Megatron Electrostatic Headphone Amplifier

[sorenb]

1 minute ago, JoaMat said:

Emission Labs 20B and 300B, is that the 4 pins socket or the octal socket with center tap filament?

20B octals
300b UX4

I got inspired from your Octal/UX4 pattern ;o)

[JoaMat]

Great work - more Sort Guld.

[Faust]

-resolved-

[Justaguy]

I'm looking at the 300V power supply for the Megatron XL (megatronpowersupply300v2xl in the drive) and it looks like pin 5 on the pcb has an unlabeled component connected to the +300V line. Looking at the schematic (megatronps300.pdf and megatronschem2.pdf in the drive) have a 22 ohm R27 connecting pin 5 to ground.

I'm very much a novice, but I don't think connecting a heater pin to the output line is right.

[jamesmking]

41 minutes ago, Justaguy said:

I'm looking at the 300V power supply for the Megatron XL (megatronpowersupply300v2xl in the drive) and it looks like pin 5 on the pcb has an unlabeled component connected to the +300V line. Looking at the schematic (megatronps300.pdf and megatronschem2.pdf in the drive) have a 22 ohm R27 connecting pin 5 to ground.

I'm very much a novice, but I don't think connecting a heater pin to the output line is right.

22ohm to ground on the heater/filament looks like its there to stop the heater floating and might reduce hum and is therefore probably optional. It looks like this resistor is not present in the megatronpowersupply300v2xl gerbers. It also looks like there is a 221 ohm resistor between the anode and G2 of the el84 in the gerbers which is not present in the megatronps300.pdf.  I am afraid I did not build this power supply so I can't comment on which is the more up to date or preferred version. But in general its the Gerbers which people get pcbs made from and therefore is assumed to be correctly functioning unless the threads on headcase say otherwise.

[Justaguy]

On 3/24/2023 at 12:53 PM, kevin gilmore said:

Power supply #3   tube rectified lc filtered 4 x 6l6gc fully regulated voltages. 300 volt tube rectified and 6bq5 regulated power supply. solid state regulated bias

Referring to my last post, I saw the images of the Felista audio repair and that region of the board is unpopulated.

So another question:
Is there a board for Power supply #3? (reference quote above)
megatronpowersupplyxl has the gz34 tubes and the LC filter, but no 6l6gc tubes. The 300V supply has a el84, which is the 6bq5 quoted. This theoretical full tube build would be 3 on the 300V PSU,  6(?) on the 400V supply and 12 on the amp? That would certainly be something to look at.

[Shawn]

On 3/24/2023 at 10:53 AM, kevin gilmore said:

so i posted this elsewhere because people really want to spend large amounts of money on an electrostatic amplifier that uses a pile of 300b.

this is the first of two.

Megatron xl  with 4 x 300b

Megatron xxl with 8 x 300b

Power supply #1   goldenreferencehv all solid state

Power supply #2   tube rectified lc filtered no regulation solid state regulated bias.

Power supply #3   tube rectified lc filtered 4 x 6l6gc fully regulated voltages. 300 volt tube rectified and 6bq5 regulated power supply. solid state regulated bias

high voltage rectifier tubes are gz34 (5ar4)  300v rectifier tube is ez81

 

I figure a megatronxxl with power supply #3 should come in at a retail built price of $40k

evidently one or more companies are making dc supplies that slam 7 amps into cold 300b resulting in tubes that don't last very long. so this will have a slow start quad 5v power supply board. the xxl will have 2 of these boards.

amp chassis will be about 14 x 14 x 4 inches. power supply box should be the same size.

 

the spice simulations indicate that minimum distortion is set at about 33ma, so this is likely to eat about 300 watts total power consumption.  distortion will end up about .01% at full voltage swing.

 

Based on the megatronxl layout, the second 100k resistor on the top left is float, any reasons for that?

And what is the purpose of those 10K trimmers? I assume they are for grid bias.

[kevin gilmore]

going to look at it in a few minutes. should be ground i think.

 

nope, thats the feedback on the output.  and the gerber file is correct.

not sure where you got that picture, probably here, and likely not done.

3 of the other versions had the same error, all are updated on the google shares.

no idea how that happened.

 

someone needs to go thru all 9 versions (or more) to see if there are any more errors.

i'm kind of busy waxing my brand new 2025 corvette Z06 3lz htc.

Edited January 22 by kevin gilmore

[Shawn]

9 hours ago, kevin gilmore said:

going to look at it in a few minutes. should be ground i think.

 

nope, thats the feedback on the output.  and the gerber file is correct.

not sure where you got that picture, probably here, and likely not done.

3 of the other versions had the same error, all are updated on the google shares.

no idea how that happened.

 

someone needs to go thru all 9 versions (or more) to see if there are any more errors.

i'm kind of busy waxing my brand new 2025 corvette Z06 3lz htc.

The picture is from your earlier post. I will go through other files to check for any errors this weekend.

Congrats for your new driving toy. I remember the Z06 comes with a flat-crank engine which sounds insane. Does it have bucket seats?😍

[robm321]

13 hours ago, kevin gilmore said:

i'm kind of busy waxing my brand new 2025 corvette Z06 3lz htc.

💪 Congrats!

[psyduck]

@kevin gilmore Is MJW18020G a good substitute for 2SA1968?

MJW18020G

 

[kevin gilmore]

definitely not.

use ksa1156.

not sure if its the same pinout.

kind of busy at the moment. but you would not know that

 

[spritzer]

It's an NPN so no... not a replacement.  There are no PNP parts that are past 600V these days. 

[psyduck]

Isn't the 2SA1968 an NPN? https://www.alldatasheet.com/datasheet-pdf/view/38672/SANYO/2SA1968.html

[kevin gilmore]

the 2sa1968 is definitely a pnp.  there were datasheets that were definitely wrong.

for this application you need a 400v pnp.

[Craig Sawyers]

The bizarre thing with the Sanyo datasheet in the subject line says NPN, but the actual data is all negative going - which would only be the case for a pnp.

Of course all the AI data scrapers only consider the (incorrect) Sanyo tag line of NPN - and that forms the basis for every human on the web to propagate that nonsense.

Or the other way round - it matters not. As Kevin says - the 2sa1968 is a pnp

[spritzer]

Yeah and it is 2SA... that is always a PNP. 

[Craig Sawyers]

13 hours ago, kevin gilmore said:

definitely not.

use ksa1156.

not sure if its the same pinout.

kind of busy at the moment. but you would not know that

 

KSA1156 is of course obsolete, alas. OnSemi seem hell bent on obsoleting anything remotely useful from the inventory of companies they acquired. If it does not go in a mobile phone, flat screen TV, PC or car or anything with huge volume, they lose interest real fast.

Any that are on eBay are >90% fakes

[Craig Sawyers]

Any thoughts on the PZTA96ST1G ? Surface mount SOT223, PNP, 450V and 500mA. Downside is 1.5W power dissipation.

At least the darned thing is still available

[Craig Sawyers]

Or even the DXTP560BP5 . PNP, 500V, 150mA and 2.8W

Surface mount of course....

[JoaMat]

Another alternative is STN9360, 600V, 0.5A, 1.5W in sot-223. Has been used in Megatron before (less than 100mW). Solder it in standing instead of ksa1156.

[Craig Sawyers]

That is an impressive device!

And cheap as chips at UKP 0.38 (although it is very similar in USD or Euro now) from Mouser in one off.

Edited February 12 by Craig Sawyers

[psyduck]

I was also thinking of the STN9360 in an adapter board.

[Shawn]

 

38 minutes ago, psyduck said:

I was also thinking of the STN9360 in an adapter board.

The STN9360 has no problem in megatron build. Make sure to check the pin out as I remember that the adapter board needed to be rotated back.

[JoaMat]

Below STN9360 replacing 2sa1468 on a Kevin DIY T2 board. I bent outer pins slightly outwards to fit into the to220 pads.