SE GU-50/6CB6 Pentode Design

[ryuji]

Ripped up my old 5 channel 807/6cb6 based amplifier because I need to rebuild into a new rack mountable chassis. Decided to go with a completely new design.

Still havent ordered any parts yet. PSU design was inspired by Peter Millett's DCPP.

This will still be 5 channels just like the old one. About 40 mA per GU-50, probably around 8 mA per 6CB6 to match old amp. Of course ill be playing with bias as i made it adjustable

Just wanted some potential commentary before i ordered the psu parts.

Ill be using different mosfets and of course better choice/fewer zeners to get the required voltage. i ran with the stock parts in the program

 

[davorin]

?
GU50 screen need more voltage, better 300V than 150V!

 

[ryuji]

Hmm, worth a try. But i have doubts that -48v is enough to keep the anode current down. might need to order another/series two 48v supplies together.

Im guessing 300v would be sought after to obtain a flatter curve?

just realized i typo'd out the intended plate current for gu-50's (40 mA)

 

[ryuji]

Found a cheap single module dc-dc converter at digikey that offers -72v. I think that should be plenty of negative to run 300v screen on the gu-50. Runs off 12v too

 

[ryuji]

heres the fully assembled and tested psu board:
12.6v DC and 400v AC input,
500v 300v,150v DC output, no current limit or constraints. full regulation even at open circuit.
6.3v 3A output for operating the small signal pentodes
5x -0v-72v bias pots
5x -0v-5v bias pots
bias supply powered by this guy:
http://www.murata-ps.com/data/power/ncl/kdc_nmt.pdf

 

[807tubenut]

Very nice my friend, any updates?

 

[ryuji]

I am just about done with the chassis wiring. all the major components are in place but its taking a bit of time to carefully solder to the tabs on the plug i chose for inputs/outputs. Slow progress as i have been distracted and busy with work lately and have been working on it in sessions. also need to resolve a issue i had with a resistor.

 

[807tubenut]

Move to resistors rated for 750+v in that problem area, R12?

 

[ryuji]

Amp is completed, but i need a new mosfet for main b+ ripple filter. the one i have there failed due to a failed zener diode

otherwise works exactly as intended. the fixed bias on the preamp works wonderfully.

i went with 500v and 6au6a's in the preamp positions

 

[ryuji]

for anyone that cares. this reflects current state of the design:
sorry for a little bit of messiness in diagram showing lowish voltage design

the lm338 is only running one of the 6AU6A's, the other 4 are running in pairs of two off the 12.6v supply. a bit more power efficient that way anyways.