Substitutes for LM308?

[mozz]

Lm108 or Lm208 are both mil spec and will work.

[Elijah]

Guys. Has anybody tried Russian UD1408A (УД1408А)? Slew rate is like 0,05.

[iainpunk]

welcome to the forum, if you go with the 741, you definitely don't want that cap in there.
in double blind tests, most people didn't hear differences between my LM308 and OP07, but they seemed to prefer the 741's sound.

a key element in the sound of a rat is the SLEW RATE, this is how fast a chip can go up or down in voltage. the LM308 and OP07 both have 0.3v/us, the 741 has 0.5v/us. these are considered really slow, even by 1980's standards, the slowness mellows out the sound of the pedal a bit in a way a filter can't.

cheers, Iain

[D_Ex_Patria]

OP07 is pretty standard. LM741 isn't 'exactly' the same slew rate, but will probably sound great.

Another option, if you're willing to tinker more, is the NE5534 with an oversize compensation cap. I haven't tried this method yet, but I've heard other people on the forum using it with some success. it would take some trial and error to get right. I've been meaning to test it out with a scope myself.

[Fancy Lime]

A word of caution on the OP07: in a Rat, the chip itself clips. Some sound good when they do, others not. The problem with opamp clipping is that the chips are not designed for it. What's more, the chip names (LM308, OP071...) only signify the minimum specifications of the original spec sheet. That means that two chips of the same specs and name (two OP07 or two TL071 or whatever) from two different manufacturers are not necessarily technically identical. They are usually very similar but with things like clipping that occur way outside anything that appears on the datasheet, there can be significant differences. All that is a long winded way of saying: many people seem to like the OP07 in a Rat and the ones in the actual ProCo units sound good to me. But the OP07 that I have sound absolutely horrid when clipping. Fizzy, harsh, inharmonic, just awful. And no amount of circuit tweaking makes it better.

My favorite chip in a Rat, even before the LM308 is the LM301. NE5534 is really good too but needs some circuit changes.

Andy

[iainpunk]

OP07 is pretty standard. LM741 isn't 'exactly' the same slew rate, but will probably sound great.

Another option, if you're willing to tinker more, is the NE5534 with an oversize compensation cap. I haven't tried this method yet, but I've heard other people on the forum using it with some success. it would take some trial and error to get right. I've been meaning to test it out with a scope myself.

the NE5534's compensation cap isn't slew rate changing, but its more of a normal filter function.

cheers.

[PRR]

....the NE5534's compensation cap isn't slew rate changing, ....

Sure it is.

[MR COFFEE]

Well, here I go again...

I second the 5534 with more capacitance on the compensation capacitor. The 741 is just too damn noisy, unless you think hiss makes your sound "tube-ish". The TL070 has gotten really rare now, which is too bad for lovers of the Rat.

In substituting op amps, I've noticed that  the output stage "sticking to the rails" when overdriven may also be part of the "sonic signature" of the old LM308 in the Rat circuit, and most modern op amps go out of their way to design that out (and do so quite successfully). The '308 is Really noisy (100+ nv/hz), and modern super-beta input op amps are totally different animals from the '308 - much quieter, modern-day slew rates, and overload gracefully.
So if you want to get the "spit" the fans of the Rat seem to love about it, you want to select an older, externally-compensated op amp. Maybe one like the NE5534, as suggested by D_Ex_Patria, which is low noise, externally-compensated, and IIRC, some manufacturers versions did stick to the rails (although I can't remember which any more), might be a good choice to make yourself a good modern day Rat.

The OP-07 is pretty quiet and slow, although the input current is compensated to get a low input-stage current, which may now work out so well in the Rat circuit, since it is one of the op amps where balanced input resistance is pretty much necessary to avoid major output offsets.

If a '5534 seems too hard to obtain, or doesn't "spit" enough for your taste, you could try an old LM301a that could easily be over-compensated to lower the slew rate to act like the '308, and since it has a very similar output stage to the '308,  it may well "stick to the rails" in a similar fashion to get that "spit". And IIRC, it still is quieter than a '741.

It's not that hard to try out different compensation caps to find what sounds good before you "button it up". Put in a IC socket for the op amp, and use two machine pin sockets to plug in various compensation caps while you listen to how different values make it sound. You can solder the "golden" cap in to the socket pins if you like, but unless you're using a big old silver mica cap or something, the socket pins will hold it just fine through on-the-road rock and rollin'.

Let us know what you decide sounds best to you afterwards. Your findings may end up in the design of the iteration of Rat pedals!

Bart

[Fancy Lime]

My Texas Instruments NE5534 do something additionally interesting in a Rat. If the bias resistor is too large to supply enough bias current and the input coupling cap is small, then the opamp clipping becomes strangely fuzzy and dark. In a good way, I really like it. The reason, I speculate wildly, is that when the output hits it's limits (or one of them), more current is drawn from the inputs than should be as per the datasheet. This causes the non inverting input to drift, making the signal very asymmetrical. If the input cap is large, it can temporarily supply some charge to counteract that. For a noticeable effect compare a 22n input cap and 2M2 bias resistor with a 220n cap and 470k resistor. This may not work with all NE5534 from different manufacturers or even different batches.

Andy

Edit: mind you, this effect becomes less noticable if you overcompensate too much.

[Rob Strand]

My Texas Instruments NE5534 do something additionally interesting in a Rat. If the bias resistor is too large to supply enough bias current and the input coupling cap is small, then the opamp clipping becomes strangely fuzzy and dark. In a good way, I really like it. The reason, I speculate wildly, is that when the output hits it's limits (or one of them), more current is drawn from the inputs than should be as per the datasheet.

Biasing shifting on the NE5534 is a common issue with the large resistors.    It does have a significantly higher bias current than most opamps.    (FWIW, the app notes have a scheme to inject current into the inputs.)
     

[bmsiddall]

I built a dual op amp rat last year (switchable from LM308 to CA3130).  Here are my notes/musings from the op amps tried - tests done via headrush gigboard, fendery cleans and headphones.  Note that I ended up bumping up the pin 6-2 feedback cap on the CA3130 to 1nF to kill the oscillation after 2PM on the distortion dial.  I lost some of the aggressive pick attack but it's still noticably more aggro than the LM308.

cheers,
Brett

LM308
•   Stock, it's a warm, rich tone.  Some settings a little woolly?
TL070
•   I used a 150 pf compensation capacitor.  Very hard to tell the difference between this and the LM308 using headphones and floor modeller.  Makes an excellent and classic Rat.
TL071
•   obvious difference to stock chip.  Retains a lot more high-end and has a metallic pick attack.  Thinner, probably more aggressive tone.  I like it as a point of contrast.
LM741
•   Quite similar to LM308 and makes an excellent Rat.  Retains a little more high-end - possibly a little more focused/less woolly than the LM308 at identical settings.  I really like it.
CA3130
•   Wow!  Needs at least a 330 pf compensation cap to tame the high end.  But when you get there, it's great.  More aggressive, modern and gainy than the LM308, with an oscillation whistle at about 2pm on the distortion pot on one pickup (middle position ok).  The additional gain is especially noticeable when you wind the guitar volume back a little- it certainly cleans up slower than the stock chip.  A great option if something a little wilder is required.  Personally, I think it's my favourite.
NE5534
•   less drive and thickness but more clarity/crispness than LM308.  Retains an aggressive pick attack with more high-end and would certainly cut through a mix.  Perhaps veering more towards overdrive.  Probably the least rattiest but a nice contrast with the stock chip.
LM301
•   very close to the 308, but with a little more clarity on the unwound strings and certainly less wooliness at high distortion settings.  I like this and it would make an excellent rat

[iainpunk]

I built a dual op amp rat last year (switchable from LM308 to CA3130).  Here are my notes/musings from the op amps tried - tests done via headrush gigboard, fendery cleans and headphones.  Note that I ended up bumping up the pin 6-2 feedback cap on the CA3130 to 1nF to kill the oscillation after 2PM on the distortion dial.  I lost some of the aggressive pick attack but it's still noticably more aggro than the LM308.

cheers,
Brett

YESSSS the CA3130 is so great! especially when omitting the original clipping diodes, since this opamp has a CMOS inverter as output stage, it does genuine soft clipping.
am currently building/designing a distortion/overdrive using the CA3130 or CA3160 for its perticular output characteristics, and the fact that it can be exploited for great sounding tones when you have only the output stage in an open loop configuatation.

cheers, Iain

[GGBB]

I have some LM208, anyone use it?

Saludos, Marcelo.

Same chip - tighter specs. Ditto for LM108.

[GGBB]

There are other "308" chips if you can find them at a reasonable price. Or at least one that I know of (and own) - Fairchild UA308.

[GGBB]

Lm108 or Lm208 are both mil spec and will work.

Actually that's not accurate. They are higher spec than 308, but the manufacturer data sheets do not indicate that they are "military" spec. Instead, they indicate to "refer to RETS108X for LM108 military specifications". I think this means that a military spec approved LM108 was available, but stock LM108 units are not tested for military spec ratings. TI chips that are truly military spec usually have an "M" in the suffix, but I don't know if that was also true with other manufacturer's like National (whom TI bought out).

[mozz]

Well here is the numbers. Better temp difference is always in the mil spec compared to a consumer part.
The LM108 is guaranteed from -55C to +125C, the
LM208 from -25C to +85C, and the LM308 from 0C to
+70C.

Here are the mil numbers.

JM38510/10104 LM108

Here's a cross reference from mil to consumer

JAN MIL-M-38S10 - bitsavers.org

[Kevin Mitchell]

I have a lead for purchasing NOS LM308 chips. If in the US feel free to PM me.

[D_Ex_Patria]

I wanted to take a moment just to thank everyone who chimed in with details about each of these alternatives. I know this was a thread specifically about substitutes, but I'm scribbling furiously in my notebook about op-amp properties that could come in handy outside of just replacing the 308.

[iainpunk]

I wanted to take a moment just to thank everyone who chimed in with details about each of these alternatives. I know this was a thread specifically about substitutes, but I'm scribbling furiously in my notebook about op-amp properties that could come in handy outside of just replacing the 308.

there are two factors i consider when using an opamp for clipping, the ''knee'' of the clipping, and the slew rate. i have been looking in to extremely slow opamps (like the MCP6042 dual) and slew rate reduction circuits to use as gain stages or after another distortion stage.
for the knee, you can use a CMOS (ca3130 ca3160, both quite expensive) output opamp with a relatively heavy load (about 1 or 2kohm) to soften the curve, or use a cheaper cmos inverter IC.

the down side is that the cmos knee and slew rate reduction are hard to have at the same time without making it really complicated.

cheers, Iain

[thetragichero]

purely anecdotal and i have been told that i'm totally wrong by folks, but in the circuit i built lm308 and lm318 both sounded the same to me. both were about the same price from some us-based ebay seller when purchased a year or two ago