Two Channels, One Volume, One TMB Tone Stack - Mixing Resistors.

[tubemonster]

From what I've gathered and seen using mixing resistors on the input grids of a tube is kind of norm for lack of better words I guess.
But, what I'm wondering is say you wanted to use Two Channels, One Volume, One TMB Tone Stack?

Would you put the mixing resistors right before the tone stack with the volume after the tone stack?
Would you put the mixing resistors just before the volume control, then to a stage, and then to the tone stack?
Or?

I get the general use of mixing resistors, and tone stacks, but not sure how the mixing resistors just before volume pots or tone stacks would weigh in on things.
I see the Ampeg Jet J20 uses mixing resistor just before the volume and tone control - for example.

Thoughts?

Here's a picture of what I'm trying to convey and the schematic of the ampeg jet j20.




Ampeg-J20-Schematic.pdf

Just a kind of curious thing, been reading and farting around (sorry, didn't mean to stink the forum up, lol), and on a tone stack bunny trail at the moment, and wondering what other may think and prefer on the subject.

[tubemonster]

Couple reads...

https://www.tubecad.com/april_may2001/page22.html

https://sound-au.com/articles/audio-mixing.htm

[loudthud]

Tone stacks generally want low impedance drive. I like the 2nd schematic. Hint: you can put the resistors before the caps, then you only need one cap.

[Chuck H]

The goal is an important consideration. The overall "dB" in voltage gain would be about equal either way, but...

The losses in a tone stack are considerable and a single gain stage CAN overdrive a second if not padded much, In other words, if you're building for a gainy sound/tone it's probably better to place the tone stack as late as practical and mix before the tone stack. If you're building for a cleaner amp it hardly matters. My only point here is that an opportunity for overdrive/clipping is lost if you mix and implement the tone stack in the same circuit because of huge signal losses.

[pdf64]

Just link the anodes, no need for mixing resistors, the valve’s internal anode resistance will do it.
And will provide lowish impedance output.
Many amps already use shared cathodes on channel input stages, so why not share the anodes too?

[Helmholtz]

Just link the anodes, no need for mixing resistors, the valve’s internal anode resistance will do it.

Good point (the method is shown in the tubecad article).

But would require the input tubes to have identical plate resistors and plate currents.

[uneumann]

But would require the input tubes to have identical plate resistors and plate currents.

I'm puzzled, Why? ^^^^
I see how both triodes wind up sharing the parallel plate resistors and a common plate voltage. But their currents may be different. Is that a problem?

[tubemonster]

Just link the anodes, no need for mixing resistors, the valve’s internal anode resistance will do it.
And will provide lowish impedance output.

I am not the 'mixing' pro you know, beyond the basic usage in most guitar amps.
Well, and repairing mixers, but that's just kind of tracing and fixing what's broke.
This all sounds good, easier even
But does cross-talk or anything like that come into play here with out mixing resistors or ?

Many amps already use shared cathodes on channel input stages, so why not share the anodes too?

Now this one I understand - lol, and I thought about it, but then I was like well, what if we wanted channels one and two to have say a .01 and .1 cap for different voicings type thing.

Which is what I think got me to scratching my head in the first place - lol.

[Helmholtz]

I'm puzzled, Why? ^^^^
I see how both triodes wind up sharing the parallel plate resistors and a common plate voltage. But their currents may be different. Is that a problem?

With one triode having a 100k plate resistor and the other a 200k plate resistor, both would see an external plate resistance of 67k with plates connected (and common plate voltage is determined by the sum of the different plate currents).
As gain depends on the value of the plate resistor, both tube will have lower gain than before, especially the one which had a 200k plate resistor.
Furthermore each triode will be loaded by the internal plate resistance of the other triode, even more lowering the gains.

It will work somehow, but operating points and thus performance will be vastly changed.

Of course, also with triodes originally having identical plate resistors performance will be changed - but in a symmetrical way.

[tubemonster]

The goal is an important consideration. The overall "dB" in voltage gain would be about equal either way, but...

The losses in a tone stack are considerable and a single gain stage CAN overdrive a second if not padded much, In other words, if you're building for a gainy sound/tone it's probably better to place the tone stack as late as practical and mix before the tone stack. If you're building for a cleaner amp it hardly matters. My only point here is that an opportunity for overdrive/clipping is lost if you mix and implement the tone stack in the same circuit because of huge signal losses.

My only point here is that an opportunity for overdrive/clipping is lost if you mix and implement the tone stack in the same circuit because of huge signal losses.

Very interesting.
So IF you were thinking of building an amp for jazz guitar, keyboard, or harmonica then maybe mixing and implementing the tone stack in the same circuit might be favorable in it's own way?
Maybe not the best or only way, but at least another tool/trick type thing?

[tubemonster]

New picture showing a bit different cap values, etc on the two channels.

[uneumann]

It will work somehow, but operating points and thus performance will be vastly changed.
Of course, also with triodes originally having identical plate resistors performance will be changed - but in a symmetrical way.

OK - the point I was asking about was your use of the word "require". That seemed a bit too strong. Yes, the circuits work differently with plates connected, but they do work and they do mix and there's nothing wrong with connecting them if you're OK with how they work once connected. As you also point out, even when the stages are identical, they will work differently too. In fact, I'll go one further and point out that even with coupling caps and mixing resistors - the stages will perform differently since gain is lost due to the mixing resistors. There are many options described here and none mentioned so far don't work. They are all viable options.

[Helmholtz]

OK - the point I was asking about was your use of the word "require". That seemed a bit too strong.

Feel free to suggest a better wording. I am not a native English speaker.

The challenge with mixing signals is to minimize source interaction and especially to preserve the original signal level ratios.
The latter will not be achieved with different plate resistors, that's what I meant.

Furthermore, if we consider the plate signals the actual sources to be mixed (as the 2 triodes seem to have different gains and probably voicings), connecting the plates will result in considerable (undesirable?) interaction between the plate signals - so the method misses the target.

[tubemonster]

Just a couple more reads someone may like:

http://www.valvewizard.co.uk/resistive.html

http://www.valvewizard.co.uk/localfeedback.html

[tubemonster]

@uneumann and @Helmholtz , or anyone else who keep me on the right track for that matter.

So, with out mixing resistors, and the two triode anodes linked together for mixing, the two triodes are seeing the two plate resistors as 100K || 200K = about 67K per triode, which is why it messes with the gain, etc?
We effectively have 67K plate resistors per triode?
Hopefully I have that some what correct.

So, what if the mixing resistors are right after the coupling caps that are tied to the triode plates, what do we have going on?

Here's were my lack of electronic engineering shows and as Enzo states "Education is what you're left with after you have forgotten what you have learned."
I must have forgot something - lol...

If the mixing resistors are tied directly to the tone stack it seems like a mixing resistor would be in series with the slope resistor of the TMB?
Or, does the tone stack (including the slope resistor) see a combined impedance of some sort between ra, Ra, and the mixing resistor(s)?

I guess I'm wondering if the mixing resistors are tied directly to the tone stack if you have to calculate the input impedance going into the tone stack differently than if the tone stack was being fed by a plate or cathode type thing.
And, if so, possibly have to adjust tone stack values also.

Hopefully I'm not to far off the bunny trail here...