to tune the level of the dithering to get just the right amount.

[Paul Marossy]

Seen as how Paul is talking about the Black Cat OD-1 and that uses only opamp clipping to get it's distortion it makes sense that different opamps could sound different. Different opamps will clip at different levels and how they clip could sound different.

Yeah I was pondering that the lack of clipping diodes probably is what makes the Black Cat OD-1 so responsive to your playing.

[Hides-His-Eyes]

wow! I wish I could hear 2MHz.

You want to be real careful making wishes like that, Mr Keen.

Take a second to appreciate how little you liked hearing 20kHz back when you could, and extrapolate.

[culturejam]

Yeah I was pondering that the lack of clipping diodes probably is what makes the Black Cat OD-1 so responsive to your playing.

Yep. My first thought on seeing that the circuit was "touch sensitive" and had an op amp was that it probably didn't have clipping diodes.

[R.G.]

You want to be real careful making wishes like that, Mr Keen.
Take a second to appreciate how little you liked hearing 20kHz back when you could, and extrapolate.

Dang! You're right! It's one of my personal rules - be careful what you want, you just might get it! 

I did at one time have a friend who could "hear" police speed radar. He heard it as a "pressure" in his ears.

It quit after he had some dental work done. 

[Paul Marossy]

Yeah I was pondering that the lack of clipping diodes probably is what makes the Black Cat OD-1 so responsive to your playing.

Yep. My first thought on seeing that the circuit was "touch sensitive" and had an op amp was that it probably didn't have clipping diodes.

I've been out of DIY for a while, at least not like I used to be into it. I never really thought about this much, but I think that must be it. The creator of the Black Cat OD-1 however attributes it to the qualities of the opamp, and I kind of have to disagree with that.

BTW, I love that a guy like RG will take the time to chime on these topics. Really an amazing wealth of information he has to contribute!