Q's about Vox VR30 Poweramp

[mth5044]

Hola.

I've been looking at the poweramp based around the LM3875 in VOX's VR30. Here is a link to the manual, schematic in question is mostly on page 7. http://www.blueguitar.org/new/schem/misc_amp/vox_vr30reverb.pdf

As I understand it, the transistor bit in the beginning is used to flip the phase to drive the tube for a tube poweramp simulation. IC1A and circuitry is used to get the phases back in sync and add them together. Signal splits into the chipamp driver as well as down in to the connector pins 8/9/10 which connect, on page 6, to a Master Power Level pot.

Then it gets weird for me.

Looking at IC2A/B, they split the phase back in to something that can be mixed in to what is coming out of the poweramp. Why inject signal back in to there instead of after the mixing stage of IC1A?

There are a lot of parts of the circuit feeding back in to the IC2A/B section of the schematic. It has a connection to the master power level control from post IC1A as well as the sleeves of the headphone and external speaker jacks. With nothing in either of those jacks, it also connects to the gain setting circuitry of the LM3875 as well as one side of the internal speaker.

So what I'm guessing is that the signal being fed back through R43 is some sort of negative feedback and the signal coming from the master power level control through R33 is sending boosted signal back through pre-mixer stage. By why? And how is this a power level control?

Thanks for any help!

[mth5044]

In an attempt to see what's going on, I sim'd the parts without the poweramp as I'm not sure how to get that in the library at the moment. Although, I suppose it's just an opamp 

(R13/14 are the Master Power Pot and R17/18 are the trim on the opamp)

V1 and V2 are set 180 degrees apart as I'm assuming that's what's coming out of the tube. Both are set at 1V from simple math on how the voltage changes. The output of U1 gives me a 2V output in phase with V1, but after C4 and R5, with the Master Power control up, I'm getting a > 90% signal reduction. U4 boosts it up a bit, but only to ~600mV or so.

The signal coming out of U2 and U3 is so minuscule, me thinks I need to add the poweramp in with the feedback coming back in to U3/U2 as the original schematic in order to get the full picture. Otherwise, it's not really doing anything.

EDIT: Nope, can't simulate the LM3875 with a simple opamp, and probably not a speaker with an 8ohm resistor.

[PRR]

Me too, and I just get dizzy around the tube.

The Power Amp, devoid of frills and switching, is a basic well-known thing. Small series resistor feeds-back information about speaker Current. The output (source) impedance is raised from "zero" to some larger value (I would suspect about 8 ohms), with little loss of output voltage/power.



You can stimulate this with an ordinary opamp by scaling 8r to say 1600r (20X) and 0.22r to 4.4r (20X), omit any other low-Z parts which may strain the opamp. Try different values at the "speaker" terminals. Vout in a short (very small signal!) and an open gives you the output impedance. For funny curves, find a simulated speaker, scale up 20X, and plot response.

[mth5044]

Thanks for taking the time to clear up the circuitry around the chip. Without all the switching jacks, it is way easier to digest. I will attempt the scaling resistance for the speaker.

I took the same approach to looking at the maze post-tube.

Phase is relative.

So we have current feedback going to the powerchip as you describe, as well as back to the gain stages feeding the tube outputs.

I think what is throwing me is the master volume setup. When turned fully up, you have maximum resistance between ground and the signal feeding the poweramp, but you are grounding out the current feedback from the speaker and the stages feeding the tube signal mixer

[PRR]

> When turned fully up

> maximum resistance between ground and the signal feeding the poweramp

More gain...

> grounding out the current feedback from the speaker

More gain.

You lose the low damping-factor on the speaker. Perhaps with everything else slamming and banging, you don't need it, maybe don't want it.

This obviously was not designed in an ivory tower with silver pen. I can picture years of tinkering a breadboard, and multiple re-thinkings.

[mth5044]

More gain.

Doh!

[slacker]

but you are grounding out the current feedback from the speaker and the stages feeding the tube signal mixer

Looks like you've got this figured out just thought I'd mention it doesn't ground the feedback from the speaker, the signal goes though a voltage divider made of R43 and R33 plus the gain pot, at max gain the bottom of R33 is grounded so you've still got some negative feedback.