OPAmp substitution...mixed results

[reubster]

Over the years I’ve had a lot of success with opamp substation in hifi equipment so I thought my Peavey MX-VTX [SS preamp, 6L6OP] would be a good candidate.

I replaced the 5 x 4558’s with OPA2134’s and got mixed results.

If I played gently, they sounded much better: quieter and cleaner with a more open “valve-ish-ness”
However, when I dig in there is an unpleasant distortion/sag effect.
It seems transient: Following the “pick” attack [as the note fades] the distortion goes.

The power supply in this amp uses basic zener regulation that requires 250 ohm resistors in each of the supply rails.
Also, there are no filter+filter bypass caps mounted close to the supply pins on any of the op amps.
I suspect the OPA2134 opamp does not like this supply arrangement, I tripled the filter capacitance and got a very noticeable improvement, but the problem still persists. .

For the gain in sound quality, I think the substitution is worth persuing.
I could do either or both of the following:
1. Replace the supply with fixed regulators.
2. Create 5 plug in DIL boards with OPA+ low ESR filtercaps

Electronically, this is straightforward, but mechanically and geographically these things are always more challenging.

So just wanted, some opinions and suggestions before make the final decision.

Cheers

[Robert M. Martinelli]

Hi,
have you tried to take some readings with an o-scope to check the waveforms and track down where the "unpleasant distortion" begins?

Anyway....I think at least one of the reasons is the very high slew rate of the 2134s.....they have a SR of 20V/uS while the good 'ol 4558s only have 1,7 V/uS, this means 2134s have a much better and faster transient response/capability, so they could overload the following stage during transients ( e.g. note/chords attack )

Hope this helps

Best regards

Bob

[Enzo]

I am with Bob here. get out your scope. Don't assume anything, don;t guess at the power supply. Why would we assume the filtar caps needed to be closer to the ICs? Just tripling them back where they were might also make this difference. For that matter, why not replace the zener regulation with plain old three leg Vregs.

But what leaped out at me was "I replaced the 5 x 4558’s with OPA2134’s." SInce each IC does a different job, and if PV had not already installed sockets, certainly you could. Swap them ONE at a time

Always keep in mind that guitar amps are NOT HiFi. Guitar amps are not REproducing sound, they are a primary SOURCE of sound - they are part of the instrument. That is why someone tells you he plays a Marshall or a Fender. If they were HiFi, they'd all sound the same.

[Robmosis]

Go an read the interview section in Dave Hunters book "The Guitar Amo Handbook". There is a good section talking about hifi amps compared to guitar amps. How guitar tube amps are not exact instruments and how trying to hifi them never gives really desirable results do to the characteristics of the guitar itself. Its short but is a great opinion.
Rob

[reubster]

Thanks Robmosis,

I'll look out for that article.
I took on board Bob and Enzo's comments, which I suspect will be inline with the article you have recommended.

Nevertheless, I bought this amp to play around with and the tinkerer in me got the better.
I've since upped the Opamps, regulators and the filter capacitance and replaced all signal path caps with better.
I must say when playing with distortion pedals or the onboard distortion there is a big improvement in clarity and openness.
However, when I play clean [particularily with a strat] I still get some kind of harshness when the notes are picked aggressively. I'm guessing the higher "fidelities" are revealling more of the guitars response [and my sloppy playing technique]
Overall though, the mods are a keeper.......

[tubeswell]

T
However, when I play clean [particularily with a strat] I still get some kind of harshness when the notes are picked aggressively.

IMHO that is the difference between transistors and tubes. Transistors can't soft-clip like tubes can, so you will always get harshness on clipped signals with transistors.

[Robert M. Martinelli]

IMHO that is the difference between transistors and tubes. Transistors can't soft-clip like tubes can, so you will always get harshness on clipped signals with transistors.

Hi tw,

I agree, no SS amp can replace a tube amp!


OTOH, in this case I think it's more a matter of different slew rates and transient response capability, as the original op-amps ( 4558s ) are BJT op-amp as well, but with a slower slew rate ( 1.7 V/uS, while the new ones' slew rate is 20V/uS ).

A Strat, with its great "attack" makes this phenomenon more audible IMHO; however, if sockets ( for easier / fast replacement ) have been used, it could be worth trying some JFET op-amps as well ( TL072s and the like ) to try to "sweeten" the sound somehow.

Cheers

Bob

[Steve Conner]

I never understood the point of op-amp substitution. Surely they all sound the same in their linear range, as they all have THD in the sub 0.1%. You're not going to hear the difference in distortion products between a TL071 and a NE5534, or whatever.

You're not going to hear slew rate either, as they all have enough of it to reproduce a 20kHz full-scale sine wave without slew limiting. The slew rate of a sine wave is just (IIRC) two pi times the frequency times the peak voltage: for 20kHz and 10V peak that's 1.2V/us.

They might sound different when overdriven, but all op-amps sound horrible when clipping, so why would you want to go there? If you want to sweeten things, try discrete JFETs.

Or just leave it alone. Different op-amps have different strengths and weaknesses, and a good designer chooses the best (ie, most cost-effective) one for each socket. And Peavey in particular always seemed to have a lot of good designers.

If you just replace them all with whatever part is cool on diyaudio.com this week, without understanding the reasons for the original choices, you're asking for trouble. And judging from this "unpleasant distortion" you mentioned, you got it.

[Robert M. Martinelli]

Hi Steve,
while I can understand your "hate" against op-amps, ( we all already agreed nothing can take the place of a good tube/valve amp ) I must honestly say I've tried op-amps substitution in the past and got good results in many cases, especially talking about noise reduction as the noise figure ( equivalent input noise voltage ) was "awfully" high in old op-amps.
Cheers
Bob

[Steve Conner]

Hi Bob,

I'd have to agree with that, lower noise is about the only valid reason I can think of to swap out op-amps.

Even so you have to be careful. For instance, the NE5534 is lower noise than a TL071, but it also has much higher offset and bias currents, because it's entirely bipolar whereas the TL071 has a JFET input stage. So in some circuits, it might not work as a substitute for the TL071.

Also, some precision/low-noise op-amps may have lousy clipping characteristics. If they achieve their precision and low noise by very high open-loop gain, this means a low open-loop pole frequency. What this means in non-geek-speak is that they can charge up their internal compensation caps and "hang" on the rail for quite a while.

Maybe this is the cause of the OP's unpleasant distortion on pick attacks: the cure is to add a clipper circuit, such as diodes in the feedback loop, LEDs or whatever, so the op-amp itself never saturates.