Megatron Electrostatic Headphone Amplifier

[kevin gilmore]

the biggest problem is that the tube sockets are up to .02 inch off in both x and y. really hard to make a circuit board with tube sockets fit into the holes with a top panel that is .35 inch thick. and all the blind holes for the other board mountings were hand tapped and then the holes in the boards were made bigger till it fit. this is so hopelessly fucked. might have to make the holes in the tube sockets much bigger to allow for the slop. No room for a regulated power supply.

do want to know how the dremel tool work was done. also corrosion all over the place when the electrolytics let loose.

at least we finally have an example that is much worse than the singlepower es2. not sure how to continue.

[spritzer]

Those boards might have been milled on one of those cheap table top CNC's...

[kevin gilmore]

joamats hand made boards look much better than that.

posted latest schematic. Does anyone recognize the tube power supply? And the output tube has 300v cathode to filament on a tube rated at 100v. Does (did) it even work?

birgir mentioned this earlier. have a look

Image-Hifi.com-62015-Viva-Audio-Solista.pdf (audiograffiti.com)

so lots of very high voltages, lots of heat and flammable felt and a horrible point to point job. 

But wait, there is more, my good friend ray samuels and his A10. 5687 output tubes in srpp configuration running with a plate voltage of 500v. So at peak output voltages the cathode of the top tube sees 500 volts with respect to the cathode on a tube rated at 100v. How does this even work?

megatronschem2.PDF

Edited September 19, 2021 by kevin gilmore

[spritzer]

I think the main issue that tubes take a crap ton of abuse and are nearly impossible to kill (except with excessive power) so they are the refuge of the hack job. 

[Fitz]

11 hours ago, kevin gilmore said:

at least we finally have an example that is much worse than the singlepower es2.

[mikeymad]

Now I want all of my amps felt lined 🔥🔥😱

[johnwmclean]

8 minutes ago, mikeymad said:

Now I want all of my amps felt lined 🔥🔥😱

💦

[kevin gilmore]

hey joamat wanna make a couple of test boards before real money is spent?

 

[JoaMat]

Send me gerbers and I’ll take a look into it.

[kevin gilmore]

So this is what its going to look like when its done.

And Hammond is willing to do the custom transformers. With a price that matches the complexity.

Anyone with experience with tube rectifier power supplies driving 5 Henry inductors?

Edited September 21, 2021 by kevin gilmore

[Craig Sawyers]

Well candidly, fuck me. I've just dipped into this thread, and I feel the need to barf. I have never seen such a construction nightmare. Handmade dremel boards FFS. Ugh. And that was not the worst by a long chalk.

When I started building amps aged 15-16 I was doing far, far better job than that nightmare. Although an early one is festering in the attic for decades I'll wager it would still power up fine.

Bit of a labor of love Kevin to rebuild the horror show.

[Craig Sawyers]

Are you talking about an LC filter following a tube rectifier, Kevin?

[kevin gilmore]

636v center tapped transformer dual tube diode, 220uf to ground 5H inductor 2 x 220uf to ground.

The turn on thump puts about 700v on the first cap. which is why the caps blow up.  adding 50 ohms after the diodes softens the bump. whats the best way to calculate the required primary voltage for 400v at 100ma dc output.

[Craig Sawyers]

You might end up with a soft start into the first C to protect the rectifier - and clearly the capacitor too!. Sort of timed relay shorting a resistor between rectifier and first C once the rectifier is nice and warm and the capacitor is charged via the limiting R. Or a MOSFET or similar device instead of a relay.

What transformer is needed depends on the resistance of the inductor, the ripple on the first C (~3V p-p), and other resistive losses in the rectifier and winding resistances of the transformer itself. You could lose ~40V all told.

[kevin gilmore]

yep that sounds about right. The resistance of the inductor is 100 ohms.

[kevin gilmore]

boards and transformers ordered.

 

[Craig Sawyers]

What rectifier tube are you using Kevin?

[kevin gilmore]

gz34

[Craig Sawyers]

Maximum for a capacitor input is 60uF. So 220uF significantly exceeds spec for a GZ34, https://frank.pocnet.net/sheets/030/g/GZ34.pdf

Edited September 30, 2021 by Craig Sawyers

[luvdunhill]

On 9/24/2021 at 5:01 AM, kevin gilmore said:

whats the best way to calculate the required primary voltage for 400v at 100ma dc output.

https://www.duncanamps.com/psud2/

[Craig Sawyers]

56 minutes ago, luvdunhill said:

https://www.duncanamps.com/psud2/

Except it doesn't do CLC filters. You can do RC, LC, or just C.

He's just about to issue a new version, which *might* do CLC.

 

OK - he's released it, and it looks like it indeed does CLC. Must have a play.

Edited September 30, 2021 by Craig Sawyers

[Craig Sawyers]

I'm talking rubbish - of course PSU2D can do CLC!

[dsavitsk]

6 hours ago, Craig Sawyers said:

Maximum for a capacitor input is 60uF. So 220uF significantly exceeds spec for a GZ34, https://frank.pocnet.net/sheets/030/g/GZ34.pdf

Modern gz34's can be even more sensitive than the datasheets indicate. You want to keep the cap small. PSUD is your friend here.

On 9/24/2021 at 6:01 AM, kevin gilmore said:

whats the best way to calculate the required primary voltage for 400v at 100ma dc output.

You can change the psud load to constant current to determine this. And you can add a current change to check for ringing in the PS.

[dsavitsk]

example of the above.

[kevin gilmore]

boards actually fit. well shit.