Mesa Nomad 45 Reverb Circuit

[Jazz P Bass]

Can anyone tell me why the electrolytic capacitor (10uf / 35v) on the drive opamp output is wired backwards.
I was under the assumption that the + should be the input.
But here they have the - as the input. (:

[Helmholtz]

The opamp is supplied from +/- 16V, so the output should normally be at 0VDC. Maybe it gets a somewhat negative when overdriven.
I'd think an NP cap would be best here.

[Jon Snell]

If you notice, the Op Amp is configured as a frequency selective tuned amplifier with possitive feedback.
It has a gain of 10 at around 1200HZ and quickly rolls off above 2400kHZ and below 560HZ.
The output pin does sit at around Ov so no need for alarm.


Have you measured the DC across it and noted the polarity or it this all guesswork?

[Helmholtz]

If you notice, the Op Amp is configured as a frequency selective tuned amplifier with possitive feedback.

No positive feedback but wrong pin numbering.

[Jon Snell]

No positive feedback but wrong pin numbering.

In that case, please disregard my last comment.

[Helmholtz]

Pin 2 and pin 6 are inverting inputs. https://www.ti.com/lit/ds/symlink/sa...oogle.com%252F
With positive feedback there would be oscillation.

[x-pro]

The polarity of the capacitor in such circuits is usually made in the process of tuning on the layout. Then it is executed as a matter of course.

[Helmholtz]

When there's no significant DCV across a cap but ACV+DCV exceeds 1Vp, a non-polar cap should be used.

[g1]

The Nomads came out in '98. By then, Mesa had been building amps for over 30 years. You'd think they could get the schematic symbols for a JFET or a reverb tank right.

[Mick Bailey]

If they can get IC pin numbers wrong, should we trust the cap polarity from the schematic?

[Helmholtz]

If they can get IC pin numbers wrong, should we trust the cap polarity from the schematic?

As the opamp output should be at 0 VDC, there's no correct cap polartiy.

[g1]

I'm sure the 3 parallel inductors the cap connects to have something to do with it.

[Mick Bailey]

As the opamp output should be at 0 VDC, there's no correct cap polartiy.

In theory, yes, but often there's a small DC offset on outputs, depending on opamp type. I have a couple of amps on the bench right now populated with TL072s. They vary between 0v and +0.005v on their outputs and where electrolytics are used the + end goes to the opamp output. However, swapping any of those opamps for an NE5532 results in a pretty hefty -1.2v to -1.43v. These are taken from 3 different batches from different manufacturers. A side effect is swapping the ones in the reverb circuit had a drastic effect in reducing background noise when the reverb level was increased, so it will be worthwhile me keeping the ICs in those positions and installing a NP cap, or reverse the existing electrolytic polarity. I also checked JRC4558 and LM1458 in the same circuit - similar offsets as the TL072. I think the circuit would need to be designed with the NE5532 in mind to result in 0v on the outputs.

I think x-pro hit on the answer. Mesa maybe tested the circuit with a 5532 and found there was a negative offset, so oriented the cap accordingly for production. It would be interesting if anyone gets one of these in for repair that they could check the voltage.

[Helmholtz]

In theory, yes, but often there's a small DC offset on outputs,....

It seems that at least some ecaps can stand a reverse voltage of 1.5V (https://www.vishay.com/docs/25001/al...oederstein.pdf), though I assume that extended exposition to reverse voltage will increase leakage over time, because it prevents self-reforming.

So if the reverb signal+offset exceeds a peak reverse voltage of 1.5V, I'd use a NP cap.

[Mick Bailey]

I think the offset arises due to the 5532 being a bipolar device and can have a differential across the inputs due to bias current - not an issue with FET or CMOS opamps due to their very high input impedance. In the Nomad circuit the offset could be nulled by a compensating resistor in the inverting input. The problem is that different opamps have very slight differences that are amplified by the gain of the opamp, so there would still be variation either side of 0v. I think the designer deliberately decided to allow a negative offset to remove any unpredictability, and chose the cap orientation accordingly.