Op amp slew rates

[jm22]

I'm not a pedal builder, but in any event I've been reading about op amps in pedals like the TS808, The Rat, etc., and the slew rate of the pedal's op amp emerges as a defining characteristic of a particular pedal's sound. I know slew rate varies considerably from one type of op amp to another, but what about the same model and manufacturer? If ten random Motorola LM308's from the eighties were measured, what kind of variation in slew rate might one expect to see across the ten?

[Mark Hammer]

I'm of the rather cynical view that slew rate matters much less than people think it does.  If one is designing or building a preamp for music reproduction, or a mixer, where lots of gain is to be applied to something with very wide bandwidth and a broad dynamic range, then being able to provide voltage increases at very fast speed is likely pretty useful.  But rock/pop guitar has fairly limited bandwidth, and generally modest dynamics, such that the ability to produce big voltage swings very very quickly is of no particular use; especially since the majority of guitar pedals are designed to operate off 9VDC, with a floating ground.  There's not a whole lot of voltage swing there.

Having said that, limits to open-gain bandwidth can sometimes work in one's favour, by reducing the amplification of upper harmonic content.  While not equivalent, gain/bandwidth product and slew rate are somewhat related.  Slew rate describes the capacity for instant voltage swings, without any direct reference to gain, while gain/bandwidth product indicates just how much of a voltage gain one can expect at different frequencies.  In some instances, gain/bandwidth product (the same thing as "open loop gain") can be the more informative thing to look at, when you have a better sense of what sorts of gains the circuit is aiming for, or requires to achieve its particular sound.  A bunch of different op-amps may well behave identically when providing less than 24db gain, but start to pull apart if one exceeds 30db.

[marcelomd]

Taking JRC4558 for example. The stated worst case slew is 1.2V/1uS.

Someone correct me, I'm bad at math, but this is a theoretical 60V swing at 20kHz, right?

Same calculation for LM308's 0.3V/uS gives 15V at 20kHz.

That is so beyond what a guitar pedal needs that a bit more, a bit less will make little difference, I think.

[jm22]

I can't dispute your observations Mark, but anecdotally, I've been going down the rabbit hole of Rat pedals, and after spending some time with a friend's '87 with a Motorola LM308, I read up on the various iterations of the Rat and discovered there's two camps, one that essentially says the op amp makes a difference, and the other that says the newer Rats with the OP07 op amp sound the same.

Hoping the latter group was correct I bought an inexpensive newer Rat with the OP07 and A/B'd it against my friend's Rat. No comparison. Might as well be two different pedals, with the early one being vastly superior. I then bought an '88 with a Motorola LM308, and while it was also superior to the new one, it still wasn't as good as my friend's, which made me start thinking about what could cause such a variation in the quality of tone if not some parameter of the op amp.

I have an '86 Rat on the way from ebay, and oddly it has a Motorola LM308 dated 1982. How a 1982 chip ended up in a pedal made in 1986 is anyone's guess, but I remain on the hunt for the sound of my friend's '87 and continue to think about what sets certain units apart from others.

[antonis]

I do second Mark's "cynical" view.. 

IMHO, LM308 versions exhibit a non-linear frequency responce during signal amplification. At high gain levels, LM308 can't amplify high frequencies to the same level as low frequencies..
(a signal generator and an oscilloscope can ensure you..)

All time classic: 
https://www.diystompboxes.com/smfforum/index.php?topic=127737.0

P.S.
Have you ever wondered why there is no LM308 soft clipping remarkable configuration (like Tube Screamer..)??

[Kipper4]

Yer man I hope the ebuy is your dream rat sound.

I can also imagine Jimi Hendrix buying a fuzz, He wouldn't settle for the first one without trying some more before handing over a wad.
Pedals do vary so much. At least you're using your ears for the right reason.
Electronic component tolerances aren't always on spec especially when employing Mojo parts.
The variation of componants real measured values might contribute to the golden sound.
Atmospheric condition may contribute in some cases too.

Personally I prefer my op07 to vintage 308 I have.
Just think of all those great punk rock songs made with a rat. Blows my tiny......


Good luck
Rich

[antonis]

Good luck
Rich

Yonks man.. 

[Gus]

With a RAT the two small value caps in the -input to ground make a difference.

The older ones I saw were tants(good for a RAT)
What did the OP07 have for caps at the locations?
I would not use aluminium electros at the two locations. I am not a fan of 10uf or lower value electros in some places in circuits like a RAT

OR

You could use films at the two locations

[marcelomd]

... How a 1982 chip ended up in a pedal made in 1986 is anyone's guess...

Economies of scale in electronic components do that kind of thing.

Weird would be a 1986 chip in a 1982 pedal =)

[PRR]

The '741 was almost good enough for 20kHz at 1V in hi-fi and home studio.

Walt Jung wrote an extended analysis and comparison; I dunno if he still has it online.

Guitar rarely gets to 1/4 that level or 1/4 that frequency.

And distortion is not unwanted.

You are probably hearing what is NOT on datasheets. How it flips-out when overloaded. Phase reverse? Hiss duck? Thump?

[PRR]

Walt Jung wrote an extended analysis and comparison; I dunno if he still has it online.

https://hifisonix.com/wp-content/uploads/2018/03/SID_and_TIM_W_Jung_77-79.pdf

[FSFX]

The '741 was almost good enough for 20kHz at 1V in hi-fi and home studio.

Walt Jung wrote an extended analysis and comparison; I dunno if he still has it online.

Guitar rarely gets to 1/4 that level or 1/4 that frequency.

And distortion is not unwanted.

You are probably hearing what is NOT on datasheets. How it flips-out when overloaded. Phase reverse? Hiss duck? Thump?

The 741 has a slew rate very close to that of the OP07 and LM308. In fact a lot of the fake LM308s test out as being 741s. However, the original 741s and the more recent ones are quite different when you do a full I/V analysis.

So it is not just slew rate that is the only factor here in the sound. The hard clipping of the overdriven LM308/OP07 in the RAT means that, as you said, how the op amp behaves when overloaded is more likely a factor in the resultant sound.

[jm22]

With a RAT the two small value caps in the -input to ground make a difference.

The older ones I saw were tants(good for a RAT)
What did the OP07 have for caps at the locations?
I would not use aluminium electros at the two locations. I am not a fan of 10uf or lower value electros in some places in circuits like a RAT

OR

You could use films at the two locations

Haven't looked, but I think I'll do some experimenting on the new Rat, and I'll try your suggestions.

[andy-h-h]

Here's a man that knows RATs.  https://tymguitars.com.au/blogs/blog/lets-talk-about-rats

Somewhere between serial number 235390 (pot dates 49th week '95) and 243304 Pro Co started fitting the new IC. On paper it's VERY similar to the 308. I could bore you with slew rates and gain/freq bandwidths between these 2 ICs and several others like the TL/LM/NE/etc series but I won't. The reason I won't is that I own those 2 serial number RAT2s. They are both IDENTICAL except for the IC and pot manufacturer and they both, wait for it, sound identical to me.

As someone who makes pedals, and makes RATs I'd like to think I know a little about what I'm hearing. I have never claimed to have golden ears or know "good tone" (no such thing) but I can hear 2 pedals side by side and tell if there's any real difference.

[Mark Hammer]

A few years back, a local buddy brought his Timmy pedal over, because he had heard/read things about other chips sounding "better" than the higher-quality dual op-amp the pedal came with.  I unsoldered the existing chip, installed a socket, and we plunked about 6-8 different chips in it, to see if any sounded better to our ears.  Much to my great surprise, I preferred the LM1458, and so did he.  Its limitations yielded a warmer overall tone in that context.

[Elektrojänis]

If you compare two whole pedals, you are not comparing only the opamps. The other parts could make a lot bigger difference than the chip. It is very common to use potentiometers that have tolerances of 10% or 20%. Same for capacitors... Nowadays many use resistors with 1% tolerance, but 5% is still common and 5% was about only thing that was used in something like pedals in the 80's. There is one opamp in the rat and dozens of those other components and the effect of those tolerances can stack up...

Other thing is that opamp circuits are usually set up with components around it so that the other components define how the opamp is to work. This tends to minimize the effect of opamp differences and the result will be affected more by the external components.

It would be nice to measure all the components of the "magic unit" and try to replicate it, but it would have to be dismantled and I'm pretty sure the owhner wouldn't like that.

Building your own will enable you to experiment. You can socket the chip and try different ones. You can try also swapping diffferent value resistors and caps and see what you like. You might even end up with something you like even more than the "magic unit".

Edit: The LM308 and many other opamps use an external compensation capacitor. If I've understood it correctly, it affects bandwidth and slew rate. That one probaly has 10% or 20% tolerance too... How about experimenting with that? Try 22pF or 47pF instead of 30pF.

[FSFX]

Edit: The LM308 and many other opamps use an external compensation capacitor. If I've understood it correctly, it affects bandwidth and slew rate. That one probaly has 10% or 20% tolerance too... How about experimenting with that? Try 22pF or 47pF instead of 30pF.

Another thing is that most pedal implementations use a capacitor in the feedback from the output to the inverting input so creating an integrator type of circuit. This is similar to that of the Miller capacitor. In fact the compensation capacitor works in a similar way but is usually just on the second internal stage of the op amp but it too makes that stage an integrator in conjunction with the constant current output characteristics of the first stage of the op amp.

All this is explained in detail in the Texas Instruments (TI) Precision Labs tutorials on op amps that can be found here:   https://www.ti.com/video/series/ti-precision-labs-op-amps.html

Even a fast slew rate op amp like the common TL071 can be 'slugged' with a Miller capacitor to give it a similar gain-bandwidth of the op amps like the LM709, LM741, LM308 and OP07. With a RAT circuit it takes a really large capacitor like a 2.2nF to get close to the LM308 bandwidth.