Megatron Electrostatic Headphone Amplifier

[Shawn]

Quick update:

it’s 6 AM here in California and I just wrapped up some late-night tweaking. Someone PM’d me earlier asking how I solved the offset and balance issue, and I can finally say: done. Offset is now around 0.3V and balance is about 0.2V.

If anyone else is running into similar issues, here’s what worked for me:

1.CCS Set Resistor: I kept the CCS set resistor at 1.5k, which gives around 22–23mA static current. You can try lowering to 1k if you want around 25mA, but I found 1.5k sufficient. The more plate current you have, the higher the offset you will get.

2.Cathode Resistor: Changed from 3300Ω to 7k–7.5k. A larger cathode resistor gives you a deeper grid voltage. I could only find suitable 7k wirewound resistors on Mouser – they’re a bit bulky but still fit. Note: Larger values will increase the offset, so some trial and error is expected.

3. Grid Pot: I replaced the 10k pot with a 50k, which gave me a much wider range to trim both offset and balance.

Final bias: around 24–25mA static current, Vgk ≈ -75V.

Hopefully, this helps anyone else dialing in the Megatron XL  and happy tweaking!😊

[JoaMat]

Glad you find a way to adjust the offset.

With EL34 as CCS, +400V and offset close to zero, I believe 1.5K gives ~23mA and 1.0K gives ~34mA.

Edited July 20 by JoaMat

[Shawn]

12 hours ago, JoaMat said:

With EL34 as CCS, +400V and offset close to zero, I believe 1.5K gives ~23mA and 1.0K gives ~34mA.

Thanks for the correction.

I actually paralleled a 3.3k resistor with the existing 1.5k CCS set resistor, which gives an effective resistance of around 1031Ω. Based on the measured voltage drop, that gives me roughly 24–25mA of current.

I believe the difference might come from my cathode resistor and grid voltage setup — possibly the cathode current is already limited before reaching the CCS. That might be why the current didn’t rise as much as expected even with the lower set resistor.

[JoaMat]

After paralleling with a 3.3K resistor an increase of a few mA seems reasonable, but what happened to the offset?

[Shawn]

8 hours ago, JoaMat said:

After paralleling with a 3.3K resistor an increase of a few mA seems reasonable, but what happened to the offset?

Paralleling a 3.3k resistor with the original 1.5k gave an effective resistance close to 1k, which I expected to behave similarly to using a single 1k resistor. After doing that, the offset increased significantly from around +12V to about +60V. So in the end, I reverted back to the original 1.5k setting resistor.

[JoaMat]

Thanks @Shawn.

I have ”gathered” some info from three sets of 300B, one 2A3 and one 6A3. To me it seems that they behave quite similar and need about 7K – 7.5K cathode resistors.

Your 50K trimmer (Grid Pot) will probably make it easier with the “fine tuning”. I guess you have put trimmers on the tube side of the board for easier access.

[Shawn]

16 hours ago, JoaMat said:

Your 50K trimmer (Grid Pot) will probably make it easier with the “fine tuning”. I guess you have put trimmers on the tube side of the board for easier access.

I actually kept the trimmers on the back side of the board. The PCB is mounted upside down near the top of the chassis, and I didn’t drill dedicated holes in the top cover for adjusting the trimmers. So each time I need to fine-tune them, I just tilt the whole chassis on its side and yes a bit inconvenient, but still manageable. A better solution would definitely be to make access holes in the top panel.

[JoaMat]

New solid state CCS for Megatron.

Now with a third pin for ground. The pins are made from 2.5 mm brass rod, which is almost 0.10 mm thicker than tube pins. PCB is a CNC milled one side 1.5 mm copper clad. So, no copper on top side of PCB. Holes for the three pins are 1.0 mm deep and you have 0.5 mm epoxy laminates left.

Transistor pins are protected with heat shrink tube and the exposed metal tab is protected with a 3D printed cover. The screw holding transistor to heat sink is electrical connected to it and to the ground pin.

The white base with guide pin is made of three 3D printed parts.

Edited yesterday at 09:10 AM by JoaMat

[simmconn]

Interesting design! I’d be a bit worried about the thermal and mechanical side of things. The thermal resistance of the heatsink this size is probably going to run above 5C/W without forced air flow. It is going to run too hot to touch at 350V to 400V/20mA (7 to 8W). With 3C/W theta JC on the 10M90S, there is still margin on the die temperature.

If I were to design a PCB to be plugged into a tube socket with soldered pins, I’d use slightly undersized pins and whatever copper I can get for adhesion to the board (double layered board with plated-thought holes in 2-mm or thicker PCB), especially if the counterpart is the unforgiving “modern” tube sockets with tulip type spring contacts.

Last but not least, when the whole thing gets very warm, I’d use materials with higher working temperature such as ASA or ABS.  PLA is probably not going to last very long.

[Shawn]

23 hours ago, JoaMat said:

Now with a third pin for ground. The pins are made from 2.5 mm brass rod, which is almost 0.10 mm thicker than tube pins. PCB is a CNC milled one side 1.5 mm copper clad. So, no copper on top side of PCB. Holes for the three pins are 1.0 mm deep and you have 0.5 mm epoxy laminates left.

About the pin choice. I think this one might be closer to the original tube pins:

MPN:6835-0-00-15-00-00-44-0

https://www.mouser.com/datasheet/2/273/MMMC_S_A0004813843_1-2556643.pdf

I haven’t ordered it yet to confirm, but maybe it helps you.

Also, I’d like to get your advice:

At 400 V / 22 mA, would you consider setting Vgk around –95 V a good operating point?
I’m also thinking about increasing the static current to around 27–30 mA. From what I see, the EL34 max plate dissipation is ~25 W. At 400 V / 30 mA, that’s only about 12 W, so it looks safe on paper. I’d love to hear your thoughts on this.

[JoaMat]

16 hours ago, simmconn said:

Interesting design! I’d be a bit worried about the thermal and mechanical side of things. The thermal resistance of the heatsink this size is probably going to run above 5C/W without forced air flow. It is going to run too hot to touch at 350V to 400V/20mA (7 to 8W). With 3C/W theta JC on the 10M90S, there is still margin on the die temperature.

If I were to design a PCB to be plugged into a tube socket with soldered pins, I’d use slightly undersized pins and whatever copper I can get for adhesion to the board (double layered board with plated-thought holes in 2-mm or thicker PCB), especially if the counterpart is the unforgiving “modern” tube sockets with tulip type spring contacts.

Last but not least, when the whole thing gets very warm, I’d use materials with higher working temperature such as ASA or ABS.  PLA is probably not going to last very long.

Thanks for your comments, @simmconn.

The heatsink gets warm and user is aware of it. I think he is running at 400V/22mA and so far, it seems to work. If it gets too hot than he can’t use this thing - that’s life.

For pin dimension. The first version used Neutrik XLR3 pins. They were perfect, diameter slightly undersized. But to short to be soldered direct to main PCB. When looking for brass rod the best I found was 2.5 mm so, I decided to try that. 2.5 mm works with my Teflon tube sockets. The user is informed of that pins are soldered only on one side and if pins come loose, we have to find a better solution or perhaps abandon this thing.

I use Addnorth HT-PLA Pro filament. Better heat performance than PLA but not as good as ABS. I’m confident the socket isn’t a problem. It might be a problem with the transistor tab cover. I haven’t thought about the cover until your post - thank you simmconn. Now the user is informed, and he promised to keep an eye on the cover.

[JoaMat]

6 hours ago, Shawn said:

About the pin choice. I think this one might be closer to the original tube pins:

MPN:6835-0-00-15-00-00-44-0

https://www.mouser.com/datasheet/2/273/MMMC_S_A0004813843_1-2556643.pdf

I haven’t ordered it yet to confirm, but maybe it helps you.

Also, I’d like to get your advice:

At 400 V / 22 mA, would you consider setting Vgk around –95 V a good operating point?
I’m also thinking about increasing the static current to around 27–30 mA. From what I see, the EL34 max plate dissipation is ~25 W. At 400 V / 30 mA, that’s only about 12 W, so it looks safe on paper. I’d love to hear your thoughts on this.

Thanks for the pin tip.

This beyond my understanding. But as I see it, with 400V/22mA the Vgk will be what it will be with offset zero.

12 W with EL34 shouldn’t be much of a problem. My EL61 will also handle that, I hope.