Shorted Reverb Tank Causes Distortion

[Phostenix]

I have a Crate G1600XL that has the same 22 ohm resistor to ground on the Reverb Send - output, but connecting the shorted tank to it doesn't have the same effect. It makes me wonder if there is something about the preamp grounding in the GT3500H that makes it succeptible to this issue.

[Tom Phillips]

Wow.
Good scope screen photos.
Hard to do with the average camera. You must be into photography.
Tom

[Steve Conner]

Does this amp have the typical cheap-ass +/-15V power supply with zeners and dropper resistors? If so, then I suspect the shorted reverb send is causing the zeners to drop out.

Photographing a scope screen is easy, just turn the lights down and the flash off.

[Enzo]

That was my initial response too, but he reported later that lifting the input cap C55 to the driver stage stopped the symptom. That tends to support his theory that the shorted IC output was affecting the input pin.

[Phostenix]

Wow.
Good scope screen photos.
Hard to do with the average camera. You must be into photography.
Tom

It's a basic Canon Powershot S5 IS using the super macro setting. If you're gonna sell pedals on ebay, you've got to have a camera with a good macro setting. Lights are still on in the room, too.

[Phostenix]

Does this amp have the typical cheap-ass +/-15V power supply with zeners and dropper resistors? If so, then I suspect the shorted reverb send is causing the zeners to drop out.

The +16/-16 supply is as you described.

[Phostenix]

That was my initial response too, but he reported later that lifting the input cap C55 to the driver stage stopped the symptom. That tends to support his theory that the shorted IC output was affecting the input pin.

Does it make sense that the signal ends up on the ground & power rails? It still doesn't make sense to me. Especially since the 22 ohm resistor is shorted to ground by the reverb tank once the cables are connected. I'm not sure what it's function is.

[R.G.]

I think it's simpler than that.

Point F is the dry signal that gets sent to the rest of the amp. If the input of the send opamp changes from an input to a set of clipping diodes, then this becomes a diode clipper on the dry signal, right?

Shorting the output of the opamp causes ugly things to happen internally. In this case, I would guess that the short on the output is making the opamp effectively not there at all and perhaps activating some internal junctions between the + and - inputs, clipping the signal.

It's a guess, but it does match the observed description and scope pictures, as well as the pulling C55 loose at F making it stop.

[Steve Conner]

I think RG just won a cookie!

The NE5534 has back-to-back diodes between its inverting and non-inverting inputs on chip, to protect the input transistors from excessive common-mode voltage.

The NE5534 is also the original low-noise op-amp, so other low-noise op-amps may well be the same.

[Phostenix]

But I didn't see any degradation of the signal on the "F" side of R77. It seems like the op amp that was impacted by shorting the output of IC5:B was IC4:B. It makes me wonder if the reverb send signal was injected onto the ground trace which is the + input of IC4:B. The GT3500 is re-assembled & back in service at this point, but I might breadboard some of this & do some testing. Enzo ws right - replace that bad reverb tank & move on.

In this case, the reverb tank was a 99 cent ebay buy (non-working) and this Crate amp just happened to be the amp out when I tested it. So much for a quick test to see if it really was bad....