Megatron Electrostatic Headphone Amplifier

[steefdebruijn]

I answer based on the little knowledge I have.
1. Yes 
2. You need lower resistor a bit 
3. Yes
4. offset is measured with each channel with respect to ground. I'm not sure what you're doing (I think you're measuring the balance)
5. I use a large heatsink. maybe it would be convenient for you to think about that when you do your chasis. 
First check the offset, if it is low and the Bias is correct you should not have problems connecting your stax.
 

Thanks for the quick response. I guess I should lower the resistor to about 17k5 then if I calculate correctly.

I measured the DC on the output pins per channel. Should I measure each pin against gnd?

The LV heatsink stands tall on the PCB so can’t bolt on chassis.

Steef


Verzonden vanaf mijn iPhone met Tapatalk

[jose]

5 minutes ago, steefdebruijn said:

 

I measured the DC on the output pins per channel. Should I measure each pin against gnd?

The LV heatsink stands tall on the PCB so can’t bolt on chassis.

 

 

Yep,  each channel with ground. 

Sorry I got confused (HV vs LV) I used a large headsink on my Square maybe you can used this  (mouser reference 588-EV-T220-64E ) on LV  

 

[steefdebruijn]

Yep,  each channel with ground. 
Sorry I got confused (HV vs LV) I used a large headsink on my Square maybe you can used this  (mouser reference 588-EV-T220-64E ) on LV  
 

I used 588-EA-T220-64E. At 8c/w vs your 10c/w but should I worry about the 60 deg celcius? I guess I should use a lower sec winding voltage transformer then. Maybe 12-13vac or so?

To be sure what I am supposed to measure on outputs. As far as I understand the outputs are balanced outputs. So a “+” and a “-“ pin per channel - no gnd output. I measured dc between these pins. If I measure pins to gnd I get 2 measures per channel.


Verzonden vanaf mijn iPhone met Tapatalk

[jose]

Here when you measure both pins (same channel) you are measure the balance between - and +. If you want to measure offset you need to check each output with ground reference.

Of course Maybe I'm wrong but if you have dudes about this you can use the search oprtion and read some many interesting treath about electrostatic amps.

Are you using the LV for the heating filaments?

[steefdebruijn]

Thanks Jose, I learn a lot by reading a lot but it takes time to sink in ?

DC offset one channel -23vDC (shifted to -40 after a while) and +1vDC, other channel -52vDC (also drifting to -35) and -13 vDC, this results in balances of 23-40vDC for both channels.

I read here that too low 300v line can cause things to oscillate. So maybe I should lower the drop resistor to about 18K first before measuring again?

I did not measure my tubes myself but did buy matched stuff from banzai-music in Germany. So I really want to rule out improper matched tubes ?

About the LV I did see that mw used a 12v secondary supply for his recent build so I think I should go that route. A bigger heatsink on the LV will not fit on the board.

Yes I use the LV for the 12Ax7 tubes (x as in X and U ? )

Thanks for the help

Steef

[jose]

I think the offset depends more of the well matched tubes (in my case it was like this). This amplifier needs a very well matched octet and sorry but I think your valves are very badly matched.

Regardless of this, you have to raise the voltage to 300v. I was working for a while with 290 volt but I finally increased it to 300v as it should be.

I use a separate PSU for the heating filaments.

I recommend that you carefully read the thread (especially the last post) so that you familiarize yourself with all the information there is about it.

 

[steefdebruijn]

Thanks for all the info and pointers. Will investigate further and make some changes. I'll be back ? 

[steefdebruijn]

Question: Is it a good or a bad idea to take the 7812 off board to a bottom-of-case-mounted 532-529802B25G (at 3.7c/w) for example? (assuming the 7812 produces the most heat of the 7812/7912 pair). The 7912 stays onboard on the 588-EA-T220-64E in this setup.

Question: should I overshoot or undershoot the 300vDC? If I calculate correctly I get 292vDC with 18k drop resistor (too low according to Jose 's findings) and 310vDC with 16k drop resistor. Or should I parallel the current 21k resistor with about 100k to get 17k36 drop resulting in about 298vDC? [edit] combining a 40k+30k gives me 17k14 resulting in 300vDC so I think I should go for that option.

Update: I consulted with a local tube seller for testing/matching my tubes to create a box of 12 good-matched tubes.

Edited November 14, 2018 by steefdebruijn
added 30+40k option to get to 300vDC

[jose]

Used a separated PSU (with a LM7912, LM317 or LM350 for example) If you're careful, you can use the floor of the box as a heatsink.

For 300vDC, I used parallel resistors. I  think it's even better than using only one resistance due to the heat dissipating.

Edited November 14, 2018 by jose

[mwl168]

Congrats on a successful build!

You can experiment by moving the 6CA7 tubes around to lower the balance and offset. On my Megatron, I use a matched quad of Tesla EL34 for the output and a matched quad of new production Svetlana EL34 for the CCS that were purchased from different vendors. I was able to get the offset and balance to within 2V.  

Edited November 14, 2018 by mwl168

[steefdebruijn]

nice trick @mwl168 thanks

[mwl168]

About the LV supply for the 12.6V filaments, I agree the 15v secondaries are too high. The drop-out voltage of the 7812 is 2.5V according to the datasheet. So a solid 12 vac secondaries should do the job and spare the regulator from having to drop lots of voltage.

In my build, I did end up using a 12 vac trafo with higher current rating which allows me to use tubes like E80CC that draws twice the filament current (300mA) compared to standard 12AU7. I had to find a 7812 regulator that's rated for 1.5A. I monitored the heatsink temperature on the 7812 and it never got hot to the touch. Sorry, don't remember the exact height of the heatsink and its temperature now but your 64mm height one looks right to me.    

EDIT: By the way, if you decide to get new regulator with higher current rating, I suggest you buy a few so you can pick the one that gives higher output voltage. These 7812 regulators typically have about a 5% tolerance. If you are unlucky you may get one that outputs 11.5VDC or so which would annoy me  

Edited November 14, 2018 by mwl168

[steefdebruijn]

By the way, if you decide to get new regulator with higher current rating 

I already have the NJM7[89]12FA which have the 1,5A current rating. Outputs are +-12.06 vDC so I am good. Will buy a 50VA 2x12V trafo and try again.


Verzonden vanaf mijn iPhone met Tapatalk

[steefdebruijn]

Pity, local tube seller is not that helpful. Pondering about another DIY project to help myself out: http://www.dos4ever.com/uTracer3/uTracer3_pag0.html Is it up to the task?

Is it helpful to power up the amp without tubes to measure voltages at the pins to verify that all tubes (output vs CSS of course) get the same input?

[jose]

Why don´t you buy a good matched quartet and tests? You have several websites with good prices for JJ EL34

[steefdebruijn]

Why don´t you buy a good matched quartet and tests? You have several websites with good prices for JJ EL34

I thought I did that already but seems not the case. I deliberately avoided vague eBay sellers and went for a seemingly reputable seller (not through eBay) in Germany. And maybe the tubes *are* okay and something else is wrong with my build. That is why I seek for alternative measurements on the tube side as well as on the amp board side. Will ponder about buying extra tubes though. Any suggestions for sellers in Europe?


Verzonden vanaf mijn iPhone met Tapatalk

[mwl168]

Have you tried re-arranging the EL34 tubes combination to lower the offset (measured from output to ground)? It'll take a few tries but if you get the offset down, the balance will be down as well. 

The amp is capacitor coupled and there aren't that many parts in the output/CCS section, unless wrong value or defective parts were stuffed, I would look into the tube matching for the cause of high offset.

I would take notes through the process of rearranging the EL34 combination and pay attention to see if one of the outputs or one of the EL34 tubes constantly shows or contributes to high offset.  

    

 

Edited November 16, 2018 by mwl168

[steefdebruijn]

Hi @mwl168 will do but first I try to get the 300v line closer to 300v and lower the lv heatsink temp. 

[jose]

Ok guys it starts to get cold. it's time to light the stove…

 

 

[Blueman2]

7 hours ago, jose said:

Ok guys it starts to get cold. it's time to light the stove…

 

 

Essentially 100% of the wattage that goes into that amp will turn into heat.  So just as efficient as a small space heater.  But far more wonderful to look at, not to mention listen to.  We should get carbon offset credits for creating wonderful music while heating our homes.    

Edited December 5, 2018 by Blueman2

[jose]

You should do one.  

I do not know if I have mentioned it before but I have remained definitively with the E180CC as a pre-valve (it´s somewhere between the ECC82 and the ECC81). I have not noticed any "negative" effects when adding that extra gain in the first stage. The E80CC also works very well here but is more in line with the ECC82 / 12AU7 with a little more gain.

I think that this change and the adjustments that I did in the circuit (thanks to Michael, Jim and Kevin) have left the amp in its final configuration.

 

 

[sorenb]

 

Finally got around to finalize my Megatron, fitted into one chassis with GRHV. 
 

 

Edited January 14, 2019 by sorenb

[chinsettawong]

Wow!  Very, very nice!

[steefdebruijn]

What a beauty!

?

[Pars]

Really nice Søren!


Sent from my iPhone using Tapatalk