LM308 and OP07 - the plot thickens

[Laundryroom David]

Following on an observation from @Chuck D. Bones re: the audible difference between the LM308 and OP07 in the Rat, I find myself in a rabbit hole. It’s pretty nice here.

So here’s what I hear: the OP07 has, for lack of a better description, a ‘thinner’ tone at low gain setting. I hear this akin to an increased presence, and, to my ears, it’s not a beneficial increase. I dialled in both Rats to sound the same (or as close as I can get, which is pretty GD close) at high gain. Then I rolled back the gain on the 308 Rat and OP07 Rat and that’s when Inheard the difference.

I am not an EE. I am trying to understand why I hear this difference (maybe it’s all a figment of my imagination!). I have looked at the schematic diagrams of the ICs. I get pretty overwhelmed pretty quickly. Clearly there’s a difference when the clipping is coming almost entirely from the IC itself. Maybe i need to breadboard each of the ICs to figure this out

 

[HamishR]

It could even just be down to tolerance variances, especially in the filter pot.

 

[derevaun]

NB @jimilee's Super Rat has an IC-clipping-only option

 

[jesuscrisp]

It could even just be down to tolerance variances, especially in the filter pot.

I have a fancy Rat that can switch between 4 ICs (Drunk Beaver Heavy Bat) and there definitely is a difference between OP07 and LM308 (it doesn't come stock with an OP07 but you can put it in one of the slots which is what I did), and that is WITH ALL OTHER PARTS being the exact same. It's not only about tone, it's also about feel, the 308 is spongier and more saggy at high gain, while the 07 keeps it together a bit more. Some may prefer that, but if I wanted a tight distortion tone, I wouldn't go for a Rat anyway.

Is it something important for some guy building his first DIY clone to play over a Boss Katana? Absolutely not, but saying there is no difference is just flat out wrong and I don't care about simulations or whatever Josh Scott says about the topic.

 

[szukalski]

If you're slamming an IC, there will be differences in how they react. Are those differences worth a rat's ass? That's in the eye of the beholder..

 

[hamerfan]

I also like the TL072 in that rat circuit. Start vid at 3.38

 

[Laundryroom David]

Got my hands on some metal can LM108s. Sounds very good in my Muroidea. Hmmmm

 

[Cybercow]

There's a lot to be said about the low slew-rate for both the 308 & OP07. Each having a slew-rate of 0.3V/µs puts it within the range of the human ear to hear the loss of high-fidelity at such a low slew-rate. The upper frequencies & harmonics are simply zero-crossing faster than the slew-rate can keep up, and the result is that distortion the Rat is famous for. And considering their respective block/schematic diagrams are not completely identical, I imagine there would be some margin of notable differences.

 

[spi]

I just swapped an LM308 into my madbeans rat. I thought I'd quickly pop it in, but it took way longer to get the legs of the round chip lined up to fit the rectangle socket, and forgot what the OP07 sounded like. They both sound like Rats though. I got my mojo on.

 

[DAJE]

All right, ya got me, I'm in.

Just ordered one of these from a local seller. Around US$8 including postage. I will swap out the OP07 in my Rat when this thing arrives. I can feel the mojo increasing already.

 

[Guardians of the analog]

I just swapped an LM308 into my madbeans rat.

Just one bean

 

[Chuck D. Bones]

It's not just about slew rate, although slew rate is definitely a factor in the Rat's tone.
The other factors to consider are:
1. Gain as a function of frequency aka open loop frequency response
2. Output impedance and drive capability
3. Recovery from saturation

The first one is characterized in the datasheets. The datasheets provide graphs showing typical values. With a 30pF compensation cap, the LM308 has around 52dB gain at 1KHz. With DIST dimed, the gain of a Rat peaks out around 60dB and 600Hz. Gain falls off above that due to the limited open loop gain. Not only does the treble roll-off at high DIST settings, but the lack of loop gain means that there is not enough feedback to correct the inherent distortion in the opamp. We can play with the LM308's comp cap to vary the slew rate and open loop freq response. The OP07 has a higher open loop gain and does not support external compensation.

The output impedance varies with frequency. Feedback around the opamp effectively reduces the output impedance. Since the loop gain falls as frequency goes up, the output impedance rises as freq goes up. When the clipping diodes are conducting, the opamp is driving a 1KΩ load. With DIST dimed, the LM308's output impedance is around 400Ω. The OP07 output impedance is not specified. Output drive capability is not well characterized in the LM308 datasheet and not at all in the OP07 datasheet.

The third one is not characterized in the datasheets. You get what you get and the only way to find out is to audition various opamps in the Rat.

What this comes down to is the Rat's signature tone is partly controlled by the opamp's limited bandwidth & slew rate and partly controlled by undocumented features in the opamp.

 

[Chuck D. Bones]

All right, ya got me, I'm in.

Just ordered one of these from a local seller. Around US$8 including postage. I will swap out the OP07 in my Rat when this thing arrives. I can feel the mojo increasing already.

One quick way to tell if that part is legit is to measure the power supply current. An LM308's power supply current will be between 250μA and 300μA. This is not a conclusive test, but if the power supply current is far above 300μA, then it's counterfeit. The other way to verify the part is to measure the slew rate when you vary the comp cap. Need a scope & sig gen for that.

 

[Laundryroom David]

It's not just about slew rate, although slew rate is definitely a factor in the Rat's tone.
The other factors to consider are:
1. Gain as a function of frequency aka open loop frequency response
2. Output impedance and drive capability
3. Recovery from saturation

The first one is characterized in the datasheets. The datasheets provide graphs showing typical values. With a 30pF compensation cap, the LM308 has around 52dB gain at 1KHz. With DIST dimed, the gain of a Rat peaks out around 60dB and 600Hz. Gain falls off above that due to the limited open loop gain. Not only does the treble roll-off at high DIST settings, but the lack of loop gain means that there is not enough feedback to correct the inherent distortion in the opamp. We can play with the LM308's comp cap to vary the slew rate and open loop freq response. The OP07 has a higher open loop gain and does not support external compensation.

The output impedance varies with frequency. Feedback around the opamp effectively reduces the output impedance. Since the loop gain falls as frequency goes up, the output impedance rises as freq goes up. When the clipping diodes are conducting, the opamp is driving a 1KΩ load. With DIST dimed, the LM308's output impedance is around 400Ω. The OP07 output impedance is not specified. Output drive capability is not well characterized in the LM308 datasheet and not at all in the OP07 datasheet.

The third one is not characterized in the datasheets. You get what you get and the only way to find out is to audition various opamps in the Rat.

What this comes down to is the Rat's signature tone is partly controlled by the opamp's limited bandwidth & slew rate and partly controlled by undocumented features in the opamp.

Thanks, Chuck. Fascinating stuff and good to know that the differences I heard in op amps weren’t entirely figments of my imagination

 

[DAJE]

One quick way to tell if that part is legit is to measure the power supply current. An LM308's power supply current will be between 250μA and 300μA. This is not a conclusive test, but if the power supply current is far above 300μA, then it's counterfeit. The other way to verify the part is to measure the slew rate when you vary the comp cap. Need a scope & sig gen for that.

Trusted seller and there's good feedback for his LM308s, so I have no doubt they're real. Looks like he found a stash of them here in Australia and he's been selling them steadily on eBay in the last few weeks. Not sure how I'd test the power supply current - with a multimeter?

 

[Chuck D. Bones]

Yes. Get a DMM that can measure current with 100μA or better resolution. Connect it between 9V and the + power lead of the opamp. You can do this on the actual pedal board or a breadboard. If you're doing on the pedal board and the IC is socketed, just bend pin 7 up so it doesn't go into the socket and connect your DMM to pin 7.

 

[Feral Feline]

Geebuzz!

I've got: metal-can LM308H, a plastic LM308 and an OP07 — I stuck the plastic 308 in my 1590A Filigree Siberian Hamster and left it there, 'cause the metal can won't fit and I couldn't hear a difference in a youtube comparison of plastic 308 and OP07;
as well, another builder of a multi-chip Rat said he couldn't hear a significant difference and after all that build-effort sold his multi-chip Rat clone online.


So now this thread makes me think I better make a new bigger Rat and roll some ICs...


Will my ears hear a difference, or do I have the same average aural abilities as my pedestrian-palate which can't discern a 1978 Cabernet Sauvignon from a 2016...

 

[Chuck D. Bones]

I don't trust YouTube demos. Too many things between the demo person's ears and your ears that will alter the sound. Too many unknowns from things the person in the demo forgot to tell you or chose not to tell you.

 

[Coda]

I second the sleeper TL071 in a Rat…sounds pretty good…

 

[Feral Feline]

Spot on Dave.

I couldn't tell much of a difference in tube-rolling vids, but when my amp-sifu friend rolled a Chinese Gold Lion vs an NOS GE vs an NOS RCA...

Now I'm a believer.


So I will be rolling some ICs...