Two Separate Input Channels into same tone stack?

[shaun]

I'm trying to design an amp that has 2 input channels: one preamp uses a 6U8a (which sort of comes with its own triode CF built in), the other uses a 12AX7. Ostensibly, these would both drive a typical Fender style TB tone stack if used separately, but I was hoping to run them both into the same tone stack. I just don't see - and can't find - any examples where this is done.

Often when there are 2 input channels, they have dedicated tone stacks, but I don't see why this has to be the case. I'm including a rudimentary schematic - you can see where my understanding evaporates....right before the tone stack.

Thanks for any help.

[pdf64]

The CF has a lowish output impedance. From the perspective of the other channel, imagine replacing the CF with a 1k resistor (to 0V common), and what that would do to its signal level.

[shaun]

The CF has a lowish output impedance. From the perspective of the other channel, imagine replacing the CF with a 1k resistor (to 0V common), and what that would do to its signal level.

I suspect you're getting right to the heart of the matter. If the CF has a lower output impedance than the 12AX7, there will be an imbalance between the 2 preamp stages as they go into the TS. I guessed I would need resistors between the .022s and the TS to create separation of preamp channels. Not sure how to compute their values though.

[pdf64]

How about 2 tone stacks, by using dual gang tone stack pots?
Then mix after them to feed the volume control.

[PRR]

Are you playing into both inputs at the same time??

> ...run them both into the same tone stack. I just don't see - and can't find - any examples....

Your tavern sound-dude has 16 inputs and possibly a Master Tone EQ to affect the entire mix. So you need a mixer.

The sound dude thinks the cowbell is weak and that guitar player   is way too loud. So there is a gain-trim for every input. Small percussion on 9, Fender on 1-1/3rd. (Where is our cowbell icon??)

With a pentode on one side and a triode on the other, I'd expect a 2:1 to 5:1 difference between inputs. So unless this *happens* work like your accompaniment/SOLO level difference, or you like working that volume knob at every change, I think you want a mix-stage with individual level trims ("Vol1 & Vol2 knobs").

[shaun]

How about 2 tone stacks, by using dual gang tone stack pots?
Then mix after them to feed the volume control.

I've seen the double TS examples and considered that. It would be more expedient and simple to have one.

Dual gang idea is quite inventive, to be sure, but twice the components.

[shaun]

"I think you want a mix-stage with individual level trims ("Vol1 & Vol2 knobs")."

Thank you! Yes indeed. I'll consider that. ie. a vol pot after each .022 cap, then from the vol pots to TS.

Then, what if I moved the original vol pot further along and made it a post PIMV. Might be an interesting experiment and offer greater control.

[tubeswell]

Take a look at 5F6a tweed Bassman schematic

[shaun]

Thanks Tubeswell. I had looked at that but took another look to see what I'd missed. I now realize what a great example it is.

It seems fairly standard that when two channels are combined, two resistors are used - one from each channel - inserted into the signal path to keep the channels separate. I believes this prevents the two channels from "fighting" each other.

The 5F6-A has the channels running into a 12AX7, as opposed to a TS, but I will give it a try. Looks like a 270k resistor is a minimum starting point.

[pdf64]

It may be that one of the reasons for including a cathode follower is its low output impedance. So it can drive the tone stack or whatever ’better’.
Hence putting a mixing resistor between it and the tonestack will act counter to that.
And the tonestack will operate a bit differently.
Have a play with the ‘RIN’ setting https://www.guitarscience.net/tsc/marshall.htm#RIN=1300&R1=33k&RT=220k&RB=1M&RM=25k&RL=517k&C1=470p&C2=22n&C3=22n&RB_pot=LogB&RM_pot=Linear&RT_pot=Linear

[shaun]

Thanks pdf64. Yes, PRR did mention the impedance imbalance between the 2 channels in an earlier post. Having built preamps with CFs in the past, I find that design makes the TS very responsive, which I put down to the nice impedance match between the CF and TS.

As I understand it, the output impedance of a tube is determined by its inherent resistance to ground; ie., if we replace the tube with a resistor to ground, what would the resistor be? In the case of a 12AX7, it's 62k approx.

The CF runs in parallel to the EF86 and reduces the high impedance of the EF86 to a much lower impedance to ground (not sure what this impedance actually is, but I know it's lower ;).

If I add a 270k in the signal path after the CF, that does not seem the same as adding to the CF's impedance to ground because it will run into the TS before it runs to ground. ie. the 270k resistor is not parallel to the CF, so it doesn't necessarily change the CF's impedance. Does that make sense, or am I barking up the wrong tree here?

Ostensibly, what you're saying is that if I put a vol pot after the CF, it will shift the TS parameters every time I change the volume. Is that right, or it is another tree I'm barking up?

I fooled around with the TSC simulator, and I like the way it changes the pretty shapes:). But I have found it doesn't always translate in real-world application.

Sigh...there's just so much I don't, and I worry I won't learn everything before...well, before.

[PRR]

.... Yes, PRR did mention the impedance imbalance between the 2 channels ...

I was looking at gain. Where's impedance?

[pdf64]

The cathode resistance is lower than (typically), and in parallel with, the cathode resistor.
Aiken discusses it here http://aikenamps.com/index.php/designing-common-cathode-triode-amplifiers

[shaun]

Sorry for misquoting you, PRR. It seems I know even less than I thought, which I didn't know was possible. You clearly mentioned only gain, but I suppose I'm so focused on trying to understand impedance that I jumped to the assumption that gain and impedance were inextricably connected, and that I should therefore be thinking in terms of impedance matching rather than gain. My mistake.

[shaun]

The cathode resistance is lower than (typically), and in parallel with, the cathode resistor.
Aiken discusses it here http://aikenamps.com/index.php/designing-common-cathode-triode-amplifiers

Sorry, but I'm confused. Do you mean the cathode resistance of the CF side of the 6U8A, or of the pentode side?

To paraphrase, it seems you are saying this: The cathode resistance is lower than the cathode resistor.
Okay, I'll go with that and see if I can fill the enormous void that is my brain. And also, I'll re-read Aitkin :).

Thank you!

[pdf64]

….
Sorry, but I'm confused. Do you mean the cathode resistance of the CF side of the 6U8A, or of the pentode side?

To paraphrase, it seems you are saying this: The cathode resistance is lower than the cathode resistor.
…

Generally, if the cathode is used as a signal output, the impedance of that signal is lower than the value of the cathode resistor.
So specifically, that applies to the CF section of the 6U8A.
I’ve not looked at the characteristics of that valve, but with a 12AX7, the output resistance is less than 1k.

[shaun]

Ah. Yes, thank you for the clarification.

[shaun]

I liked PRRs suggestion of using one vol pot per channel, so here's the first draft of what I've come up with. I figured that, as the vol pots are early in the design, I would move the typical pre PI vol pot to post PI.

It should be an interesting experiment. I am open to feedback. The thing I'm most curious about is how the two vol pots will effect impedance between the stages.

Thanks.

[sluckey]

The thing I'm most curious about is how the two vol pots will effect impedance between the stages.

The two vol pots are very interactive. Turning either vol pot all the way down will also totally kill the other channel signal. You need two mixing resistors added to the wipers of the pots. 220K is a popular value.

Need to draw the PT center tap so the rectifier can work.

Why the oddball dropping resistor values in the B+ string.

[shaun]

 

Thank you, Steve. Much appreciated. It never occurred to me that the interactivity would function quite that drastically. But when you point it out, it seems obvious. That interactivity would probably be why it is not done too often. I guess that by the time I add the mixing resistors to the vol pots, I've impacted the signal to a considerable extent, maybe not for the better. I might re-think the 2 vol pots and simply go with the mixing resistors. Can the mixing resistors be different values? I suppose so, but an imbalance might occur. Hmmm. Big math. I suppose I could start out with a couple of 270k, and then tweak on the bench to find balance.

Yes to the CT on the PT.

Oddball values in the B+ string are due to what I had handy and seemed to work in the past. I usually guesstimate the approximate plate voltages I'll need in terms of ballpark figures, and I'm still ham-fisted enough that "close-enough" is all I can manage. I've been studying plate resistors lately, and how they effect gain and signal, so hopefully I can relate that information back to the power dropping resistors and bring a deeper understanding to bear sometime soon. I am usually trying for a clean tone with max headroom, but my learning curve is...slow.

[thetragichero]

I guess that by the time I add the mixing resistors to the vol pots, I've impacted the signal to a considerable extent, maybe not for the better. I might re-think the 2 vol pots and simply go with the mixing resistors.

why? many of the most coveted american and british amps used channel volume controls into mixing resistors. you may be overthinking this. for a four input amp individual channel volumes and a short patch cable allow for the most tonal options

[shaun]

Good point, tragichero. My thinking is that the fewer resistors the signal has to pass through the better. But yes, I do tend to overthink things.

I just took a look at the 5E3 schematic and realized why those controls are so interactive - I'm posting a partial schematic which offers an option to reduce the interplay. I guess if it's good enough for the 5E3....

[PRR]

....by the time I add the mixing resistors to the vol pots, I've impacted the signal to a considerable extent, maybe not for the better....

I guess that is why NObody uses the 5F6a or its many clones?

[shaun]

Well, yes, you're right of course PRR. What I don't know would fill an olympic size swimming pool.

[nandrewjackson]

PRRs most recent post would be the best, most simple way to get where you're headed.

[shaun]

PRRs most recent post would be the best, most simple way to get where you're headed.

Yes indeed! Thank you PRR for that schematic. It would mean adding another pre tube before the tone stack, which I'm not sure I'll have room for. Do I need it? Can I run straight into the TS, or is that extra triode stage acting as a buffer for the TS?

[thetragichero]

the extra triode is a cathode follower and indeed prevents the lossy tone stack from loading down the previous gain stage. that being said there are plenty of fender (and other) amps that have the tone stack directly off the plate of the previous stage. this is a design decision so neither is wrong nor right depending on what you're going for

[shaun]

Well, here's what I came up with, and I am very happy with the results, which doesn't happen too often. With both channels, I was trying to get gain switching without popping. I tried several methods and then came up with the ideas represented in the schematic. They're not perfect, but they seem better than anything else I've tried. The wiring is a bit messy in the preamp section, which always happens once I start experimenting.

There seems to be little interactivity between channels, and the of course pentode has higher gain than the Triode, but it is not much of a problem because the triode kicks in nicely when the amp is cranked to about 70%. Both channels respond well to the TS.

The amp is surprisingly clean with a lot of headroom and decent volume. I wanted more break-up from the triode section, so I suppose that instead of using parallel configuration, I could have an extra gain stage. But boy, there is something really nice and chimey about parallel stages, so I might leave it as it is. Not sure whether the Master Vol is actually needed because the amp is so clean overall, but it might come in handy when using pedals to drive the preamp stage harder.

PT is a salvaged Philco - ancient but solid. OT is a new Edcor 15w.

Cab is Baltic Birch. Knocked out the finger joints on my tiny Ryobi table saw using a cheap dido set. Finished with shellac and True oil. Bought a 24" bending brake to shape the chassis from .063 5051 sheet aluminum. Bending brake cost about $300. I initially tried a cheapo Harbor Freight 18" and found it was good for bending very light material but not much else, so I wasted money being cheap - as usual.

Jensen Mod 10" ceramic 50w - cost a mere $45 or so, and sounds very good for this design. A very nice surprise.

[sluckey]

I suggest a coupling cap on the plates of the parallel triodes in order to keep the plate B+ from messing with the cathode B+ of ch1 cathode follower and vice versa.

[shaun]

Thank you Steve - that's a very good suggestion. Strangely enough, I had some .022 caps right before the 270k resistors, and I removed them without thinking that through.