1 triode into 2 channels, possible pitfalls?

[Bergheim]

I have an amp with two 12ax7's in the preamp, and I'm looking for a way to build two preamp channels out of these.
What I'm picturing is:
Standard V1a input stage with 1k5/25uF and 2k7/.68uF cathode on a switch.
Two coupling caps, one into a fender preamp and the other into a marshall preamp.
Fender preamp: tonestack -> volume control -> V1b gain stage -> coupling cap -> 220k mixer resistor
Marshall preamp: volume control -> 470k/470p -> V2 DC CF -> tonestack -> 220k mixer resistor

As far as I can see this should be a pretty simple and straight forward build.
Will there be any problems having two coupling caps off of the same plate, aside from the extra AC load due to the Fender tonestack and volume control being in parallel with the Marshall volume control?
It's clearly not gonna be exact replicas of the Fender/Marshall preamps, so some tweaking and compensating here and there is totally fine. I really like having two different voiced channels in the same amp, and even more having the possibility to blend them, which to me is the holy grail of adjustability
I've read a lot through the forums about parallel preamp channels, blending, in and out of phase and what not, but never seen anyone hook two coupling caps to a single triode, which except from an extensive switching setup may be the most effective way to make the most out of a single triode. Anyway, it got me thinking there has to be a reason I've never seen this before

All ideas and thoughts are welcome.

[Stevem]

My main thing I need, and all should want out of a 2 channel amp is two sets of tone controls ( tone stacks) to have the needed EQ flexibility between clean and lead!

If you do not have needed gain stages to get that done them to me atleast your wasting your time, I would sooner then use a SS or better yet tube stomp box in front of the amp for my added gain ,distortion and EQ changes that I need!

[roberto]

Find a compromise for the first stage, keep the 680n always on ans switch the 22uF for the clean channel.
Decouple and reference to ground, then switch between EQ or gain-pot-ensamble.
Do not mix with 220k but decouple then again switch to a final source follower to the output.

How will it be the power amp?

[Bergheim]

My main thing I need, and all should want out of a 2 channel amp is two sets of tone controls ( tone stacks) to have the needed EQ flexibility between clean and lead!

If you do not have needed gain stages to get that done them to me atleast your wasting your time, I would sooner then use a SS or better yet tube stomp box in front of the amp for my added gain ,distortion and EQ changes that I need!

Thanks Stevem. For this amp it's really more about having a simple but versatile setup that suits as many guitars (pickups) as possible, from spanky cleans to cranked 5F6A/JTM45-territory, rather than switching between crystal clear and hi gain.
Both preamps are fairly clean on their own, but the early vs late tonestack, different component values and especially the cathode follower in the Marshall channel should make for a very wide range of sounds and characters.

Find a compromise for the first stage, keep the 680n always on ans switch the 22uF for the clean channel.
Decouple and reference to ground, then switch between EQ or gain-pot-ensamble.
Do not mix with 220k but decouple then again switch to a final source follower to the output.

How will it be the power amp?

I've found that shunting series caps almost completely removes switching noise, so I'll place a 1uF in series with a 22 uF with a switch to short out the 1uF. 1uF and 22uF in series is .69uF.
Switching between channels is the standard route, but it kinda defeats the whole idea of being able to blend the channels, and also requires a fair amount of wires and parts for the switching setup.
Power amp is a brownface LTP phase inverter and 6L6GC tubes with adjustable NFB (brownface to JTM45 levels). I'm considering putting the 820R bias and 6.8k tail on a switch to select between that and 470R/10k.

I guess I'll just build the thing and see how it works

[Bergheim]

Today I finally got the amp wired up and tested.
Notes and thoughts:
1. The Fender channel with its heavy tonestack load is hardly affected by the extra 1 Meg Marshall channel volume pot in parallel. The Marshall channel on the other hand I think will definitively have som extra AC-load coloration from the Fender channel, but I haven't scoped it so I don't know how much it is affected. It still sounds and feels like a Marshall preamp so I don't mind anyway.
2. I ditched the switched cathode setup for the input stage and went with 2k7/10uF. To me, the 2k7 sounds like it has a little more "color" and string definition (probably a little more harmonic distortion) than the 1k5. With the cathode being fully bypassed, the Marshall channel needed a smaller 1n coupling cap to keep the LF roll of point of the entire gain stage about the same as the original Super Lead bright channel circuit. A 680p would actually be even closer to the Super Lead frequency response, but I felt the tone was lean enough as is.
3. I modded the cathode follower quite a bit from the original values to increase the compression/overdrive and reduce the headroom a bit. It actually ended up with the Vox Top Boost values. Although the plan was to keep the circuits as close to stock as possible, I realized this was most important with the Fender channel. And honestly, while there clearly are tonal and dynamic differences between the channels, they weren't quite big enough to justify the hazzle of wiring two channels in parallel (a little nod to Stevem ). The Top Boost values got me halfway there, and eventually I added the 150n bypass cap which increases the gain of the stage at the same frequency as the 470k/470p treble peaker. That cap really did the trick.
4. The PI is setup with Brownface values (820 bias and 6k8 tail). I wired a 1k2 on a switch in parallel with the 820R bias resistor (1k2||820=~480R) to decide if I wanted the Brownface PI bias or the 470R found in Marshall/5F6 and BF Fender circuits. I honestly couldn't hear any differences at all, so I left the 820R in place..
5. Variable NFB level. Goes from tight JTM45 to loose Fender levels. Most people seems to wire this one on a switch instead of a pot, as the differences along the sweep are rather subtle. They do have a point