5881 single ended poweramp (powered with 12V SMPS)

[jhjh217]

This was my recent project.
My first project was this preamp project:
a SMPS powered tube preamp(Dumble SSS#002), mountable on a pedalboard. and it was a great success :
https://ampgarage.com/forum/viewtopic.php?t=36345

Then my second thought was "how about a power amp on a same form factor?"

20230810-IMG_0947.jpg

It's powered by a 12v DC power supply, then it boosts the voltages up to around 360 volts and feeds the power tubes.
(converter I used: https://www.ebay.com/itm/354131738261?_ ... %3A2047675
I switched some component on the original PCB though. I had to replace the mosfets to much higher quality ones because of heating issue.)

Actually making this took me about 2weeks, but it had some really bad noise issues..
So further debugging took me about 2 months, and it's finally finished!

I'm currently filming a video about this project, also I'll upload a sound demo video using this power amp paired with my Dumble SSS preamp. I'll upload it by next week

It doesn't have 60cycle hum, it doesn't have heater hum, it's a dead quiet power amp

This is the schematics:

5881SE Power Amp Schematics.png

It's biased at 94% of max dissipation. (JJ 5881 has 23W max dissipation)
If I want different bias settings, there's no need to change the cathode bias resistors when biasing, even though this is a cathode biased amp! I can just adjust the B+ voltage by turning the trim pot at the boost converter! So very precise biasing and very precise B+ adjustments are possible. Also switching to different tubes is much easier. Bias resistor stays the same, only changing the B+ voltage through the trim pot will do it.

I have one question about screen voltages though.
Even though I'm using a 1K screen grid stopper resistor, screen voltage measures 361v, which is 4.5v higher than the plate voltage.
(I didn't take into account of the resistance of the primary windings when planning out in the first place..)
And I don't really want to raise the screen grid resistor value any higher than that.
It does seem to work flawlessly though. But 4 volts higher screen voltage is okay right..?

(In the pictures, the tubes seems to glow very red and hot, but that's because it's photographed with long exposure settings in a dark lighting, so it's not red plating or anything. Just photographed that way.)

[bepone]

a lot of svitchers (and generated noise) for that simple circuit

[jhjh217]

a lot of svitchers (and generated noise) for that simple circuit

Actually it wasn't the circuit, just bad grounding. Solved it by simply moving some ground point, which took me 2 months to figure out.

[Stevem]

The tubes screen voltage applied to it is actually one of the main factors in how a tube is biased.

[bcmatt]

Wow! Cool and creative project. I look forward to checking out that video!

[jhjh217]

It's been a long time since my last upload..
I've been busy lately..
But, finally here's the video! with a sound demo.

It's paired with Dumble SSS preamp, with celestion vintage 30 speaker.


https://youtu.be/ZmnO8MMYe9k

Thanks for watching!

[Stevem]

Just how are you getting 360 volts out of a dc to dc
Converter that’s rated at a max of 250 volts?

Here’s some info on running 5881 in class A like you are.

Your screen voltage as you can see is far too high.

[bepone]

Your screen voltage as you can see is far too high.

his voltage is far too high? how did you arrive to this conclusion?
what about fender?
which tubes are used in bassman 5F6A, and what is the voltage on g2 ?

[Stevem]

Did you happen to notice he is running only one 5881?

Would that not have to be in Class A then?

Look at the specs I posted where class A on the screen should be a max voltage of 250.

[bepone]

@stevem
Class A doesnt change anything?

And why you believe that 250V from datasheet cannot be exceeded? because in the datasheet is stated this number? Did you ask yourself a question who put this data in datasheet and why?

From your responses around looks like you know what are you talking about, so, you should know allready that numbers in datasheets are stated coservative, for long and reliable run, but this doesnt mean that you cannot run over that voltage, and by experience from last 70 years knowing that still will be reliable.

[bepone]

It's biased at 94% of max dissipation. (JJ 5881 has 23W max dissipation)

to be ontopic, the best for SE use is to keep the dissipation lower, 70%, (i wouldnt even go to 80%), not 95% - you will lost 1W, your ear cannot hear that, but your tube will be grateful and life extened.

[Colossal]

It's biased at 94% of max dissipation. (JJ 5881 has 23W max dissipation)

to be ontopic, the best for SE use is to keep the dissipation lower, 70%, (i wouldnt even go to 80%), not 95% - you will lost 1W, your ear cannot hear that, but your tube will be grateful and life extened.

I have found lower dissipation in SE to be better sounding as well.

[jhjh217]

Just how are you getting 360 volts out of a dc to dc
Converter that’s rated at a max of 250 volts?

Here’s some info on running 5881 in class A like you are.

Your screen voltage as you can see is far too high.

Hmm. interesting.. When I ordered that converter, it clearly stated around 400v-450v max. but now it listed at 250v max? Image seems to show the exact same converter that I bought 3 months ago though. I'm not sure why the max voltage is edited to 250v..?
Boost converter is capable of putting out around 420V-450V when I tested.
But the problem was that the mosfets got really hot. It seemed almost impossible to put out 70W(their stated max wattage) at around 400v without burning those mosfets. (without sufficient heatsink).
I think that's why they edited the max voltage? to a more conservative level?

I actually switched mosfets to more higher quality ones. (Mosfets with lower RDS(on), lower gate charge, and input capacitance) and the heat issue was dramatically improved.

mosfets.jpg

I guess when using that boost converter, 70w at 250v is the safe limit (without additional heatsink or switching out different components).
But with better components(mosfets especially) and better heatsink, I guess max voltage can go up to 400V..?

and for the screen voltage, the data sheet surely states screen voltage of 250V, but that is not the 'max' screen voltage. I always thought that 250v was just picked to provide an example of a typical operation, and plate characteristics graph. For other data sheets (pdf attatchment, though it's 6l6gc not 5881) it provides several graphs with screen voltage of 250v 'and' 400v. So I guess you can choose whatever screen voltage you want?

Fender champ amps (5f1 etc.) uses sinlgle ended poweramp section and they don't even have a screen resistor. Their screen voltage is higher than their plate voltage..!
And majority of other fender amps uses 470ohm screen resistors. which will also have very high screen voltages.
To lower the quiescent screen voltage to 250v on my amp, I'm going to have to use a 50K ohm screen resistor..!! No guitar amp uses such a large screen resistor as I know of.
It can be seen on hi-fi amps though.
I don't really know the real logic behind this, but all guitar amps seem to do it that way..

[jhjh217]

Link:
https://el34world.com/Forum/index.php?topic=29132.0

This was a very interesting topic about screen currents and screen resistors. I thinks it's worth checking out!

[jhjh217]

It's biased at 94% of max dissipation. (JJ 5881 has 23W max dissipation)

to be ontopic, the best for SE use is to keep the dissipation lower, 70%, (i wouldnt even go to 80%), not 95% - you will lost 1W, your ear cannot hear that, but your tube will be grateful and life extened.

It's biased at 94% of max dissipation. (JJ 5881 has 23W max dissipation)

to be ontopic, the best for SE use is to keep the dissipation lower, 70%, (i wouldnt even go to 80%), not 95% - you will lost 1W, your ear cannot hear that, but your tube will be grateful and life extened.

I have found lower dissipation in SE to be better sounding as well.

Oh interesting. All I knew was that SE should be okay at near max dissipation, and never considered lowering it..
I might try lowering the dissipation to 70~80% and see how it sounds. I guess it shouldn't sound that different, while definitely better for tube life. Thanks