vibro king example. taking half the voltage from the capacitors in series?

[pietro_moog]

hi guys.
i'm building an high gain amp with 2 channels. one is an ordinary Soldano OD, the other is a Fender BF channel with a master volume.
the point is that i wanna add reverb to the clean channel, so i was thinking about the standard fender reverb with tone and dwell controls.
i have a 5w OT and a 6v6 spares from a previous 5f2 build. so i thought about using those instead of the 12at7 arrangement, and i think it would be a much sweeter reverb.
but my main voltage is too high, like almost 500v on the 6l6 plates, and using that for the 6v6 is out of discussion.
i took a look at the vibro king power supply and that seems a nice solution. it has a similar voltage for the 6L6, and it take the power for the 6v6 reverb driver half way in the capacitor arrangement (it has the classic capacitor-in-series solution, like the others big Fenders and my Soldano)

i've never seen that solution. is it safe? how does it work? is it gonna put stress on the power main line?
it seems i could get almost half the voltage of the HT, and that would be perfect.
what do you think?

[Jazz P Bass]

The Vibro King power supply is a voltage doubler.
I do not know that a "classic capacitor-in-series solution" was used by Fender.
As to the 6V6 plate voltage, why not 500V?
It's the current through the tube that kills them.
Not so much the voltage.

[pietro_moog]

hi. i'm sorry for my newbieness (actually i've been studying a lot of electronics in the last few years)
what do you mean for voltage doubler? is it the 2 caps, one on the top of the other, to handle a voltage which correspond to the sum of the 2?
because that's what i meant. can i take the power from the middle of that? is it safe? i don't wanna risk that the amp to explode in the middle of a gig, for everybody's safety (mine especially).
500v on the 6v6 i don't think it can handle that.
i've seen this thing on the vibro king, i'd love to understand how it works
thanks

[Steve Conner]

The voltage doubler is more than just two capacitors in series.
Voltage Doubler,Voltage Tripler,Voltage Quadrupler circuit
There is a related circuit using a bridge rectifier with a transformer centre tap connected to the midpoint of the two capacitors.

These are the only two circuits that let you draw current from the midpoint of the capacitors, since there's a transformer winding there to supply it. If you try it with any other circuit, the voltage just collapses and the upper capacitor dies from overvoltage.

[pietro_moog]

mmm
i understand. this circuit looks like the marshall 1959 too.
maybe it's not worth it, maybe it is better to use a 12at7 for the reverb driver, instead of complicating things.
easier and fewer components.
is it so worse sounding?

[Chuck H]

Why not just branch the power supply and decouple it again. FWIW the BF circuit has the reverb driver on the screen node of the power supply. Probably because there is less ripple at that node. But if you branch from the first filter node and decouple again it's the same thing. Leo wouldn't have done it that way because it would mean an extra resistor and capacitor.

[pietro_moog]

i didn't understand. what do you mean? cutting down the voltage with a resistor? but i would need something like a 25W resistor. it's crazy..

[Chuck H]

Yes, Cutting down the voltage with a resistor. Since the reverb driver is operating in class A the current draw should be fairly constant, so you could branch from the screen node. Just think of the word "branch" and it should become clear. From the screen node, add a resistor/decoupling cap PARALLEL to the existing +V rail. So the resistor doesn't need to be that large because it will only be supplying current for the reverb driver. I was trying to avoid making a drawing. But if it's still not clear I will.

NOTE: On your existing +V rail the resistor or choke between power tube plate and screen node would also need to be bumped up in rating to handle the extra current from the branched reverb node.

[pietro_moog]

i don't understand what you are saying to me. i should run a 6v6 at 250-300v, from the screen node, having there 500v? all with a resistor and a cap?
it seems crazy to me. am i missing something? i suppose i could run the tube at 380v using a 10W resistor, but it seems crazy too.

[Chuck H]

You wouldn't be asking the 6V6 to make anything close to it's wattage potential. You can't put five watts into a reverb pan without failure. You only want about one clean watt of potential. As to the dropping resistor, suppose you get 380V using a 10k. Your dropping, what? 100V from the screen node? So 100/10,000=.01 amps. .01*100=1, or 1 watt. Since the tube will operate in class A that shouldn't change significantly. Use a three watt resistor. You'll have plenty of margin and the resistor will likely only get warm. Aside from that, I never said to use the 6V6 (Though I didn't say not to. That's up to you.). I was only suggesting a solution to your high voltage problem as it applies to ANY reverb circuit.

[pietro_moog]

well, thanks. it's so obvious i need to study a little more, especially power tubes. where did you get the 10mA from? the 12at7 idea? that's cool, and that's probably what i'll do. the 6v6 would have been super-cool, mainly because i have already a spare 6v6 and a trafo for it. but the 12at7 it's ok for space reasons and easiness.
thanks

[Chuck H]

If space is a consideration you could use a 12dw7. The low mu half is good to three watts and works just dandy for driving a reverb. The high mu half has an amplification factor of 100 and works just dandy as a recovery stage. A one tube reverb that actually works very well.

If you already have a recovery stage for some other circuit you could use a 12au7 as the driver and skip the trafo all together by using a higher input impedance tank. The Valve Wizard site has a good method for doing this outlined in detail. Lots of options.

I got the .01A by dividing the intended voltage drop across the resistor by the resistance. And it was just an example. I don't actually know that your resistor value would need to be 10k. It would depend on your idle current for the reverb driver. And you don't actually need to idle the 6V6 all the way into class A for class A operation at low signal levels. So the current demand wouldn't be so large. lots of options.