adding a tube driven reverb to a champ clone

[wagdog]

Greetings,

I have an AA764 champ clone that I built... with the exception that it runs a beefier PT and OT and uses a 6L6 (it was originally a Dave Hunter two-stroke, if you're familiar with that).

It sounds really, really nice, especially with the 8" and 10" speaker combo.

I'd like to add an internal, tube driven reverb.

Because it was a two-stroke, I've got a hole in the chassis where the 2'nd 6V6 used to live, so I can easily add another tube to drive the reverb. Since the PT was overspec'ed, powering the reverb won't be an issue either. I was thinking of using a compactron type tube (got a socket for one in my junk box), like a 6D10, which is 3x 12AT7's in one bottle, using 2 in parallel to drive the tank and one for recovery. Or, possibly going with a 6CX8 and using the pentode to drive the reverb and the triode for recovery (the super twin reverb does this, except it uses one triode from a 6C10 for recovery).

I'm kind of leaning towards the 6D10.

Anyone have thoughts or experiences to share with either of these approaches? Perhaps a different approach?

Thanks in advance.

[imaradiostar]

If you can find a compactron and socket for cheap it sounds like a great plan!

Failing that you might want to just use a dual triode or even a 6v6 to drive the verb and an LND150 to recover it. Set the LND150 up like any normal triode stage with a 100k "plate" resistor. You might have to mess with source resistors to get the gain and current draw of the mosfet right.

Where do you plan to insert the verb circuit? Seeing as a champ is at it's best relatively clean I'd probably drive the verb from the plate of V1a (first triode) and have the recovery amp drive the grid of V1b (second triode) through a blend resistor of some sort. You could use a 1meg pot as a dwell control. Use a second 1 meg pot as reverb level- drive the wiper of the pot with a 10nf cap from the drain of the LND150. The CCW end of the verb pot would be grounded and the non-grounded end connects to your blend resistor at V1b's grid.

If you stick with the compactron I'd use more or less the same layout with the triode instead of the mosfet.

jamie

[wagdog]

Thanks for the reply.

I have the compactron socket left over from my old super twin reverb that I gutted and built an ab763 inside of it. I have 2 6C10's as well, one old and one new. The 6D10 is $12 from tubedepot - I'm thinking that would be a better driver then the 6C10 as the 6c10 is 3x12ax7's, and the paralleled 12at7's in the 6D10 can handle the current better.

I kind of like the idea of driving the tank w/a 6V6 and using a mosfet for the recovery stage, that would be pretty cool.

The compactron tubes are kind of neat though, and I like the idea of having the driver and recovery in the same bottle.

I was actually thinking of putting it after the .022 coupling cap on v1b and having it straddle a 10pf/3M before v2 (the 6L6) like a typical fender.

BTW; here's a clip of the amp: http://soundcloud.com/wagdog-1/sweetampclips

[wagdog]

After giving this more thought (and seeing the threads linked below this thread), I remembered that a typical fender 3M/10pf mixer on the reverb has another recovery stage after it and that wouldn't work.

Googling "single tube reverb" brought up A LOT of previous stuff, so I'm thinking I could go a lot more simple then using a compactron and just split a 12ax7 and get decent results, or use a pentode/triode tube.

[jmaf]

After giving this more thought (and seeing the threads linked below this thread), I remembered that a typical fender 3M/10pf mixer on the reverb has another recovery stage after it and that wouldn't work.

Googling "single tube reverb" brought up A LOT of previous stuff, so I'm thinking I could go a lot more simple then using a compactron and just split a 12ax7 and get decent results, or use a pentode/triode tube.

Has anyone tried a 12DW7 in there? If you drive the 12AU7 hard, via a bigger capacitor, the 12AX7 in there is more than enough for the return, too much gain in fact and you might get too much single coil noise if you don't limit the return IME.

I'd experiment driving the 12AU7 in a 12DW7 hard and see what happens, it's a current production tube(JJ ECC832 for U$ 9 JJ ECC832 / 12DW7 ). Never tried it but I'd be curious to know the results.

[Chuck H]

Right on. Often ignored is that a reverb pan is driven by current. Some guys stick a lower mu tube into a typical BF reverb driver socket and think that the lower reverb drive is due to the lower amplification factor. Wrong. The lower drive is due to the tube that was stuck in there being used in poor operating conditions for that tube type. With circuit modification it is possible to drive a BF reverb even harder with a 12au7 instead of a 12at7. But the circuit would need to be idealized for the 12au7, and it never is.

Half a 12au7 should provide excellent results driving a pan. And half a 12ax7 should be very good for summing and re-amplifying the wet and dry signals. Vote 2 for the 12dw7.

[wagdog]

With circuit modification it is possible to drive a BF reverb even harder with a 12au7 instead of a 12at7. But the circuit would need to be idealized for the 12au7, and it never is.

Thanks for that information!

Give a typical fender style reverb driver, what modifications would use suggest to use half a 12au7 (the 12au7 side of a 12dw7) as the driver instead of a paralleled 12at7?

[Chuck H]

After a quickie tube data check on Google I see that 1/2 a 12au7 (one triode) has a max diss of 2.75 watts. But that would be extreme. IME you need at least one good solid dynamic watt to drive a reverb pan. You should be able to do that.

Only three things really need to be different from the typical Fender circuit. One is the reverb drive transformer. You'll want about half the primary impedance that Fender uses IIRC. About 7k primary. Then, the bias condition for the drive triode needs to be set relative to the plate volts for max clean dissapation. I have no specifics, sorry. It's the kind of thing I would do on the bench. And last, the signal voltage for the driver triode will likely need to be larger than is typical for a BF type circuit. That also shouldn't be a problem since you'd be starting at the tail end of a two triode preamp. Fender used a 500pf cap to couple the reverb driver tube grid. Not so much for frequency limiting as voltage limiting. Remember that the function of the driver tube is to drive transducers into shaking springs. How much frequency control can there be. OK, there is some, but... The cap size is primarily limiting the voltage swing. A bigger cap would increase drive. As would increasing the plate resistor of the stage feeding the reverb/mix resistor send triode.

Two things come to mind that may be problematic. One is that you'll be asking the PT to provide a pinch more filament and HV current. Basically asking the amp to make an extra one or two watts. That's what drives the pan. Since that amp is only about 3.5 watts now, that would be 30% more current you'll be asking from it. And that amp has an NFB loop that includes the triode you'd be driving the tank with. This puts the reverb in the NFB loop. And it's never a good idea to have any kind of effect, EQ or phase shifts within the NFB loop. I can see where reverb would be especially problematic due to phase errors and such. You'll need to move the NFB loop to the new "mixer" stage that will feed the power tube grid.

[wagdog]

fwiw, here is the schematic of what I built:

http://ampwares.com/schematics/champ_aa764.pdf

except I used a bigger PT and OT and am running a 6L6 into a 8" and 10" speaker (parallel, 4 ohms)

Only three things really need to be different from the typical Fender circuit. One is the reverb drive transformer. You'll want about half the primary impedance that Fender uses IIRC. About 7k primary. Then, the bias condition for the drive triode needs to be set relative to the plate volts for max clean dissapation. I have no specifics, sorry. It's the kind of thing I would do on the bench.

That's cool - I can experiment/work it out as I build it.

Two things come to mind that may be problematic. One is that you'll be asking the PT to provide a pinch more filament and HV current. Basically asking the amp to make an extra one or two watts. That's what drives the pan. Since that amp is only about 3.5 watts now, that would be 30% more current you'll be asking from it.

w/the 6L6, I'm thinking it's about 8w actually, and I have plenty of capacity left over in the PT as it was intended to drive 2x6V6's originally. It's this weber transformer: https://taweber.powweb.com/store/022772sch.jpg - and it's currently driving a single 6L6 (60ma) and a 12ax7.

And that amp has an NFB loop that includes the triode you'd be driving the tank with. This puts the reverb in the NFB loop. And it's never a good idea to have any kind of effect, EQ or phase shifts within the NFB loop. I can see where reverb would be especially problematic due to phase errors and such. You'll need to move the NFB loop to the new "mixer" stage that will feed the power tube grid.

Not quite sure how that would work... the nfb is feeding into the 2'nd triode, and if I put the mixer resistor between that triode and the 6L6, you're saying that puts it in the nfb loop (seeing that now)...

What if I put it before the second triode, just after the vol and before v1b's grid? I could also put it before the tone stack... ?

[Mandopicker]

A freind of mine wanted to do the same thing. Add a tube reverb to a SF Champ. The cab was in bad shape, but the original chassis was in great shape.

My apologies to the collector purists should you continue reading. This was after all made to be a players amp.

My friend who commissioned it is great with wood working so I had him actually copy all the dimensions of a Princeton cabinet, but had him make it the width of a Champ. Installed a 10" speaker and also runs a 6L6. In this configuration, the bass response is excellent. I also added DC filaments, a middle control, adjustable bias and the amp really sounds unbelieveable.





The OT has since been changed to a 125ESE. The 125E worked, but didn't seem to sound as good..less articulate...a bit more "brown" sounding and seemd to break up a little bit more.

He calls it HIS ultimate bedroom amp.

Enjoy!

[Chuck H]

First, I gave some misinformation above. I corrected my post but it's still in my quoted area of your post. If you used a standard Fender reverb transformer you would actually need a tank of half the typical 8 ohm impedance to reflect the correct primary Z on the 12au7. Pretty sure no one makes a four ohm tank. So a small 6.6k to 8k single ended OT would be needed.

What if I put it before the second triode, just after the vol and before v1b's grid? I could also put it before the tone stack... ?

No good. Not enough voltage to drive the 12au7 to the needed power for driving the tank. It may work, but the reverb would be weak and anemic sounding. Not lush and full.

There are three ways to do it. One is to use the high mu triode of the 12dw7 for the second preamp triode and use the triode driving the power tube grid as the mixer/recovery stage. That would mean you could basically leave the second triode in your amp wired the way it is now on the plate and cathode. The second way would be to remove the NFB loop from the second triode and use that triode to drive the reverb triode, then use the high mu triode of the 12dw7 as the mixer/recovery stage that will drive the power tube and hook up the NFB loop on that triodes cathode the way your second triode is wired now. The third way would be to do it any way you want and just remove the NFB loop all together.

[Chuck H]

Interesting. I can't decipher from the layout all the circuit specifics (actually I don't want to. a schematic is much less work to read). So I can't tell where the reverb is placed in the circuit path. I can see that the amp has two extra tubes instead of just one. I'm not personally into the braid insulated wire and carbon comp resistors but I'm sure it sounds great. Nice job on the twisted DC filament leads too. Done for nostalgia no doubt.

The scheme we're working on should only require one extra tube. As is the OP's original intent. DC filament seem to be the wave but I've done a couple of single ended amps and used a plain Jane AC twisted pair with the filament CT DC elevated on top of the cathode resistor. Super easy, no extra circuitry and quiet as a mouse.

[wagdog]

A freind of mine wanted to do the same thing. Add a tube reverb to a SF Champ. The cab was in bad shape, but the original chassis was in great shape.

My apologies to the collector purists should you continue reading. This was after all made to be a players amp.

Interesting - can you share a schematic?

[wagdog]

First, I gave some misinformation above. I corrected my post but it's still in my quoted area of your post. If you used a standard Fender reverb transformer you would actually need a tank of half the typical 8 ohm impedance to reflect the correct primary Z on the 12au7. Pretty sure no one makes a four ohm tank. So a small 6.6k to 8k single ended OT would be needed.

Thanks for that - noted!

There are three ways to do it. One is to use the high mu triode of the 12dw7 for the second preamp triode and use the triode driving the power tube grid as the mixer/recovery stage. That would mean you could basically leave the second triode in your amp wired the way it is now on the plate and cathode. The second way would be to remove the NFB loop from the second triode and use that triode to drive the reverb triode, then use the high mu triode of the 12dw7 as the mixer/recovery stage that will drive the power tube and hook up the NFB loop on that triodes cathode the way your second triode is wired now. The third way would be to do it any way you want and just remove the NFB loop all together.

I'm going draw some of this out to get my head around it and then figure out where to go.

This looks interesting as well: http://music-electronics-forum.com/t22535/

[Chuck H]

This looks interesting as well: http://music-electronics-forum.com/t22535/

For me, personally, that wouldn't be as good either. Because your amp has a tone stack the grid drive in that location is going to be tiny. And if you take the signal before the tone stack you won't have the actual processed signal being reverberated. Just seems like a compromise to me. The tube choice seems fine though. My only reservation might be that it isn't a humbucking filament. Considering that your amp is single ended you might want as much hum abatement as possible built in.

EDIT: If you look up dchang0's posts you'll find two more threads about this amp and problems he was having with the circuit.