Does it exist a class D tube amp ?

[kagliostro]

Do you know if a class D tube amp exist ?

There are schematics about such a configuration?

Thanks

Franco

[DummyLoad]

Do you know if a class D tube amp exist ?

There are schematics about such a configuration?

Thanks

Franco

Why? Nothing but waste of heat and space and unnecessary complexity with inherently unstable elements. Maybe for the input stages, yes, but really, what are the benefits? 

--Pete

[kagliostro]

Ciao Pete

I was thinking it will stress less the power tube

and just curious about

Franco

[PRR]

Transistors have been described as "ideal switches". They aren't ideal, but they are very good.

Tubes are terrible switches.

A tube class D is possible but much complexity for little or no advantage.

[kagliostro]

Thanks PRR

I believe, it was only a curiousity

born thinking to the class A SE amps and the high heat they dissipate

Obviously there are class we can use in our amps that are more suitable to stress less the power tubes

Franco

[shooter]

don't know about Class-D but in Class-C we would hammer a 500W tube to make 25KW, now that's stress'n a tube.
my understanding of tubes, as long as the "average dissipation" stays within the design specs, the tube will live a "normal life cycle" regardless of operating Class.

[kagliostro]

Ciao Shooter

The SE amp to wich I refer uses a pair of tubes (it is a PSE) with an output of around 36/37 W

And also if they work within spec a lot of heat is transferred to the chassis

So the curiousity about the class D

Franco

[shooter]

i've built may SE and PSE, and yes they do get warm, I use them to heat my little workshop this time of year, 100% efficient heater 

[kagliostro]

It is like to have a tube organ amp in less space

😁

Franco

[pdf64]

Ciao Shooter

The SE amp to wich I refer uses a pair of tubes (it is a PSE) with an output of around 36/37 W

And also if they work within spec a lot of heat is transferred to the chassis

So the curiousity about the class D

Franco

Would it be vastly simpler to change it to push pull AB? That would enable the same pair of output valves to idle cooler than they have to in parallel single ended.

[kagliostro]

Ciao pdf64

Yes, we know, but it was planned as PSE to obtain what a PSE amp can give

Franco

[HotBluePlates]

Do you know if a class D tube amp exist ?

There are schematics about such a configuration?

A few years ago, I found schematics somewhere for a Class D tube amp.  So yes, it can be done.

Once you understand how "Class D" works then you'll understand PRR's comment that "tubes are terrible switches, though transistors can be good switches."

The SE amp to wich I refer uses a pair of tubes (it is a PSE) with an output of around 36/37 W

And also if they work within spec a lot of heat is transferred to the chassis

So the curiousity about the class D

... it was planned as PSE to obtain what a PSE amp can give

Forget about "Class D" for a minute.  Instead, look at something like a Fryette Power Station that is a "re-amper."

   -  The main amplifier had a reactive-load/attenuator, to reduce the speaker output to "about zero."
   -  There is a Class AB tube amp after the reactive load to bring the speaker output up to anything between "zero" and "50w" or "100w" (depending on the model of Power Station).

We can't "bias into Class D."  It's a change to the fundamental way an amplifier amplifies.  And there's a bunch of support-circuitry that goes with the "different way to amplify."

[blackcorvo]

The way Class D works in a nutshell is, you switch the power device (transistor, IC, what have you) on-and-off real fast (ideally so fast it's way beyond the audio range), so it passes as much current to the load as possible on that small timeframe while it's on, and the average of time between on and off makes the device not heat up as much as it would when being fully on.
It's kinda like if the sampling technique we use for digital audio, but each sample is a pulse going to a power device to reproduce it louder (but it's not digital because there's no conversion into digital information on this process, despite a lotta people calling them Digital amplifiers...).


I just don't know if any possible gains in performance could justify the increase in complexity of the circuit. Tube amps are pretty straight forward and easy to grasp how they work...


but hey, if you REALLY wanna have a go, maybe try something like Usagi's OTL amp "but make it class D" for a proof of concept. I mention the OTL approach because sticking with Output Transformers and high-frequency switching will have The Fairy of Filter Calculus pull your leg at night in your sleep, and the switching stuff will be more than enough in one's plate to deal with.

[shooter]

The Fairy of Filter Calculus pull your leg at night in your sleep

more like the psycho-math instructor wielding a bloody axe 

[kagliostro]

Hi Friends I apologize for replying only now, yesterday was a NO day (unfortunately)

@ HotBluePlates

A few years ago, I found schematics somewhere for a Class D tube amp

from the way you say it I understand that they are lost and cannot be seen, however it would have only been a technical curiosity

Once you understand how "Class D" works then you'll understand PRR's comment that "tubes are terrible switches, though transistors can be good switches."

OK, now I think to class D as a "sort of PWM" Amp

Instead, look at something like a Fryette Power Station that is a "re-amper."

I didn't know the existance of such unity but, to tell it all, when the problem (of heat) was highlighted a thing I tinked to was a Super Scaler (Kevin O'Connor, KOC, on his books discuss about it and he planned a real one), you can read about it on the TUT Vol.4 Chapter 7, TUT Vol.5 Chapter 15 and Chapter 19

On the page of London Power, here, https://londonpower.com/super-scaling/, you can read some explanations (also follow links on that page)

Also on the pages of Tubecad the Super Scaler was discussed

The Fryette Power Station remember me a lot a Super Scaler (and an Herzog, because at the input the signal is lowered)



But this will be a way only to transfer the heat to one other chassis instead to don't produce it

Some people also build a small tube amp (or only preamp) and buy an off the shelf Transistor class D Power Section (tremendously powered) but it is not in the spirit of what was thought of with the PSE we are talking about

We can't "bias into Class D."  It's a change to the fundamental way an amplifier amplifies.  And there's a bunch of support-circuitry that goes with the "different way to amplify."

Now it is clare

@ blackcorvo

The way Class D works in a nutshell is, you switch the power device (transistor, IC, what have you) on-and-off real fast (ideally so fast it's way beyond the audio range), so it passes as much current to the load as possible on that small timeframe while it's on, and the average of time between on and off makes the device not heat up as much as it would when being fully on.

I don't know if is the correct interpretation, but to think to a PWM seems to give me the right understand on what is happening in class D (and to me suffice)

I just don't know if any possible gains in performance could justify the increase in complexity of the circuit.

I agree, the added complexity it doesn't seem at all a problem to be underestimated

@ shooter

You know, I always struggle with math 

THANKS to ALL

Franco


p.s.: BTW I have a distant memory of tubes designed to be used as switches, am I wrong or did they also exist for this use (although perhaps surely not suitable for our case of class D amplification)

May be 5963 Tubes an example ?

[shooter]

I have a distant memory of tubes designed to be used as switches

seems they did ok in computers, at least long enough to get the kinks worked out of transistors, then IC, then...

[kagliostro]

Yes, on the vacuum tubes computers

it seems to me also that there was other kind of on / off tubes used in industrial machinery but don't remember more

(and also if what I remember is real or a fake)

Franco

[PRR]

Gas tubes are good switches up to thousands of volts and hundred of Hertz.
https://en.wikipedia.org/wiki/Thyratron
https://it.wikipedia.org/wiki/Thyratron

[kagliostro]

Ciao PRR

Many Thanks

Interesting reading

Franco