Franken-Amp: 5E3 + AB763

[cryptonomicon]

Have been mulling over an amp design - as one does when one reads RobRob's site a little bit too much. What started as the idea of a two channel amp with a AA1164 Princeton Reverb on one channel and a 5E3 Deluxe the other, tone king imperial sort of thing, has morphed into replacing the normal channel on an AB763 Deluxe Reverb with a 5E3 pre-amp and tone stack.

Conceptually can you see any major problems?

Basically it is the tone stack and first two preamps of a 5E3 up to and including the coupling cap on V2A and plugging that into the input of the 3.3M mix reverb mix resistor - al-la the robrob normal channel mod. And then putting in a Lar-Mar trainwreck 2 Master Volume and Negative Feedback Mod.

The weird bit, and am still thinking about it, is a blend between Channel 1 and Channel 2 so that you can dial in some tweed dirt into the blackface cleans. I am seeing a mid scooped, mid pushed thing going on here.

It makes sense to me, but then I don't know enough to know what I don't know.

 

[Dacious]

A lot of the sound of a 5E3 is the cathodyne phase inverter which the Princeton has but the Deluxe Reverb does not. The Princeton power rail is too much b+. You can drop that of course. Then there's the fixed bias of the Princeton vs cathode of the 5E3, big coupling caps. Then there's the negative feedback loop of the Princeton. You'd have to split the preamps as the 5E3 has only one preamp stage to push gain whereas the PR has several extra stages to push signal through the reverb tank.

 

[Lowerleftcoast]

The weird bit, and am still thinking about it, is a blend between Channel 1 and Channel 2 so that you can dial in some tweed dirt into the blackface cleans. I am seeing a mid scooped, mid pushed thing going on here.

I don't see these circuits the way you do.

The 5E3 does it's well known interaction between the channels in the preamp but there is no 5E3 dirt there. There is simply not enough gain. The 5E3 dirt is power tube distortion. The ease of driving the 6V6 tubes is part of the equation.

The other circuits you are looking at, especially considering the use of a MV, are getting dirt from the preamp and/or PI. This may or may not be blended with power tube distortion.

For your vision, the major barriers I see are the difference in B+, the types of PI, and the primary impedance of the OT.

To maintain a 5E3 sound, imo, you will need to build a 5E3 and add a clean channel.
I would try to push the B+ up to about 380VDC (imo, that's about as high as a 5E3 can go before it loses 5E3-ness).
Add a channel with a scooped tone stack. It would be good to provide this channel with as much B+ as you can.
Add a switched NFB loop similar to what Rob shows.
Add control grid stoppers and grid resistors on the 6V6. Just follow the Deluxe Reverb circuit for these resistors.
Add a switch to change from cathode bias to fixed bias.

Starting with Blackface-ish amp. Maybe adding a compressed 5E7 type channel along with lowering voltages, increasing the OT primary impedance, and doing some of the switching described above will get some tweed-ish character. The Blackface preamp would need as much B+ as you can find.
Perhaps a 6V6 version of a 5F6A and a raw control to get less scoop from the tone stack. Either way, I think to get the tweed sound from a blackface-ish circuit, the Blackface clean will be diminished.

 

[D'tar]

You may like our friend @printer2 has many ideas and real live amps that fit this catagory! You may need a 20p20t switch

 

[schmee]

I always wanted to get an old Twin reverb and make it into two amps in one. The Normal channel would be a 5D3, the Reverb channel might be a Vibrolux or similar. Each would have it's own preamp, OT and speaker, only sharing the PT. The 5D3 would be set for lead tone. Simple A/B switching for channels. Vibrolux clean on one side and thick "hotel california" lead tone on the other side.

I wonder if a Vibrolux Reverb PT could handle both amps? Maybe without any reverb on board...? Maybe I'd go Brown Vibrolux on that side... although I'd want LT PI.....

 

[printer2]

You could use the normal channel with a Tweed volume/tone control to get some mids to blend in. As the others have said the sound of the 5E4 mainly comes from the output stage and the power supply. You could reconfigure the normal channel to have a bridged two channels of the 5E3 with two volumes controls and tone control in place of the volume, treble and bass controls. Rather than run into the 3.3M resistor I would feed it into a 470k mixing with the existing 3.3M and 470k resistors. This will get you some Tweed flavor also but will not be a 5E3 loose sound due to the PS and NFB. While a multi-position switch is not needed (only two position) more than a few poles is always a good thing. The simplest way to go for an amp switching between circuits would be to add a hefty sag resistor, a pair of poles in front of the master volume that come from either the LTP or Cathodyne inverter (need an extra triode). Before the inverters the regular preamp circuits.

 

[cryptonomicon]

The 5E3 does it's well known interaction between the channels in the preamp but there is no 5E3 dirt there. There is simply not enough gain. The 5E3 dirt is power tube distortion. The ease of driving the 6V6 tubes is part of the equation.

That's the bit I am going to have to drill down on, to work out exactly where the dirt comes from. Build a test rig I suppose, expose test points after the coupling capacitors and measure it.

Some sort of box with all the high tension circuits safely beyond reach. I suppose I could go further and expose the bias circuits and tone stack with some sort of bread board thing stuck on top.

 

[Lowerleftcoast]

Google *Fender 57 Deluxe schematic*

Fender was kind enough to include the AC signal voltage test points on this schematic. The VAC is in the oval test points. You can extrapolate how there is not enough voltage in the signal path to overdrive anything until it reaches the 6V6 tubes. For fun check the difference from TP13 to TP14 to see how much signal is lost with the volume and tone pot configuration.

 

[cryptonomicon]

TP13 to TP14 92% of the signal is lost ??? So if I plug the 300 mA at TP14 into a load line diagram for V2A, it does bugger all - a bees fart would probably do more than that.

TP14 to TP15 is about 58 times gain - that seems to be normal for an AX7 according to Wikipedia.

Cool, thanks for that

 

[cryptonomicon]

Had another look at the schematic of the 57 Deluxe. Trying to work out where the signal went nagged at me.

It turns out, according to the notes on the schematic, the voltages are measured with the pots at 50%, which would make sense. Except they are log pots. That means that at 50% of a turn as a voltage divider 80% to 90% is sent to earth. When the pots are wide open, then its a different story, with the full 3.79 volts is hitting the front of V2A.

When I look at robrobs loadline chart for V2A of 5E3 , with the operating point close to say the -1.5v curve , then that would distort on both positive and negative swings.

 

[Pete Farrington]

Except they are log pots

But they’re not wired up as variable potential dividers. So every setting is a different AC loadline.
Rather they’re arranged to be a variable load on their channel’s input stage.
With the remainder of the track, along with that of the other channel, acting as mixing resistors.

https://www.thetubestore.com/lib/thetubestore/schematics/Fender/Fender-57-Deluxe-Schematic.pdf

 

[cryptonomicon]

So every setting is a different AC loadline

So every time you adjust the controls you re-voice the amp? That would also affect V1 as well methinks.

Perhaps I used the wrong word when I said divide. I couldn't understand how 3.79V at TP13 went through R12 to come out as 0.3V at TP14, so did a little spice analysis of the tone stack in Kicad. I modeled the pot as two resistances on either side of a wiper, al la voltage divider circuit. At 50% resistance either side of the wiper it gave the wrong answers, but when I used a 10A taper (10% at 50% rotation) that dumped 90% to ground, it worked.

The question though is whether V2A contributes to distortion, and by how much. At 0.3V it would appear not much, at 3.79 V quite a lot.

By way of a sanity check, if the pot is wide open and out of the circuit, TP13 should be at the same potential as TP14, implying there should be swings in the order of 3.79V around the operating point. Presuming 3.79 V is RMS then that is a 10.7V peak to peak swing. Even if the operating point moves 3.79 V is likely to distort.

Anyway now I have the tone stack model I suppose I should connect it to an 12AX7 tube model, but have have no idea how well spice models work when pushed into distortion. Probably a lot worse than amps.

And yes I read 10A on the schematic as 10 Amp the first time as well. Turns out 10A is a type of taper, https://www.potentiometers.com/potcomFAQ.cfm?FAQID=29

 

[clarksdale]

Gerald Weber has a tweed-tone conversion for the normal channel of a Fender, leaving the other unaffected. It's in the "Tube Guitar Amp Essentials" book. I did it on a silverface SR, and it really sounded pretty good...

i might have a copy of the details somewhere- pm if you'd like a copy.

 

[Lowerleftcoast]

.

By way of a sanity check, if the pot is wide open and out of the circuit, TP13 should be at the same potential as TP14, implying there should be swings in the order of 3.79V around the operating point. Presuming 3.79 V is RMS

The 5E3 is an interesting study isn't it?

Imo, maintaining 3.79V at V2a is unlikely. As the *other side* is turned up we actually hear the loss of signal due to the interaction of the *that side* of the circuit. I am not saying V2a can not contribute some distortion as the signal voltage rises. What I am saying is your expectation of getting *5E3 tweed dirt* by inserting the 5E3 preamp into a AB763 type circuit is too optimistic. That is why I suggested the 5E7 preamp. It has something happening with the cathode follower (maybe tweed dirt/compression)... anyway it has something more than what the 5E3 preamp can contribute to a Franken AB763.

 

[cryptonomicon]

The 5E3 is an interesting study isn't it?

I certainly seem to have gone down the rabbit hole just on that very first preamp.

Because the volume control is in parallel with the load resistor of V1, turning the volume control in effect creates a variable load resistor, that enables you to rotate the AC load line from almost horizontal (lots of gain) at maximum, to almost vertical at a minimum (lots of headroom) at minimum.

Playing around with the "Amp Books" calculator, "AC load of following stage", lets say from 1k to 1000k, allows you to just rotate the AC load line about the operating point like turning a tuning dial.

Amp Books 12AY7 Calculator

The question being why would someone design things that way? All I can come up with is 1957 , no real standard for electric guitar pickups. So possibly leo fender simply designed an amp that could handle anything. Remembering he used to make PA's , so perhaps his vision was a person could plug any guitar and mic in, and be able to adjust the gain to handle their particular equipment.

Actual control of volume and tone was intended from the guitar.

Along comes the 1960's and broke teenagers playing with second hand equipment find themselves with an amp with what amounts to a tube screamer and a fuzz built in to it, that can be controlled by stomping on a volume pedal.

Well one can imagine.

 

[Phrygian77]

I always wanted to get an old Twin reverb and make it into two amps in one. The Normal channel would be a 5D3, the Reverb channel might be a Vibrolux or similar. Each would have it's own preamp, OT and speaker, only sharing the PT. The 5D3 would be set for lead tone. Simple A/B switching for channels. Vibrolux clean on one side and thick "hotel california" lead tone on the other side.

I wonder if a Vibrolux Reverb PT could handle both amps? Maybe without any reverb on board...? Maybe I'd go Brown Vibrolux on that side... although I'd want LT PI.....

You could do it with a Twin Reverb PT, but you'd end up with a BFDR-like B+. I'd probably use a lower voltage 200VA Antek. Although, a TR chassis is long enough, you could probably adjust the layout enough to fit a second toroid underneath.

I think it isn't a very practical approach, since you're still dependent on the power tubes for your overdrive, and with that comes the same old volume issues we've always had.

Having two completely separate channels, one for clean, and one with extra gain stages for overdrive is also inefficient, especially if you're also trying to cram reverb and/or an effects loop in there. I'm starting to realize that Dumble was on the right track with his topology.

 

[schmee]

You could do it with a Twin Reverb PT, but you'd end up with a BFDR-like B+. I'd probably use a lower voltage 200VA Antek. Although, a TR chassis is long enough, you could probably adjust the layout enough to fit a second toroid underneath.

I think it isn't a very practical approach, since you're still dependent on the power tubes for your overdrive, and with that comes the same old volume issues we've always had.

Having two completely separate channels, one for clean, and one with extra gain stages for overdrive is also inefficient, especially if you're also trying to cram reverb and/or an effects loop in there. I'm starting to realize that Dumble was on the right track with his topology.

Yeah, the first issue to come up was voltage for sure. For the weight of a Twin PT, I suppose one could avoid paying much weight penalty just using two smaller PT's. The 5D3 one would be small I imagine. The other, maybe a Vibrolux type is not too big.
In the end though, it's a big, 2x12 amp to carry around even if it's not a Twin anymore!

 

[Gris]

Just find one of those old Gretsch 6166 Fury amps with two output transformers. I had one for a long while. They are already 60% of the way there. The reverb channel sounds blackface like. And the non verb channel is more Marshally.

 

[Wally]

Have been mulling over an amp design - as one does when one reads RobRob's site a little bit too much. What started as the idea of a two channel amp with a AA1164 Princeton Reverb on one channel and a 5E3 Deluxe the other, tone king imperial sort of thing, has morphed into replacing the normal channel on an AB763 Deluxe Reverb with a 5E3 pre-amp and tone stack.

Conceptually can you see any major problems?

Basically it is the tone stack and first two preamps of a 5E3 up to and including the coupling cap on V2A and plugging that into the input of the 3.3M mix reverb mix resistor - al-la the robrob normal channel mod. And then putting in a Lar-Mar trainwreck 2 Master Volume and Negative Feedback Mod.

The weird bit, and am still thinking about it, is a blend between Channel 1 and Channel 2 so that you can dial in some tweed dirt into the blackface cleans. I am seeing a mid scooped, mid pushed thing going on here.

It makes sense to me, but then I don't know enough to know what I don't know.

Check out the Tone King Imperial MKII. The schematic for the Imperial 20th Anniversary amp is online And might be a help to you.