OPA2134 Burr Brown

[HamishR]

Oh dear - I think I'm starting to like this chip! Is this bad? I don't normally notice much difference between chips and will happily use a TL072 where the circuit calls for a JRC 4558/9 or whatever... I don't think it really makes much difference. But a few days ago I built myself a dual OD with a couple of Rockett style ODs in there and used JRC4558s in it. Then I compared it with a similar pedal I had previously built with those two exact same circuits in it and found I preferred the older one - which had OPA2134s in it.

So I swapped the chips in the new pedal to OPA2134s. And I like it better now - it just sounds a hair more solid and goes more smoothly between clean and dirt. Surely I am imagining this?

If using the OPA2134 improves my enjoyment of these pedals I guess I'm ok with that. Yes it is a lot more expensive than a TL072 but really - Tayda sells a OPA2134 for US$3.59. I think I can afford that for my own pedals. Is this cork-sniffery?

 

[music6000]

I'm afraid it's all in your imagination, You have put a Hi-Fi op amp in your pedal so your brain thinks it sounds better.
While you were not looking, I snuck the JRC4558's back in the pedal and you have not picked up on it!
Hi- Fi purist will tell you that you can put anything in a pedal and it will sound the same....
Of course they could be Cork Sniffers not Guitarists !!!

 

[HamishR]

It really wouldn't surprise me if it was all in my head. I was worried about that but it's best to confront these things.

And keep outa my pedals!

 

[Chuck D. Bones]

It's possible you are hearing a difference. Most opamps make a tiny bit of crossover distortion. The OPA2134 was specifically designed to not do that. Here's an old trick used in hi-fi preamps to get rid of crossover distortion: bias the output into class A. You can do that by installing a resistor from pin 1 to pin 4 on IC1. 2.2K is a good value to use, but anything from 1K to 10K will probably work. No point doing that on an OPA2134, but it could improve most other opamps. YMMV.

I have some $50 TL072s for sale. If you like expensive chips, you're gonna LOVE those.

 

[zgrav]

I guess $3.59 for the cork is not too bad compared to tracking down NOS for some rare pedal builds. Sniff away. We certainly put enough extras into the eye candy for many of these boxes, so a bit more for the ears is well spent if it makes you feel good.

I doubt if the remainder of the sound chain from the pedal through the amp will necessarily preserve those fine points, but I could see it making a difference if you were recording from that pedal directly.

 

[HamishR]

I guess I like hi-fidelity distortion.

Most of this stuff is a mystery to me. Most times I'm amazed that it all actually works when I plug it in!

 

[Chuck D. Bones]

Most times I'm amazed that it all actually works when I plug it in!

We are too.

 

[music6000]

I believe this sounds 2 X's Better than it did before.
How High can you Stack Em !!!

 

[jimini]

How much could the difference of the slew rate count?? Especially thinking of the old LM308 versus some modern opamp

 

[HamishR]

Aha! One I know a little bit about - well I know nothing about slew rates - but I use the CA3130 instead of the LM308 and it works perfectly. I figured that seeing as BJF pedals are often much modded Rat circuits and he uses the CA3130 how bad could it be?

I thought slew rates were how much your car slid around on icy roads.

 

[jimini]

It s ok if it works for you , my understanding was that this added to the opamp distortion in the rat, I could be wrong

 

[Chuck D. Bones]

I thought slew rates were how much your car slid around on icy roads.

No, you're thinking of Street Curling.

 

[zgrav]

I would be interested in anything showing that slew rates made audible differences in effects pedals. Seems unlikely.

 

[Chuck D. Bones]

Slew-rate limiting is a real source of distortion, let's see if it could come into play in a pedal...

The maximum slew-rate of a sinewave is πfA where f is the frequency in Hz and A is the amplitude in volts p-p. An 8Vp-p 1KHz sinewave has a peak slew-rate of 0.025 V/μs. Even if that signal is overdriven 10x, it's only 0.25 V/μs. I picked 1KHz because that's close to the fundamental tone of the high E string on the 20th fret. The LM308 is externally compensated, that's the capacitor between pins 1 & 8. That capacitor stabilizes the opamp by limiting the gain at high freq. It also limits the slew-rate. A 30pF cap limits the slew-rate to about 0.12 V/μs. So it's possible that under extreme conditions, the LM308 does slew-rate limit the signal a little bit. But it requires the opamp to be overdriven which by itself makes plenty of distortion even without slew-rate limiting.

All opamps have a finite slew-rate, most have their compensation cap on the inside where we can't fiddle with it. Here's a comparison of typical slew-rates:
LM308 (30pF cap): 0.12 V/μs
JRC4558: 2.2 V/μs
LM833: 7 V/μs
CA3130: (56pF cap): 10 V/μs
TL072: 13 V/μs
OPA2134: 20 V/μs

Keep in mind that once you have more slew-rate than you need, nothing is gained by having an even higher slew-rate.

There is something else going on as well. When an opamp is overdriven, the input stage saturates. When the signal swings back into the input stage's linear range, it takes time for the input stage to come out of saturation. The time it takes is influenced by that external capacitor, among other things. Different opamps react differently to being overdriven. Some come out of saturation quickly and smoothly, others are slow to recover or make large amounts of harmonics during the recovery. I think this is the primary reason why different opamps sound different in some dirt pedals, particularly the pick attack when the signal is the loudest. The effect is most pronounced in hard limiting pedals like the Rat, Expandora, Distortion+, DS-1, Crunch Box, etc. because they intentionally drive the opamp(s) into saturation. Soft limiters like the opamp Big Muff, TS, Timmy, Sweet Honey, etc. don't drive the opamp to saturation and are much less sensitive to the opamp selection.

 

[zgrav]

Chuck --- thanks for that detailed reply. Very interesting. My user experience with discussions about slew rates was talking about the relative performance of stereo amplifiers where the goal was to minimize audible distortion rather than create interesting distortion in an effect pedal.

 

[HamishR]

As usual, great explanation Chuck. I understood all of it except the second paragraph.

 

[Chuck D. Bones]

It's simple. d/dt A*sin(ωt) = A*ω*cos(ωt).

 

[chongmagic]

It's simple. d/dt A*sin(ωt) = A*ω*cos(ωt).

I didn't know you could speak Greek Chuck!?!?!?!?

 

[HamishR]

Um... E=MC2??

 

[Nostradoomus]

I’m pretty sure I see butts in that equation.