Purpose of diode at input of op-amp buffer

[kurtlives]

What is the purpose of D8 at the input to this op-amp buffer pictured above? To me it looks like it could cause clipping to the signal (based on the high gain transistor stage that preceeds the op-amp).


Insight appreciated. Thanks!

[Bill Mountain]

What is the purpose of D8 at the input to this op-amp buffer pictured above? To me it looks like it could cause clipping to the signal (based on the high gain transistor stage that preceeds the op-amp).


Insight appreciated. Thanks!

I think it is to limit the the voltage of the signal going into the buffer to avoid hard clipping.  It's just a guess though.  I would have guessed some sort of static protection or to keep the inputs from drifting like the inputs some amp circuits but you said it's after a gain stage?

Maybe it's an old opamp that doesn't have as sophisticated of input protection that modern opamps have???

[kurtlives]

I thought voltage protection to but wasn't sure.

Yes there is a high gain transistor stage before the op-amp. The circuit in question is the modern Boss DS-1 pedal.

Op-amp is NJM3404AL

[Bill Mountain]

I thought voltage protection to but wasn't sure.

Yes there is a high gain transistor stage before the op-amp. The circuit in question is the modern Boss DS-1 pedal.

Op-amp is NJM3404AL

Those things are strange to me.  They also have a 1k to the V+ from pin 7 which has always confused me so I just assumed they were quirks with that specific opamp.

I modded the hell out of my DS-1 but it always sounded the same to me.

I really thought I'd be the one to tackle the perfect DS-1 bass mods.  Oh...the arrogance of ignorance.

[kurtlives]

There is indeed a few interesting things. That 1K resistor confused me too.

[ashcat_lt]

That diodes can't clip the signal.  It's oriented so that it only conducts if the input swings more negative than the bottom of the battery.  The guitar signal at this point is swinging around Vref, and the preceeding gain stage will have constrained it to swinging within the rails.  It must be static protection or something similar.

[PRR]

What Ashcat said. The opamp input can't usefully swing much past its V- pin. The diode does nothing until the input is forced 0.5V _below_ the V- pin. It is protection against abuse or latch-up, not clipping. (It _could_ be to re-bias C5 on extreme inputs, like a vacuum tube grid, except the 100K+47K network makes this negligible; also an internal junction inside the chip does the same thing, but not as beefy.)

The 1K..... this M5223 is a terrible audio opamp. The output is biased deep class B. Small signals won't pass. NFB helps, and apparently this is good-enough for the unity-gain IC1A. But IC1B runs at significant gain, less NFB. Pulling a large DC current from the output cuts-off half of the push-pull output stage and makes the other half run Class A, clean (until it clips!). I suspect they first used a "good" opamp, then went to the '324/5223 chip because it was a dime cheaper, noticed an "off" sound, and quick-fixed it. (I had to do that once, 20 minutes before a show.)