Crazy tube idea?

[Diablo]

I was wondering if you could run a pair of power tubes in series. My idea is if you have some type of tubes with a relative low max plate current rating of say 250 volts, could you run two in series and apply 500 volts safely?.....like capacitors in series?

[tedmich]

you crazy!

[Chuck H]

I think there's a way to get a beam tetrode to float enough voltage and current to power another tube in series because IIRC Gibson used a 6V6 as a regulator tube in a couple of amps. But I don't know how you would make power with tubes in series.

[Enzo]

A cascode?


What would be the intended purpose for such a thing?

[Diablo]

The purpose for this would be to use widely available inexpensive NOS tubes which can't handle high plate voltages (e.g. 6AQ5) in conventional amp designs with around 350-470 volts on the plates. I know the straightforward way to use those tubes would be to just use a PT with lower B+. Just a wierd thought came into my head.....

[Chuck H]

A wierd thought that could have added benefits. I have a JL Hood book here with a Sonab schematic of a pair of power pentodes in what appears to be a cascode/cathode follwer type circuit that is intended to run OTL. I couldn't find the schem on line. I'm too tired to trouble with it now but maybe tomorrow night I'll hook up my scanner and post an image. Very interesting. Hood called it a "White cathode follower" arrangement. A search only shows cascode triodes.

[diagrammatiks]

no not really.

there are designs where you can run multiple power tubes after another but it's not going to give you the benefit of "series" power tubes.

Even a cascode like enzo said wouldn't give you the ability to exceed max plate voltage.

[mooreamps]

Even a cascode like enzo said wouldn't give you the ability to exceed max plate voltage.

yes it would...

-g

[diagrammatiks]

yes it would...

-g

how? learn me.

[Steve Conner]

Well, if I were feeling mischievous, I'd get an OT with ultralinear taps, and connect the control grids of my top cascode tubes to the UL taps, and bootstrap their screens to their cathodes.

All this strikes me as an exercise in Yugojetting. If you bolted a jet engine to the roof of a Yugo, it would probably go pretty fast. But you could buy a better car for the price of a used Pratt & Whitney.

[diagrammatiks]

Well, if I were feeling mischievous, I'd get an OT with ultralinear taps, and connect the control grids of my top cascode tubes to the UL taps, and bootstrap their screens to their cathodes.

All this strikes me as an exercise in Yugojetting. If you bolted a jet engine to the roof of a Yugo, it would probably go pretty fast. But you could buy a better car for the price of a used Pratt & Whitney.

are you imagining a 2 pentode cascode?

[Steve Conner]

I started trying to imagine a cascode made of 6AQ5s, which are like mini 6V6s. But somehow it turned into a jet-assisted Yugo.

Another problem would be heater supplies for the top tubes.

[R.G.]

The purpose for this would be to use widely available inexpensive NOS tubes which can't handle high plate voltages (e.g. 6AQ5) in conventional amp designs with around 350-470 volts on the plates. I know the straightforward way to use those tubes would be to just use a PT with lower B+.

This is a similar motivation to early transistor designs where they stacked lower-voltage devices to get more power. It was useful when higher voltage devices were unavailable.

It could probably be made to work the same way the transistor circuits worked; that being to make the upper device be a follower which follows a resistor divider from the output to the power rail, effectively making the upper device take half the power supply voltage in all conditions. The upper device becomes an active power supply which splits half the remaining power supply voltage with the active device and lets the active device take over all the amplifying.

The devil is (as always) in the details. It works pretty well with transistors as long as you can convince yourself that you'll never exceed second breakdown. With tubes, second breakdown is not an issue, but heater voltage breakdown is. The 6AQ5 is rated for 200V between heater and cathode. If you run a floating 6AQ5 with heaters tied to ground, you have to keep the cathode below 200V. This will not be true in a 400V B+ setting with a floating 'upper' 6AQ5 because on alternate push-pull swings the plate will swing up to nearly 2x B+. So a floating upper 6AQ5 in a stacked setup will see 350V or so on its cathode and easily exceed the heater rating. That means you just bought yourself a floating heater supply. Worse, in a push-pull setup, you bought two of them because each one is (typically) swinging at a different signal voltage.

There is that issue of the max plate dissipation and max plate voltage rating too.

A simpler way to do the voltage-divider is to use a power MOSFET instead of a stacked 'upper' 6AQ5. This doesn't need a floating heater. And it's working as a source follower, which is both high speed and low impedance. Bipolar darlingtons might work, but MOSFETs are just easier for this situation. Not to mention cheap for what they do.

Since you don't get higher current flow, only higher voltage swing, the output trannie ratio has to change to higher impedance (that is, higher voltage at the same current) for the plates.

It might be simpler to use a power MOSFET to just series-regulate from 400V-ish B+ down to what the 6AQ5 needs normally.

But yeah, it's possible if you dig through the details, have a good supply of 'victim' 6AQ5s for development, and keep a fire extinguisher handy. 8-)

BTW, the 6AQ5 is my favorite unknown power tube. They sound great, and could be had for $0.50 each in quantity back when I was messing with them. I have something like half a cubic foot of them in the back stock in the garage.

[mooreamps]

I started trying to imagine a cascode made of 6AQ5s, which are like mini 6V6s. But somehow it turned into a jet-assisted Yugo.

Another problem would be heater supplies for the top tubes.

No problem. Just use a separate little power transformer for the heater power for those top tubes..

-g

[mooreamps]

how? learn me.

Maximum plate voltage is the measured plate voltage with respect to it's cathode, not to ground.

-g