New build - Dual Channel Dual Mode 2204

[Snicksound]

This will be for a friend/bandmate. He's currently using a sort of JCM800 2204 that I built for him 5 years ago, except it runs on a pair of cathode biased 6V6, and has a switch to disengage NFB.
Turns out... he always plays with the NFB off. With a slightly tweaked tonestack and an OD pedal in front, into a Mesa cab, it sounds quite similar to a Dual Rectifier, except much more open and controlled.

It's time for me to build him something better.

This is what is missing or needs to be changed:
- A "clean-ish" channel (he currently does that using his guitar's volume, it's never truly clean, and being able to access that tone instantly would be a good thing)
- Being able to not need an external overdrive pedal to get it where it needs to be (this is a controversial opinion amongst Marshall enthusiasts, but practicality is always a good thing, and pedals are always noisier IMHO)
- Without NFB, the high end can be quite intense, even with Treble down it can still be ice-picky. A Mesa Boogie Dual Rectifier style Presence control would fix that

I also want to give him a stouter more powerful fixed bias power section. The type of sound he's looking for requires him to keep the Master Volume fairly low, it doesn't work if he has a gig with no PA. Current amp is maybe 15-17W at most.

I COULD just build him a Dual Rectifier kind of deal, but he also does enjoy the old school bold crunch of the 2204 circuit when NFB is engaged. He plays in multiple bands.
Also needs to still uses only 3 preamp tubes due to size constraints.

So here's what I'm going for:
- Stay with 2204 circuit as the basis of the amp, he loves his current amp, he just wants more versatility and control.
- Add a "Plexi clean" mode that bypasses the cold clipper stage and has its dedicated gain and volume (I've been experimenting with my personal 2204, it's a great sound even if the input stage remains 2.7k/0.68uF)
- Add a high voltage FET Boost on the front end, tweaked to focus the mid range and cut some low-end, so he can get it to roar as if he had his MXR Custom Bass OD in front without needing it
- Will use a dual footswitch: Boost and Channel (will be possible to boost the Plexi side too, because why not)
- Tone stack will be shared, and minimal tone shaping switches (K.I.S.S. kind of amp, he likes it simple)
- The switch that disengages NFB will also automatically engage a Dual Rectifier Presence circuit (Presence pot will be dual-ganged, the Mesa side will be lifted when NFB is engaged, and when the NFB is disengaged the Marshall side simply won't have any effect anymore)

I'm experimenting with deactivating NFB by simply AC bypassing the NFB tail resistor (if you put a 10uF or so cap instead of a 0.1uF on the Presence pot, it turns into a NFB control!). Though right now I'm finding that it tends to pop once in a while (usually the first time you use it) which I can't quite figure out. The big cap has a path to ground when not fully bypassing, and I'm switching the ground side. More on that after I play with it more.

So this is the plan so far:

(Erroneous schematic removed)

Fairly simple, but immensely versatile! A total of 8 different sounds (2 channels, boosted vs unboosted, NFB vs no NFB)

Because these are the transformers I have and because why not, it will run a pair of KT66 in the power amp, though I will make sure it can run 6L6GTs too. The transformer are a Tweed Bassman set (Hammond 290DX and 1760K).

This build may take a few months, I'm still in the prototyping stage (which I'm doing by sampling various iterations of the circuit in my personal 2204 clone) and getting parts is challenging these days.

EDIT: Updated schematic:

- Power supply added
- Fixed some values
- Dual mode Presence control using single gang pot instead of dual gang

[Snicksound]

Oooops, already spotted an error: C12 should be 3.3nF and not 33nF

[Colossal]

R15 and R16 reversed?

[Snicksound]

R15 and R16 reversed?

Oh, good catch! Thanks

Also added the power supply:
(Updated schem in next post)

[Snicksound]

Fixed an issue in the bias supply (cap polarity... easy mistake to be avoided!)

Also realized that dual ganged 25kB pots are pretty much impossible to find, but then realized I can accomplish the dual mode Presence with a single gang. Simply ground the wiper, connect the "Marshall" side to the top lug and the "Mesa" side to the bottom. Simply disconnect the Mesa side in Marshall mode. Since both work in reverse (max brightness = max resistance for Mesa but min resistance for Marshall), that should work!

[Snicksound]

Managed to fit everything on a 11.5" long board.
(NB: the 10uF caps in the DC elevation and NFB bypass circuits are reversed in this image)



Rough chassis layout to make sure everything fits:



Chassis all drilled out.

[xtian]

Very, very cool. You've addressed all the things I wanted...immediately after finishing my recent 2204! My honeymoon with the 2204 just keeps going on and on...

[Snicksound]

Very, very cool. You've addressed all the things I wanted...immediately after finishing my recent 2204! My honeymoon with the 2204 just keeps going on and on...

Thanks

Fun fact, after I'm done building this one and delivering it to my friend, my next project is rebuilding my personal 2204 clone... to the original specs.

The thing is, it's such an amazing amp for dynamic playing.

BUT, the difference is I also have a Mesa Badlander on the way, so I'll use that when I need channel switching and lots of onboard gain.
And I can't wait to compare the real 2204 to this frankenamp.

[xtian]

The master on this amp is awesome. Have no fear of the 50w power section.

[Snicksound]

The master on this amp is awesome. Have no fear of the 50w power section.

That's the plan! I want this amp to be easy to control for living room noodling, while having enough power on tap to do unmic'd gigs with a solid low end.

Too many high gain amps are porting over their master volume straight from the JCM800 architecture, which allows driving the power amp very hard at the expense of fine control at lower volumes. I've tried and a JCM800 does not sound good on 10, in fact it loses everything I like about it past about 7 (my personal opinion of course, the more preamp gain you have the less power amp overdrive you want).

With the usual logarithmic pot, even 7 is a reduction of 80% in voltage! So last amp I made (a Dookie modded 20W JCM800 type thing), I used a 250k volume pot with an inline 680k resistor before it. Effectively, when on full it's like a normal MV on 8 or so. The result is an amp with much finer volume control, and it still has enough drive to get he power tubes cooking.

Friedman does something similar, but only a 270k resistor before a 1M volume with a 1M resistor in parallel (or a 2nd pot in dual master designs). I've tried it, still too twitchy and still too much unusable range. So I'm pushing it farther with a 470k and 2 500k volume controls. Add to that the post loop Master and it'll be really easy to control at low volume. That's only my starting point though. But I'm confident it'll be a good starting point.

[gz34]

this is an extremely interesting post, please keep us up to date..

[Snicksound]

this is an extremely interesting post, please keep us up to date..

Thanks!

The amp is making sound now (still many things to hook up though),

First lesson learned: the shared dual mode Presence control was not a good idea. In theory it should work but it created some kind of oscillations issues. Now I understand why Mesa uses a dual gang pot for this (they even sell the 25k/100k power for a decent price, but shipping to Canada is too expensive for such a small part). Either way, I've resorted to applying a fix Presence boost in Marshall mode (1k to ground from the .1uF presence cap) and leaving the Mesa style Presence engaged at all times. So the Mode switch really just disconnects NFB now. It actually works really well.

My next issue is that... feels like I'm not getting enough gain from the 2204 circuit. Even with a bypass cap on V2a, it only gets dirty in the last 10% of the potentiometer's range, heck, the Plexi side can get almost just as dirty. I know I'm losing some signal by paralleling 2 gain controls, but then why is the Plexi side almost as dirty? It's not a huge deal, but it forces me to extract a lot of gain from the FET Boost to get it where I need to be, which makes for a stark contrast between boosted and unboosted. Either way, I'm working on that.

The good news is it's really quiet. No noise at all in 2204 mode, and just a bit of hiss with the FET Boost engaged (like you can stay there on boosted mode between songs and no one would notice in at a gig).

And the channel switching works perfectly, no popping. It's not the greatest clean sound in history (it's basically a Plexi Lead channel with a Master), especially with a bridge humbucker. But it's perfectly serviceable for gigs (in the studio you can always re-eq for a better clean tone). You can get it to overdrive like I said which has its advantages. You could set the amp up as your rhythm and lead sounds (instead of clean vs crunch) and the FET boost would allow the rhythm side to get fairly dirty. I'm thinking of making the Plexi gain control a push-pull for a bright boost (like add a 100pF or 220pF bright cap), could get handy with humbuckers, or when trying to get it to sound like AC/DC.

[Snicksound]

So here's where I'm at. Very satisfied with the circuit like this. Will be posting some clips soon hopefully.
While this ain't no Blackface clean, the clean and dirty channels are well balanced.

One interesting tidbit though, I wasn't so sure when playing through my G12M-65 Creamback equipped cab, and since the future owner of this amp is running a Mesa Rectifier 212 cab, I procured a Mesa sourced Celestion V30 (T4335B) and loaded it in one of my own 1x12 cabs and... wow! Amp came alive! I guess some circuits just work better with some speakers. Still surprises me though because it's still very close to a 2204 and I really liked my 2204 in the Creambacks. Maybe I'm a V30 guy after all, who knows. This is however more mid scooped and smoothed out than a proper 2204.

[Snicksound]

This will probably be the last post on this amp as it's going to its new home tomorrow.

See below the final schematic and a potato quality demo.

I did get to play one gig with it and it was GREAT. I found the boost to be a bit intense for some songs so I complemented with my pedals. When I wanted proper high gain though, I found the onboard boost to be best in terms of transparency and punch.
Did not experience any popping. Amp was super quiet, no weird noise, very well behaved.

Some late changes:
- Rk on 1st stage lowered to 1k8 (coincidentally, same as on a Mesa). With the boost on, it just sounds better than 2k7.
- Divider after the cold clipper changed to 220k/470k and no capacitor. It's a nice balance of cutting some fizz and being just dirty enough in non boosted mode.
- Presence pot increased to 250k with a 2n2 cap. I found that I was losing too much highs with the FX Loop engaged and it's not the first time I notice this with this type of Presence control coupled with this type of FX Loop. Whatever the reason, increasing the pot value and lowering the cap gives me a better sweep I find.

That said, for my next build I will probably go with a dedicated clean channel with its own tone stack. For this one I was working with reduced real estate (had to fit on top of a 2x12 Mesa cab on its side) and that meant some sacrifices. Clean channel is very decent, but EQing will always be a compromise.

Also, the FET boost PCB is super useful for getting more gain without adding a tube, but I find it has a bit of stiff feeling.

Anyway, I'm getting started on my next project next week.

[adew]

Very cool build, Snicksound.

I know you no longer have the amp, but do you have any guts photos you could share? Curious to see the final board and layout. Also, I have a couple of questions, if you don't mind:
1. The 47R resistor after the rectifier, how did you arrive at the value?
2. About the FET boost, I've heard Jason of Headfirst saying that he think it works best after the input grid stopper, but I noticed from your schematic that you placed it before, and was wondering if you had tried both ways and, if so, what difference there was, if any.

Thanks!