Help with op-amps in Vox AC30C2X

[SunTzuBean]

Schematic for reference: https://el34world.com/charts/Schematics/Files/Vox/Vox_ac30c2.pdf

Hi I know this is not technically a stomp box, but this question is mostly about the op-amps in the circuit of this amp. I figured you all know op-amps best, so I'm asking it here.

I recently bought a used Vox AC30C2X and have been reading online about how this particular amp has its signal path op-amps configured in such a way that it results in the signal is boosted by about ~4.5x from U1B to U5B. When I try to calculate the gain on my own, it appears to be only slightly above unity.

When I look at the schematic, I see that the first op amp serves as a mixer and a signal polarity inverter for the Normal channel. This op-amp definitely confuses me the most. Since its inverting input is connected to the normal channel volume pot, it appears to be missing a path to ground unless the volume pot is turned all the way down. I don't have a clue as to what the gain of this op-amp is or how to go about calculating it, since it's in a configuration I've never seen before.

Then the signal is attenuated via voltage divider network by approximately 1/4 (~.258). Next, the FX recovery op amp boosts the return signal by about 4 (~3.83, or roughly 1/.258, boosting back to unity after reduction). After the reverb, the last op amp that affects the signal is connected to R_FX_RET, which appears to have a gain factor of about 1.07, unless my calculations are incorrect.

What would I have to change to get the gain back to unity?

[PRR]

Welcome!

Your post is munged-up with backslashes. That 'link to schematic' should be:
https://el34world.com/charts/Schematics/Files/Vox/Vox_ac30c2.pdf
(no backslash before underscore)

EDIT - I see, you also posted to REDIT which seems to need escaping.
https://www.reddit.com/r/AskElectronics/comments/1ht3hqo/can_anyone_help_me_determine_the_opamp_gain_in/
trtr6842 knows his/her stuff; one point that may have been missed is that U2B is not a classic mixer. It forces a high output impedance b/c that is a better fit to a spring tank. This is NOT in the dry path (that I see) and would not be unity gain but rather hundreds to cover spring loss. Note the 47Ohm resistor in the gain-set network.

> the first op amp serves as a mixer and a signal polarity inverter

I don't see that. Which is "first"? U1A is not mixing or inverting, just gain of +6.1. U5B (from U1A) is gain of about half of +4.9.


I would suggest re-drawing. VOX's computer has made a mess of this. Leave out the stuff not in the path you are investigating (the whole C35 leg). At least re-draw as a block diagram with gains and assumed signal levels noted.

[SunTzuBean]

Hi, thanks for the welcome!

Thanks for the info about the reverb driver. I didn't know that reverb tanks like to see a high impedance, I would've though the opposite is true since it's essentially a speaker with no cone, thus lots of current and a lower voltage does it good. But now that I think about it, when I see tube driven reverbs, they're not driven by a cathode follower  I'll have more to learn!

I fixed my original post. Yes, it is up on Reddit! I am asking around because this is really a mystery to me. An experienced amp tech also made the claim of increased gain: https://youtu.be/WU8C-70tFOc?si=Y4oCvhRWwH7qQfCT&t=617 (this is timestamped).

I also found this video from the same amp tech that hints that the problems lie in mixer of the two channels, the buffer to the FX loop, and the mixer from the reverb: https://www.facebook.com/PsionicAudio/videos/the-ac30cc2c2x-is-a-pretty-good-amp-but-whoever-designed-the-opamp-mixerbuffer-s/3110542522384943/

I apologize for the confusion as far as "first op-amp." I specifically meant U1B, which is the first op-amp in the signal chain. It's a source of confusion because I know it serves its purpose as a mixer that inverts the phase of the Normal channel so that you can jumper the inputs. But that means that it appears to no longer have a connection to ground except through the Normal Vol pot in a more traditional inverting amplifier config.

Then again, the non-inverting input has a connection to ground through R35... I have no clue what's happening! I'm not sure how to calculate the gain of this stage.

U1A is not mixing or inverting, just gain of +6.1

Can I ask how you calculated this? To me, this looks like a standard non-inverting op-amp config, thus the gain should be 1 + (R30 / R31) = ~3.83. This seems to mimic the signal loss through the voltage divider network R41, R38, and R36: 1 / (R36 / (R36 + R38 + R41)) = ~3.87

U5B (from U1A) is gain of about half of +4.9.

Is it half because it's summing two inputs? Sorry, I'm still learning

[merlinb]

recently bought a used Vox AC30C2X and have been reading online about how this particular amp has its signal path op-amps configured in such a way that it results in the signal is boosted by about ~4.5x from U1B to U5B.

Are you saying you actually have a problem with the FX loop? Or have you just read something online?

[SunTzuBean]

The problem with the FX loop is that it is supposedly boosting the signal level by 4.5x, driving the phase inverter and power tubes harder than it should. The amp is fairly close to vintage Vox AC30 Normal and Top Boost channels, and I purchased the amp knowing this with the intention to change it. I think it would be a fun project! But unfortunately, my knowledge falls short when it comes to op-amps, especially U1B.

I've forgotten to report that changing the Normal channel volume while using the Top Boost channel does not appear to significantly change the gain of the Top Boost channel.

[merlinb]

The problem with the FX loop is that it is supposedly boosting the signal level by 4.5x, driving the phase inverter and power tubes harder than it should.

I am not convinced that's correct, but you can always reduce the gain before the phase inverter by reducing R80.

[ElectricDruid]

U1B is set up as a differential amp. That is to say, *all* op-amps are really differential amps, but mostly we only use one input or the other. This circuit is using both inputs.
The Top Boost channel signal comes in via C25, and then there's the two resistors you'd expect to see in the diff amp, R34/35. That's the non-inverting input.
The Normal channel has DC blocked by C7, and then there's the volume control VR1 and associated bits, before we get to the two resistors you'd expect to see on the other half of a diff amp, R37 and R40.
Talking about whether the op-amp is "mixing" or "inverting" requires knowing what phase we're expecting. Whatever happens, the op-amp produces the difference of the two inputs (A-B). However, if one of those inputs was inverted earlier in the circuit  then it's equivalent to mixing, since the previous inversion gets cancelled.
The other thing to say about diff amps is that the gain for the two inputs can be different - quite often is, in fact. So the Normal channel and the Top Boost channel aren't treated the same way. The best page I can find about it at the moment is this one:

https://www.electronics-tutorials.ws/opamp/opamp_5.html



As you can see, the equation for the output voltage is a bit of a mouthful. It's pretty clear that the gain for the inverting input is simply what we'd expect for the basic inverting op-amp ( the -V1(R3/R1) part ), but the gain for the non-inverting input is a lot more complicated (the rest from +V2 to the end!)

The AC30C2X schematic shows:
R1 = 470K
R2 = 330K
R3 = 820K
R4 = 120K

That should be enough for you to work out the gains for the two inputs.

[SunTzuBean]

U1B is set up as a differential amp

That makes so so much sense! Thanks so much for providing this info. I hadn't learned about this configuration of differential amps yet.

I did know that an op-amp is a differential amp, but I'd only seen it used in a Schmitt trigger.

Using the equation from that site, the stock inverting input amplifies by a factor of -1.74 and the non-inverting input amplifies by a factor of 0.73.

I played around with the values in Desmos and changing R40 (R3 in your post) to 470k and changing R35 (R4 in your post) to 330k causes each side to output unity gain:


I was worried at first that if I made these changes, the signals would almost be cancelled out if the inputs were jumpered, until I remembered that the incoming signals are out of phase with each other. Thus, they will actually be mixed together; negative times a negative is a positive and all that (right??).

you can always reduce the gain before the phase inverter by reducing R80

This amp looks like it IS configured in a mixer configuration with a non-inverting input. It looks like all I'd have to do to set this op-amp back to unity gain is change R80 to 1k since the summing input resistors are the same value.

The other op-amps in the chain only add lost gain from the FX loop attenuation and are used to drive a (wet) reverb signal, so their gain is not as important.

Unless I'm missing something, I believe that answers my questions as far as what to change on the amp to ensure the signal level entering R_FX_SND equals that at R_FX_RET. Thanks everyone!

If I wanted to make one last change... I'd want to increase the gain from the reverb recovery stage, U5A. There may be something amiss, but the reverb on the amp feels a bit weak. This could just be normal though, I'll do some research to find out!

[merlinb]

It's tricky because the gain of one signal path changes depending on the control settings of the other signal paths. If we break it down to the bare minimum, assuming other controls are set to zero, then we have this, which does have a little more than unity gain. But if any of the other pots are turned up a bit (which they probably will be) then the gain gets closer to unity. Admittedly Vox has made some over-complicated design choices throughout.

[BJF]

Hi


Oh this is considerably different from the AC15C in which there are OP amp buffer to buffer any point that would have a treble loss;)

Here though Normal channel has been made a tad brighter which it really needs and the Bright channel got a little ceiling via the 10pF cap to ground. Actually pretty nifty to sum both channels in phase with an OP and give it a bit of gain which it should hopefully handle considering the OP is powered by 54V´s
One could guess that the output swing is shy of supply votlage with a Volt or so per side so output at U1B can likely deliver over 50V´s pk-pk
And that's also the max output of U5B : this should really allow pushing power amp to max before any of the OP amps distort

Yes an effects loop in a tube amp needs to be a bottle neck. The Zener clamper at 27,6V indicates highest transient allowed at return !

Yes that's cool with brute force summing buffers to drive the reverb input.

Actually this design makes clever use of speciality OP amps

Hm rectifying the filaments indicates shortened tube life.

Bjorn Juhl
BJF Electronics