New AB763 based project

[SLW]

Hi to everybody on the forum.  I don’t have any musician friends that care about building things so I have come here for wisdom, advice, and general conversation about my project.  I want to build an AB763 based amp that keeps the essence of a black face Deluxe but has some modern updates.  I have a head cab.  Everything else is just in the planning stage.  To start with I would like to talk about what makes a black face Deluxe a black face Deluxe.  I have drawn a schematic of what I think is the basic essence of an AB763.  It is the vibrato channel without the reverb and the vibrato.  I would like to hear what you guys think that basic essence is.  If you see a mistake feel free to tell me about it. 

Here is the head cab I have.

[FYL]

Drop the 3M3 and related components.

[JayB]

I wouldn't drop the 3m3 and related components. To me that is the sound unless you want a real clean jazzy sound with hardly no dynamics; metal guys like that kind of clean tone.

In your first schematic, I would add a voltage divider to simulate the mixing and intensity pot just before the PI input. I always just add a 250k master pot there to allow me to tweak it to taste, You could also add the normal channel to the reverb/vibrato side and just disconnect the vibrato and reverb. Then you got two channels to switch between with an A/B switch.

It's going to be super loud though but it will sound really sweet.

[SLW]

Thanks guys.  This is the type of discussion that I was hoping for.  I had always assumed that tube stage (V2A in my schematic) was there to make up for the loss of gain from the reverb signal path.  Perhaps my assumption was wrong.  Let me know if it is.  I have never played a real BF Deluxe.   I have played the reissues and like the Vibrato channel better even with the reverb off.  I just figured that this extra stage was the reason.  That was why I started with a stripped down version of the vibrato channel.  I am interested in talking about the “voltage divider to simulate the mixing and intensity pot” suggestion.  Looking at the schematic I can see how that pot is part of the circuit even when the vibrato is off.  Does it function as a filter or does it simply act like a volume before the PI?

[bluesbear]

I built the black face Deluxe with no reverb and just left that 3rd stage out. It all depends on what you're looking for. I wanted enough headroom to actually use the amp on stage so I raise the voltages to silver face level to the phase inverter. It's clean but very warm till cranked then violin-like. Plenty of warm harmonics even when clean. If you're looking for the typical early Deluxe breakup, do it just like the black face schematic. Both sound great; it just depends on your personal style.
Dave

[FYL]

I wouldn't drop the 3m3 and related components. To me that is the sound unless you want a real clean jazzy sound with hardly no dynamics; metal guys like that kind of clean tone.

Hmmm. I've built a few BF-like amps based on the modded schematics I've posted. The amps didn't sound too clean or anemic.

[FYL]

I had always assumed that tube stage (V2A in my schematic) was there to make up for the loss of gain from the reverb signal path.

Not only. The 3M3/220K form an attenuator throwing away 90%+ of the signal, with the 10p acting as a treble bypass. You also need V2A for the normal channel...

[SLW]

Thanks again guys.  My style of playing is more toward the clean side of things but not too clean.  A little twangy on occasion.   Overdriven sounds in the classic rock style are also good.  The older I get, the bluesier, twangy, cleaner I like to play.  I hope that helps describe what I am after with this amp.  Good cleans and good overdrive.  My favorite guitar is my G&L Legacy.

I just moved and left the few guitar buddies I had behind so this amp will get played in my house most of the time.  I have been known to play in church and in jams on occasion.  I am just a 43 year old guitar hack that like to have a good time making music.  I am pretty sure that I want to add a MV to the circuit since the amp will be in my house most of the time.  I do like modulation pedals.   I want to add an effects loop.   So the MV and the effects loop are a couple of the mods that I spoke about earlier.  I have played amps with a switchable extra gain stage that I thought was nice.  If I have half a tube left over I might go that route.  A stomp box can come close enough for me.  The AB763 seemed like a good place to start.

[SLW]

I had always assumed that tube stage (V2A in my schematic) was there to make up for the loss of gain from the reverb signal path.

Not only. The 3M3/220K form an attenuator throwing away 90%+ of the signal, with the 10p acting as a treble bypass. You also need V2A for the normal channel...

I see how 3m3/220K would dump most of the signal.  Thanks for showing me that.  So the first two stages do most of the "tone shaping".  The signal would then be attenuated by the 3M3/220K divider and then boosted again by the next stage before heading to the PI.  I didn't realize the stage (V2A in my schematic) played a part in the normal channel.  I thought the normal channel went around it and into the PI.

[Fresh_Start]

FYL has built the Normal channel by itself, so that's a known-good solution.

I've only built one AB763 derivative, but IMHO that third gain stage before the phase inverter is part of the tone/sound of the Vibrato channel.

Seeing an unused triode in a design sort of bugs me though.  You could do a 1-triode bias vary tremolo like the Princeton Reverb.  Add a MOSFET as a DC coupled cathode follower if you want (that's something I need to try soon).  

Alternatively, maybe there's a way you could use the extra triode in the first stage.  Switchable as a parallel triode?  One input for each input jack with different bias points and/or cathode bypass caps?  I don't know how to isolate the two triodes for that approach off the top of my head but there must be a way to do it relatively transparently.

Just a couple of random thoughts.  

Cheers,

Chip

[JayB]

I wouldn't drop the 3m3 and related components. To me that is the sound unless you want a real clean jazzy sound with hardly no dynamics; metal guys like that kind of clean tone.

Hmmm. I've built a few BF-like amps based on the modded schematics I've posted. The amps didn't sound too clean or anemic.

Well, if your into jazz, I guess it's not to clean. 

I use the same cleans as you for the metal cleans thing. It's definitely nice. It's just not the typical blues thing you hear ever day. He likes that vibrato/reverb side and that third stage sets the tone for that channel. It compresses a little too. Just a roll back of the guitar volume and your back to a nice clean tone. I like a single coil strat into the fender. That is the tone. A les paul doesn't do it for me unless it's roaring marshall.

[bluesbear]

Generalizing "this tone for blues, that tone for jazz, etc" is a big mistake. The other guitar player in my band is about the best I've ever had the chance to play with. He uses a clean, very bright (humbucker bright, not tele) clean sound for blues. I do the blues fairly clean and warm; call it "jazz with balls". I know a good player that uses a Les Paul on the bridge pickup through a Twin Reverb. There are no rules to this, just personal preference. Decide the direction you want to go and work from there.
BTW, since this is mostly a home amp, standard blackface Deluxe Reverb (if you use reverb; I don't) should be fine for you. Keep in mind, if you don't use reverb or tremolo, you'll only need 2 preamp tubes, one for the 2 preamp stages and one for the phase inverter. Very simple!
Dave

[SLW]

I do think that you should set up you guitar and amp the way you like and have fun with it.  I like the vibrato channel so that is what I will go with. 

There are no rules to this, just personal preference. Decide the direction you want to go and work from there.

Thanks for saying that Dave.  That helped firm up the direction I want to go.

As far as reverb goes, I am still on the fence.  I don't have an amp with reverb at the moment.  I do like it and have been using a pedal.  I must have missed the reverb or I would not have bought the pedal.  So I guess I just answered my question.  I will include reverb.  I will update my schematic and repost it. 

Seeing an unused triode in a design sort of bugs me though.

Don't worrry I will not waste it.

I do love modulation pedals.  Chorus, delay, tremolo, phaser, you name it I like it.  I want to add an effects loop because modulation sounds better to me after the preamp.  I have seen effects loop with tubes and without.  I want to hear some opinions about that.

[drew]

As far as reverb goes, I am still on the fence.  I don't have an amp with reverb at the moment.  I do like it and have been using a pedal.  I must have missed the reverb or I would not have bought the pedal.  So I guess I just answered my question.  I will include reverb.  I will update my schematic and repost it. 

Layout for a single-channel AB763 with reverb, no trem:  http://www.el34world.com/Forum/index.php?topic=6781.0

[SLW]

I am updating the schematic and was wondering why are both halfs of the 12AT7 being used?  I have seen designs that only use half of a 12AX7.  That seems more simple.  I am sure it was done for a reason but I don't know what it is.

[sluckey]

More current drive to the 8ohm reverb tank. The 12AT7 reverb driver is just a low power single ended power amp.

[tubesornothing]

If you don't want super surf, 1/2 ax7 will work.  Tubenit has a lot of one tube reverb in the schematics section.

[HotBluePlates]

If you don't want super surf, 1/2 ax7 will work.  Tubenit has a lot of one tube reverb in the schematics section.

Tubenit may be the King of the 1-triode Reverb circuit!

I think he and I talked about it before he tried using 1/2 a 12AX7 for the first time. I wasn't sure it would work, but I also didn't know that it would-not-work. He gave it a go, and it turned out beautifully!

I think we talked about the fact that we tended to play with the reverb set around 2-3 and that the typical onboard Fender reverb was just way too much above 6 or so. If you don't need very long dwell time or very strong cavernous reverb, the 1-tube circuit seems to get the job done. Tubenit has several amps that prove the idea.

[SLW]

Thanks for the tip guys.  I will check out those 1 tube reverbs from Tubenit.  I do think Dick Dale is cool.  :glasses1:

Any personal experience with effects loops out there?

[JayB]

Thanks for the tip guys.  I will check out those 1 tube reverbs from Tubenit.  I do think Dick Dale is cool.  :glasses1:

Any personal experience with effects loops out there?

Try this one. Has done me well so far. It's just a serial loop, I never like a parallel one yet. Just a few parts and you could turn it into a parallel loop. Transparent so you could switch it in and out and hardly hear a difference. You might want to add a cathode bypass cap on the return to keep the volume level the same between in and out.

[SLW]

I want to publicly thank Tubenit for answering questions about the 1 tube reverb.  Many thanks.

Thanks JayB for the effects loop.  I have been studying them and have noticed a lot of similarities.  I will work something out and update my schematic.
Here is where the circuit design stands now.   In summary it is the vibrato channel of the AB763 circuit without the vibrato.  A one tube reverb has been substituted.    Additions will be an effects loop and some form of master volume.  I am not sure what way to go on this.  I have tried amps with pre- and post PI masters.  I have not tried anything that uses VVR but it does look interesting.  I am interested in hearing opinions about your experiences with MV.

[sluckey]

Noticed a couple glitches on your PA drawing. You may also want to include a shunt 47K resistor between your preamp and PI to simulate the missing INTENSITY pot. Otherwise, the gain will be noticably higher than an AB763.

[JayB]

Noticed a couple glitches on your PA drawing. You may also want to include a shunt 47K resistor between your preamp and PI to simulate the missing INTENSITY pot. Otherwise, the gain will be noticably higher than an AB763.

Way higher unless over driving the power end when it's only one is your thing. It'll be like driving a Toyota with the gas pedal stuck to the floor. 

[SLW]

It was an omission on my part.  Thanks for pointing it out.  I knew you guys would steer me straight.

[SLW]

I finally got back to drawing the schematics.  I like the JSchem program.  Kudos to the creator of it.  I added the missing parts that simulate the missing intensity pot.  They are R38, R39 and C24.  I also corrected the glitches pointed out by Sluckey.  I see this as staying pretty faithful to the Vibrato channel of BF deluxe. Even with the modified reverb.  Next up will be incorporating an active effects loop and a switchable extra gain stage to use that half of a tube that I have left over. 

Has anybody out there built a similar circuit with a MV?  It could be pre or post PI.  I am interested in hearing opinions about master volume in general.  It is the last part of the puzzle and the one that I have the least experience with. 

[sluckey]

Here's my mv... http://home.comcast.net/~seluckey/tdr/tdr.pdf

I would change R11 to a 1.5K. 820 ohms is usually seen when two 12AX7 triodes share the same cathode resistor.

[Fresh_Start]

Do you intend to bias the V1b and V2a triodes hot with the 820 ohm cathode resistors?  In the AB763 circuit, those resistors are shared by two triodes.  To get equivalent bias for a single triode, a 1.5k cathode resistor would be appropriate.

I would add a Mid control instead of the 6.8K fixed resistor, especially if you use a 25K or 50K audio taper pot.  It gives you a partial "Raw" control and a lot more flexibility IMHO.

Other thoughts:  a 10uf/15k/10uf pi filter will do a much better job at reducing ripple in the bias supply (in place of C18).  Replacing all of the power rail caps with 22uf would cost next to nothing and would improve filtering.  Being prepared to experiment with the tone stack won't hurt either - .047uf and .022uf for the bass & mid caps might sound better for you than stock.  You won't know until you try it.

Also, I'm guessing that your reverb is going to be anemic with one 12AX7 triode and the stock Blackface 3.3meg mix resistor.  Could be wrong, so build it as shown.  Just have 2.2meg, 1.1meg, and 500k resistors on hand for experiments.

More personal opinion:  Back to the power rail, having node "D" supply the reverb driver, reverb recovery plus 3 preamp stages is asking for noise.  Also, higher plate voltage for the reverb driver should help with headroom there.  Personally, I would add a new node on the power rail between B (screens) and C (phase inverter) just for the two reverb triodes and the reverb transformer.  With a 22uf filter cap for that node, a 1K or 1.5K 3-watt resistor is all you need between the new node and node B for the power tube screens.  It won't drop the overall preamp voltages much but will provide much better isolation between the reverb circuit and the rest of the preamp.  Ground with the preamp if possible.  I did this on a Super Reverb build and it worked out great.

The simplest master volume would be to put a pot (250k 500k ?) "above" the 47K R38, but that's a pre-PI master volume.  Again, my personal opinion is that a good attenuator from Weber is better than most master volumes.  However, go to the ARCHIVES here and do a search for "master volume" or "PPIMV".  Do the same on 18watt.com and/or AX84.  There are LOTS of alternatives and even more opinions.  

OTOH variable voltage regulation is a very cool concept which appears to produce the desired effects of a master volume with the tone being affected as much.

Sorry for the core dump.  Hopefully at least some of it will be useful.

Cheers,

Chip

[SLW]

Thanks Chip, that is a lot to chew on.  I will research some more of the 1 tube reverb circuits out there and compare that mixing resistor.  I do think running the reverb tube and transformer at a higer voltage is a good idea.  I had some plans to modify the standard tonestack.  This will be the most complex circuit I have ever built.  I intend to take my time ( and lots of advice).  In the end I think I will have a modern amp based on a classic amp.  I am grateful for your guys input.  I think the MV / VVR decision will take the most time.

Thanks again.

[SLW]

I haven’t gone away.  I have spent time studying several “1 tube” reverb circuits.  One of them I have actually heard.  I stayed with the circuit from Tubenit and used his values but I am not stuck on any of those.  The voltage that the reverb drive stage is run at is now the same “B” as the PA tubes.  This seems the most common.  The reverb recovery stage is now run at the same voltage as the preamp tubes.  This seems the most common also.  I have a few items about the circuit that I would like to know more about so here are my questions.
1.   C25 is not always present.  What function does it serve?
2.   The value of R36 ranges from 47k to this 750K.  Any advice about this value?
3.   C26 changes value from circuit to circuit.  I am assuming this is just a person preference.
4.   On a few circuits (including the one I heard) there is a cap connecting the anode to the cathode.  I have never seen this before.  What does this do?
5.   R42 is not there in most of the samples.  Why is it here?
In my searching I read a few things about using pentode/ triode tubes.  The tube was 6bm8.  I will look into this a little more.  Perhaps it is a good alternative.    Anybody have experience using this tube?

[tubenit]

C25 is needed. Most use 500p. I prefer .001 for more mids.

R42 was somewhat an experiment since I leave the reverb on all the time anyway. It's not needed. Just increased the verb. You can use a 2.2M reverb pot which was originally called for in the D'Lite one tube reverb but had waaay too much reverb for me so I dropped down to 1M.

R36 dwell resistor ......... where did you see 47k? I am wondering if that was a typo?  Using a 1MA dwell pot, I typically have it adjusted to about 3-5 depending on the amp.

A cap from anode to cathode tends to be a smoothing type cap such as is used on some of the Dumblish clones like the D'Lite. I think it's used on some TrainWreck clones also and some other highgain amps like BadCat (but am not sure?).

With respect, Tubenit

[mister ed]

On the single channel AB763 builds I've seen , the 47K is connected before the .001 , not the .1 cap. Is there a difference ?

Noticed a couple glitches on your PA drawing. You may also want to include a shunt 47K resistor between your preamp and PI to simulate the missing INTENSITY pot. Otherwise, the gain will be noticably higher than an AB763.

[flora]

I would like to see a reply to " mister ed ".... I was thinking the same thing.

[sluckey]

I would like to see a reply to " mister ed ".... I was thinking the same thing.

Follow this schematic (page one) from V4-6 thru the 0.1µF cap. You'll be at the junction of a 220K mixing resistor and the 47K resistor that replaces the 50K intensity pot. You can also see it on the original Hoffman layout on page three. Compare to any AB763 schematic for verification.

http://home.comcast.net/~seluckey/Hoffman_AB763.pdf

[SLW]

Back again.  I took some time to think about reverb and how I play.  I realized that I use all the settings on my reverb pedal, not just the spring.  Since I am going to have the effects loop, and some form of reverb effect in the loop, it just seems better to drop the onboard reverb.  I added a switchable gain stage to the pre-amp.  I put the gain stage together from examples I have seen.  All of the part values are up for adjustment in the extra stage.  I added the suggested modifications to the bias circuit.  I just used Mr. Hoffman's bias circuit.  I guess the effects loop will be next.  I am leaning toward a tube based one "just because" but I have been checking out solid state loops as well.  I still have to figure out what to do about a MV.  I have been looking around and reading up on MV's.  I thought of a name for it.  "763" seems like a proper name.

[Fresh_Start]

What is the purpose of the two 220K resistors at the start of the boost stage (R15 & R16)?  If I'm not mistaken, the first one going to ground will reduce the gain from your very first gain stage V1a.  Also, is the "gain" pot wired up differently from a normal volume control for a reason?  R17 may be redundant since the gain pot serves as a grid return.

Making C5 & C16 switchable "bright" caps might be a good idea depending on your taste.  As you say, the entire tone stack for the gain stage is up for grabs.  Personally, I'd start out more Marshally in terms of component values but YMMV.  You also could play with the plate resistor, cathode resistor & cathode bypass cap there. 

You may want a small coupling cap between the plate of V1a and Relay 1A to keep DC off the switch.  I'm no expert on switching though...

After reviewing this thread, I've got a couple of pratical questions.  Do you already have a chassis for this amp are you going to make one yourself?  I'm asking because of the cab you showed us in the first post.  I was unable to find anyone reliable who makes slant-front chassis for BF Fender style cabs.  If you're using an existing chassis, that might affect your choice of controls somewhat.

More importantly, how are you going to approach the layout?  Sure, you can start with Hoffman's AB763 layout (very good choice IMNSHO) but you've got a lot of changes to work on.  Plus, this clearly is going to be a prototype with a lot of tweaking after you get it up and running.  I'm asking because I find customizing a layout is more work than screwing around with schematics.  Plus, when I built a customized AB763 derivative with a lot fewer changes than you're contemplating, I had some problems making things fit during the corrections/tweaking phase of the build.  In other words, follow PRR's advice quoted below:  "Plan on something being wrong."

Not trying to be a party pooper - this is a fun discussion from my perspective. 

If you've already done your own layout from scratch from a custom schematic, please ignore my comments on that topic and rock on!

Cheers,

Chip

[SLW]

Not trying to be a party pooper - this is a fun discussion from my perspective. 

No worries Chip.  I think this is a fun discussion too.  While music and electronics are both serious hobbies for me, my ego doesn't rest on them.  I got thick skin.  Go ahead and call me a crazy redneck.  One of the reasons I came to this forum is to learn some things and get input from folks more experienced than I.  If I didn't want you guys to give me input and advice I would not have posted.

What is the purpose of the two 220K resistors at the start of the boost stage (R15 & R16)?  If I'm not mistaken, the first one going to ground will reduce the gain from your very first gain stage V1a.  Also, is the "gain" pot wired up differently from a normal volume control for a reason?  R17 may be redundant since the gain pot serves as a grid return.

To be honest, I took the gain stage from a Mesa Lonestar.  It is an amp with a switchable gain stage that I like and I had the schem to study.   I did change a couple of values.  R16 should have been 1M.  I made a mistake labeling.  I don't know how to answer your questions why there are two resistors or why the pot is wired up differently.  I wondered about these things also.  I finally decided that they must of had a reason so I stuck with it.  It is not written in stone though.  Lets talk about it some more. 

I Thought about a blocking cap before the relay.  It seems to make more sense to me to switch the signal and leave the DC behind.  I am willing to talk about that. 

Concerning a chassis.  I do not have one.  I plan on making one.  While I do have machine shop experience and I have access to the proper tools to bend up a chassis, more importantly, I have access to the “old timer” in the shop who likes helping me with my crazy projects.  We have already talked about my sloped front box.

Concerning the layout.  I plan on starting with a known good one and working it out.  If I have to build it twice I don’t mind.  If it takes me two years to build it that will be two years of fun.  I hope you guys will give some more advice.  I did the layout for my first amp.  It was a simple Tweed Princeton with just three option switches.  This one will be a lot more complicated.  I like my projects to be difficult.  They push me to learn new things and to accomplish something difficult.  I get a lot of satisfaction from the process.  Sorry for all of the philosophy in this post but it is fun too.

[PRR]

 > two 220K resistors at the start of the boost stage (R15 & R16)?  If I'm not mistaken, the first one going to ground will reduce the gain from your very first gain stage V1a.
 
Yes. The output of the first stage could be a couple dozen Volts, WAY more than is needed to shake the overdrive stage. Overdrive TOO much, it just gets messy, harsh, it fights you.

R16 cuts the signal a little, but more important gives a more constant load for C11, a modest low-cut. R15 is semi-necessary because the Drive control is backward-potentiometer which needs a very non-zero source impedance. R17 likewise refines the Drive control action. There's other ways to skin this cat, but you could save at most one 12-cent resistor, and it might not work the same. I'd "have to minimize" it just as a puzzle, but if it is stolen from a known-good plan, do it.

I too am not in love with the relay taking raw plate voltage. At the least it will want ~~22Meg resistors across the contacts so the not-enabled caps (C11 or C2 C3 C4) will hover at plate DC voltage and won't POP! so bad when switched. Alternatively a cap before the relay (and a ~~10Meg drain) but it has to be much bigger than the subsequent caps or it will shave bass. 0.1u at least, maybe more, a fat hunk.

My other thought is: here's a bells + whistles amp (well, trem and reverb) which had to be gussied-up with extra gain to cover losses in the 3.3Meg reverb mix network and the 50K term-roach node.... and there's no trem or reverb! Seems to me that C7 could go right to the PI. Maybe the extra staging adds "flavor" even when not needed for the bells and whistles. Or maybe its just more parts.

OTOH, with all that other stuff, making R13 a pot that goes to zero (shorting-out V2A C10) is a natural place to strangle the preamp (MV), one that apparently worked in tremolo context when the roach shorted the full-up 50K trem-depth pot.

[Fresh_Start]

Personally, I like the idea of C7 going straight to the PI then using the "extra" triode as a DC coupled cathode follower on the "gain" channel.  Again, I'm headed toward a Fender/Marshall choice and that may not be what you want.  Also, your "clean" channel would sound more like the "Normal" channel of a BF Fender than the "Vibrato" channel.

Back to switching:  is there a better way do this?  Anything need to be done at Relay 1B to avoid a big POP?  I don't know squat about switching other than what I've read here and in O'Connor's TUT 1, but there seem to be 2 goals:  avoid DC on the switch and try to keep switches out of the direct signal path whenever possible.

Hope this helps,

Chip

[SLW]

Just a couple of corrections to my last schematic.  Chip, lets have a look at your switching (and other) ideas.

[Fresh_Start]

SLW and I have emailed back & forth.  Here is an alternative approach to the switching.  It both keeps the signal from going through the switch(es) and keeps DC away from them.  SLW's topography rang a bell - the basic idea is close to KOC's "London Power Standard".  That's where this switching arrangement came from.

Unfortunately, I realized that KOC creates a 50% voltage divider after either of the Volume pots.  See two 475K resistors and circled grounds.  I cut the 3.3 meg resistor to 1.5 meg, but the triode right after the two channels come together would only be getting 1/2 the signal normally seen in an AB763.  Not happy after looking this over again!  

I also looked over the Lonestar schematic.  The only thing missing from SLW's schematic is a 100K grid stopper right before his V1B grid.  I changed C11 back to .005uf and R16 back to 1-meg to match the Messy Booger.  

The 475K isolation resistor before Relay 1A concerns me in terms of frequency shaping, both for the "gain channel" and in relation to the 0.47uf cap I added to keep Relay 1A from grounding DC.  I raised the cap value to .015uf in an attempt to hit the same cutoff frequency.  Is my math ok?  Should the 1 meg resistor to ground be changed to 500K so total resistance in series to ground is roughly 1 meg? (need to change cap again if so)

The Lonestar must work fine, so I'd be interested in the pros and cons of these two approaches to the switching.  KOC's is more "elegant" to my eyes, but I'm probably missing something fundamental (like the voltage divider after each volume pot ).

Also, how interactive will the volume pots be with KOC's approach?  I think that the Normal volume pot will affect the gain stage output, but the effect of the Gain volume pot on the normal volume should be minimal.

While I was modifying SLW's work to show the switching alternative, I took extensive liberties with the gain stage and its tone stack.  My idea was to get more of a Marshall-like tone although it still will be a long way off without the cathode follower to drive the tone stack (among other missing pieces).  Any comments or suggestions on this as a starting point for experiments would be most welcome!

Swapped tube triodes so that V1A is input, V1B is gain, V2B is following gain stage & V2A drives the PI.  Added node "E" to power rail so V1 is isolated from V2.

Also put fuse on "hot" side of PT primary.  Is that correct and is 2 amps an appropriate fuse value?  (AB763 uses 3 amp fuse but there are two more triodes plus the reverb transformer.)

Please let SLW know whether or not any of these suggestions has some value.  

Cheers,

Chip