Rat op-amps.

[RRJackson]

I'd never owned a Rat before, but I'd heard people talk about the original TO-99 LM308 op-amps having a particular sound, so I decided to try one. A few weeks ago I built a Rat clone using an Aion board and a TO-99 LM308 op-amp. And I really liked it. Here's the board while I was assembling it.



I used tantalum caps because the original Rat pedals used them. Also, just to embrace the historical thing.

So I started thinking that I should try an Expandora board, since that's kind of a Super-Rat with a pair of LM308s mated through an optocoupler for wider dynamic range. I used a Fuzz Dog board for that and used modern DIP-8 LM308s. The Expandora also uses a JRC4558 and that was also a modern op-amp.



The Expandora was nice, but wasn't really as inspiring as the original Rat, although it was possible to pick lightly and completely clean up the signal, which isn't something that happens with the Rat.

So this week I decided to build another Rat with another Aion board and a DIP-8 LM308 to prove once and for all that the op-amp package doesn't make any difference. Except the pedal sounded thin and buzzy, which is how I remembered old Rat pedals sounding. It's why I never bought one.

SO just to see if it would make a difference, I removed the DIP-8 op-amp and put another vintage TO-99 op-amp in the new pedal.



Now it sounds like the first pedal I built. It's very lo-fi and very compressed. It actually sounds almost like it's got an Orange Squeezer or something in front of it.

So this got me thinking about the Expandora build and I went and stuffed a pair of vintage Motorola TO-99 LM308 op-amps in it. Now it's amazing. It has the wide dynamic range of the Expandora circuit and I can still make it completely clean by picking lighter, but it's got that same full, lo-fi, compressed thing going on as the Rat builds. Which is contradictory; you can't have wide dynamics and compression, but maybe (and if anyone knows, please tell me) it's two compressed signals handling different parts of the bandwidth being mixed back together. In any case, it's a whole different pedal now.



So here's the thing; I don't see any good reason for these old TO-99 LM308 op-amps to sound notably different than the more recent DIP-8 versions, but they do. Did processes improve before the advent of the later op-amps, because the, "Magic" of the vintage ones seems to be that they're terrible. Not a lot of noise, really. I kind of expected them to hiss a lot, but they don't. However, they have a distinct low-fi thing and the perceived compression seems to be a part of that.

In any case, that's my recent journey through lands I didn't believe existed. Like, I don't buy expensive transistors because some fuzz box made in the 60's used them, because my experience has been that two transistors with similar test results will usually sound pretty much identical. I'm usually looking for hFE and leakage more than branding or encapsulation. So none of this sits all that comfortably with me. Still, that's my experience. If anyone knows WHY I'm experiencing this, I'd love to hear it.

-Rob Jackson

[strungout]

I heard this time and time again on this forum: opamp and tranny choices always come down to low cost of production. They weren't chosen because they had any stellar audio capabilities, but just cause they were cheap and easy to get and did the job 

It's mainly because of the artists that use them and how they use them that creates this magic, mojo, whatever you wanna call it, sound. A good sound is subjective, particular to each person. That's why most of us do like you did: we try stuff out until we get something that speaks to us. But if you want to emulate Hendrix for example, youre more likely to find a germanium fuzz face to be good sounding. I personally, don't like fuzzes all that much, they all sound muddy to me! Fun to play with sometimes tho.

You're just experiencing an encounter with your own personal taste. If that makes sense.

ùEDIT: You made me want to try that TO-99 LM308 though!

[RRJackson]

You're just experiencing an encounter with your own personal taste. If that makes sense.

I do like bandwidth-limited, compressed guitar-sounds. I just can't figure out why the DIP-8 version sounds so much different.

FWIW, I don't think ProCo ever used the TO-99 version in any Rat pedals.

-Rob

[thermionix]

A Honda Civic might be able to haul the same load at the same speed as a '68 Cadillac, but when you slam the doors, do they sound the same?  I know it's not really a popular point of view on this forum, but I tend to think that materials and construction usually do have some relation to the overall sound of a component in an audio circuit.  Some more than others, obviously.  I think some see only circuit calculations and/or data sheet specs.  But you're using your ears and they're telling you the whole truth.  Rock on.

(And now I'm wanting to try both a Rat and an Expandora with those opamps.)

[Markw5]

If it sounds different, have you checked with a scope to see if the waveform looks different, or if there is some high frequency oscillation going on? The latter especially, can make a circuit sound thin and with some ugly distortion.

[RRJackson]

If it sounds different, have you checked with a scope to see if the waveform looks different, or if there is some high frequency oscillation going on? The latter especially, can make a circuit sound thin and with some ugly distortion.

I haven't put it on a scope. I guess I should do that. I know that when I fired up the DIP-8 pedal it was feeding back like crazy and I pulled the gain down as soon as it started. I don't think the TO-99 pedals will do that, for whatever reason. I've had the gain dimed and they haven't tried to shriek like that.

-Rob

[12Bass]

Just a guess: there may be a significant difference in materials and construction between the old and new op amps.  The new ones might even have higher measured performance but sound subjectively worse, as high fidelity is not generally very desirable in fuzz pedals.  If so, then the case is not the determining factor.  That said, I've read many reports saying that the TO-99 version of the LM4562 sounds superior to the DIP in hi-fi applications, though in this case it would seem to be because the metal package has some kind of performance advantage.

[Kipper4]

"though in this case it would seem to be because the metal package has some kind of performance advantage. "

Metal Mojo?

[12Bass]

More metal!

Always better....

[GGBB]

Well now you've got me wondering. I just finished a RAT based build with the LM308H and I am thrilled with it - classic RAT tone but way huge dynamics and clarity. But I thought that was because of all the changes I made. I socketed the IC so now I'll have to A-B it with the LM308N to see if I experience what you did. I would never have believed it, but you have actually listened to the same pedal with different ICs which I'm not sure I've come across before. I have a Vintage RAT with a socket too so I can compare on the original circuit as well.

[GGBB]

So ... I did the A-B back and forth a couple of times and I can't hear any difference between a National LM308H TO-99 and a Fairchild UA308ATC PDIP-8. This was in my RAT-derived pedal - didn't bother trying in an actual RAT.

[RRJackson]

So ... I did the A-B back and forth a couple of times and I can't hear any difference between a National LM308H TO-99 and a Fairchild UA308ATC PDIP-8. This was in my RAT-derived pedal - didn't bother trying in an actual RAT.

It could be that my DIP-8 op-amps were terrible. I mean, that's always a possibility. They could be some counterfeit nonsense.

-Rob

[Markw5]

The metal cases where milspec for use under wider temperature variations and high humidity environments.

On another issue, I have a RAT type pedal with a plastic LM-308. I made an adapter socket so I could quickly try opamps with different pinouts. Decided I like the LM308 best of the ones I tried.

[GGBB]

The metal can alone does not make the LM308 military spec. A look at the datasheet will show the same temperature specs (actually all specs) for both the dip and metal can versions, as well as a note about referring to "RETS108X for LM108 military specifications." Not sure, but I believe the LM308 was never available in a mil spec version.

[StephenGiles]

I'd wager that with a proper drummer playing hard and a bass player with a decent bass tone, you couldn't tell the difference!!

[RRJackson]

I'd wager that with a proper drummer playing hard and a bass player with a decent bass tone, you couldn't tell the difference!!

Eh, most live bands make me wish I had a low pass filter in my ears. It's all that fizz in the high end that makes me love a circuit that can't seem to reproduce it. 

-Rob

[anotherjim]

National Semiconductors LM (Linear Monolithic) mass production base spec devices are LM3xx, better are LM2xx and best of all LM1xx. I have never seen any LM2xx in the wild. In data sheets, 1xx and 2xx usually share the same tables while 3xx gets it's own set. Application circuits often only showed 1xx devices. 3xx have narrower temperature range 0deg to +70degC. 1xx can cope -55deg up to a sweltering +125degC. 2xx somewhere in between.

Early op-amp designs could easily suffer instability (oscillating), sometimes needing little excuse. Try a uA709 (no don't - only joking). The extra lead length of the T0-99 might be enough at high gain.

[Markw5]

The metal can alone does not make the LM308 military spec. A look at the datasheet will show the same temperature specs (actually all specs) for both the dip and metal can versions, as well as a note about referring to "RETS108X for LM108 military specifications." Not sure, but I believe the LM308 was never available in a mil spec version.

Ok, but I believe cans are hermetically sealed whereas plastic packages, at least in the past, were subject to some moisture absorption or diffusion, and also some surface track leakage currents due to environmental moisture or humidity. Believe the newer plastics are better in this regard. Once is a can, though, there is also probably some difference is heat transport and that could affect some other specifications.

[PRR]

Any difference is not in the package, but on-going process changes in the die.

However the TO99 to PDIP is also an "on-going process change" and may have gone on about the same times the Silicon processes were evolving.

LM308 is a very old device. When new, only the vacuum-tube seal technology could be trusted not to spoil the low-low bias current of the 308. But in fact the glass-metal seals were far from perfect. We used to use kitchen pressure cookers to sort-out the worst leakers. Cheapskates in the production office and brilliant chemists evolve epoxy systems which equaled then bettered the metal-glass packages.

108/208/308 were not made different, they were sorted, and maybe not that. If a device barely-failed 108 specs, it could be sold as 308. However the 108 was designed FOR the silicon process of the time. If you didn't get nearly all 108 grade dice out of a wafer, Something Was Wrong. It had to be fixed before an order for cash-cow 108 came in and you couldn't actually hit the specs. Clean the masks, refresh the etchant, swab the diffusion chambers, whatever it took. I suspect most "308" were really 108-grade dice but not sold for $10+ so not marked as high-grade.

[ElectricDruid]

I know that when I fired up the DIP-8 pedal it was feeding back like crazy and I pulled the gain down as soon as it started. I don't think the TO-99 pedals will do that, for whatever reason. I've had the gain dimed and they haven't tried to shriek like that.

This part makes it sound to me like there's some significant difference of spec between the early and late versions of the amp, not just a change of package.

Your comment about fakes is a possibility too. It'd be easy to grab some cheap op-amp, wipe it clean, and then mark it up as something vintage that people will pay significant money for.

Tom