Reverb using two triodes

[tubeswell]

Does anyone know a design for a two-triode reverb that I could adapt for a 6G2? (Is there such a thing? or do you really need three triodes? All the fender schemes I can find so far seem to use three triodes)

I am procrastinating over a 5G9 vs 6G2 build (DANG! just when I thought I had become determined). But the idea of reverb and tremolo is somewhat more exciting to me than trem on its own.

I have a PT with four secondary windings (one for 50V). The High Voltage winding is 300-0-300 150mA. The recto winding is 5-0-5 3A. The heater winding is also 3A. Is this going to be enough for adding a 'verb or not? If not I'll go with the 5G9 or the 'later 5E9A' variant

I have a 5U4GB, 2 x 6V6Gt in the power section.

[d95err]

Search the Dumble-forums over at www.ampgarage.com. They have a single-tube reverb design, that apparently works quite well.

[gdgross]

Do you want to use 1 triode simply to save space? If so, and current draw is not a consideration, perhaps consider a pentode/triode in one bottle, like the 6GW8 or 6MB8 tubes.

I've used one of them (can't remember now, it's which ever one has the triode with the higher gain) for a reverb driver/recovery and it worked great. You get the power you need to drive the tank and a triode for recovery all in one package.

[d95err]

Another solution is to use a regular reverb driver stage, but use a JFET return stage. You can probably copy the FET input circuit of Dumble amps for this. OK, not for the tube purists, but you're never going to be able to drive that JFET into distorsion with the reverb return anyway.

[tubeswell]

Another solution is to use a regular reverb driver stage, but use a JFET return stage. You can probably copy the FET input circuit of Dumble amps for this. OK, not for the tube purists, but you're never going to be able to drive that JFET into distorsion with the reverb return anyway.

What sort of JFET would I use and have you any suggestion as to the particular amp circuit I should look at? I guess it uses a lower voltage power rail for the JFET. Could I run that maybe off the 50V fixed bias winding I have on my PT? How would I do that?/?

Sorry for all the questions

[Enzo]

Look at some of the old AMpegs. Didn't they drive the pan with a single tube. Hi Z pan, like a 4FBxxxx, and the signal comes off the plate through a cap. Look at the V series - V2, VT22, VT40.

[Bruce / Mission Amps]

The biggest problem is that the reverberated signal out of the reverb tank is tiny, only about 4mv-10mv under the best case conditions.
Take that as an average of say 7mv and run it right into a 12AX7 triode with a ton of gain, say x45... and you still only have a small, reverberated audio signal of around 300mvac.
Many players have guitars that can put out nearly that much signal and maybe more then that... right from the pickups!
That is why the much stronger dry signal is filtered through a resistor and parallel capacitor, to reduce it's level so it isn't too high compared to the weaker reverb signal. It is mixed with the tiny reverb signal after that reducing stage and then the sum is reamplified to a much higher level, before driving the next stage or phase inverter.
So the solution is actually simple... one or two small signal FETs.
Use J201s, (set up just like a 12AX7 triode but at 18vdc), as the new reverb tank recovery amplifier with the final output going to your reverb level pot.
Then on to the RC mixer stage followed by the second triode for mixed amplification.

[tubeswell]

Use J201s, (set up just like a 12AX7 triode but at 18vdc), as the new reverb tank recovery amplifier with the final output going to your reverb level pot.

Thanks Bruce

Are the JFETs wired like a Darlington Pair?

[tubeswell]

Look at some of the old AMpegs. Didn't they drive the pan with a single tube. Hi Z pan, like a 4FBxxxx, and the signal comes off the plate through a cap. Look at the V series - V2, VT22, VT40.

Thanks Enzo

Will Do

Can you still get those type of Pans (do you know)?

[Enzo]

Sure.

They still make them, these guys sell them, http://www.tubesandmore.com/, and I thin ,ost anyone who sells Accutronics pans probably does as well.

[Bruce / Mission Amps]

Thanks Bruce

Are the JFETs wired like a Darlington Pair?

No. Maybe you could do them as a cascode but I meant two in series and in common source mode, with a small signal divider between the two so the total gain was not too high nor too noisy.

By the way, in the search for adapting reverb configurations to be used with different amps, I have tried the speaker output source method and to be honest it does work but it is not great reverb.
Check out the Fender Champ 12 and how it's reverb is generated by stealing audio from the output transformer secondary through a 47 ohm resistor to the tank.
The tank output is recovered with a small signal FET.

[tubeswell]

No. Maybe you could do them as a cascode but I meant two in series and in common source mode, with a small signal divider between the two so the total gain was not too high nor too noisy.

By the way, in the search for adapting reverb configurations to be used with different amps, I have tried the speaker output source method and to be honest it does work but it is not great reverb.
Check out the Fender Champ 12 and how it's reverb is generated by stealing audio from the output transformer secondary through a 47 ohm resistor to the tank.
The tank output is recovered with a small signal FET.

Thanks Bruce. I see the J231 is powered by the Cathode bias of the output tube, and feeds the signal into V2B which then goes back into the grid of the output tube, so the 'verb uses one tube triode and one FET to recover the signal.

I see there is something unusual (i.e. that I don't recognise) going on in that the output tube grid seems to be biased to the output tube cathode. I don't understand that. Is this amp fixed bias?

[tubeswell]

More on this at the dumble forum

FWIW thanks to the earlier posts guys (d95err and gdgross), I found some useful threads at the Dumble forum describing an ECL86 where the pentode side was used as the driver and the triode side was used as recovery. Here is the schem (courtesy of Bob-I - gee thanks Bob)

Apparently the schematic here runs the pentode at 70% dissipation which is too high and it works better with "a 1k cathode resistor and a 27k with a 22uf / 450v cap to lower the screen voltage".

Food for thought

Now I just have to find some ECL86s somewhere. I thought I saw some at the local recycling junk shop in some old tube radios a while back. I must zip down there again and see if they are still floating around...

[hasserl]

Carvin runs a single 12AX7 reverb circuit in all of their amps, and it works rather nice, if it is not quite Fenderish. I personally don't use much reverb, only use a touch to wet the signal in a room without much natural reverb. I don't do the Dick Dale thing. So I find the Carvin system to be just fine for my needs. No external reverb transformer needed either. I can send a schematic if it helps.

[tubeswell]

ECL82s

Well I found some ECL82 s at the recycling centre just now.

The National Valve Museum has specs on both types

ECL82

http://www.r-type.org/exhib/aaa0052.htm


ECL86

http://www.r-type.org/exhib/aaa0034.htm

Looks like the ECL86 is the more powerful tube, but if I had less power supply to work with (i.e. a 300-0-300 Ht at 150mA running a 5U4GT with a 40uF first filter, two 6V6Gts, and also running two other pre-amp dual triodes in (say) a 6G2 circuit), then maybe the ECL82 would be the amp to design a one tube reverb circuit around?

Any feedback about this? That wasn't supposed to be a pun :-)

PS If I go this triode/pentode valve way, I might have to start another thread