Source-o-dyne Phase Inverter

[toshiba_nz]

Hi all,

I really want to try a source-o-dyne phase inverter, like Gnobuddy's ones, however I'm unsure about the MOSFET biasing.

My power will be plus 300VDC - 0V - minus 300VDC , and the MOSFET will be the 2sk2700

How do I work out:

1/ the ideal current thru the MOSFET?
2/ R1/R2
3/ R3
4/ R4
5/ R5

Cheers

Schematic courtesy of Gnobuddy

 

[tikiroo]

Current - depends on what it is driving (resistive and capacitive load). A few mA usually sufficient - say 5 mA. Driving a pentode can be lower, driving large triode or paralled devices might need to be higher.
Biasing - you need to know what output swing you need, and allow a minimum of 20V across the mosfet (Crss is low above 20V). Say you need to swing 200V peak to peak on each output, you could bias to 120 on the drain and -120 on the source. This would swing between +/- 220V (440V across mosfet) and +/- 20V (40V across mosfet). Then for 5mA through the mosfet:

R1 = R2 = (300-120)/0.005 = 36k or just use 33k for a slightly higher current
Gate voltage about 4V higher than source voltage = -116V
Current through R3/R4 can be very low, just needs to drive gate. Say 0.5 mA so R4 = (300 - 116)/0.0005 = 368k, R3 = (300+116)/0.0005 = 832k. Or just use 330k and 820k, will still be OK.
R5 to stop oscillation, 100 ohm-1k.

Quite a lot of leeway with all these, unless you are trying to maximise voltage swing with limited B+.

 

[jcalvarez]

C1 & C2 is very low for a typical valve output stage being driven.

 

[toshiba_nz]

Cheers Tikiroo!

The PI will be driving 45's (Tubelab powerdrive)

The input could be anything yet (maybe el33 (triode) , 6c5g, 6j5g, 843, 7b4.)

 

[toshiba_nz]

Well, today is the day where I've said goodbye to my smoked glass RCA VT-231 Phase inverter, and a welcome and happy "hello!" to my Wolfspeed C2M1000170D MOSFET Source-o-dyne Phase Inverter - which is now sounding absolutely fantastic!


I ended up putting 0-960VDC (and not the +/-300vDC) on the fet. Partly due to the gain in my DHT finals, I had to implement a CLC arrangement for the power supply to the Source-o-dyne.....Sic Schottky diodes -> 22uF PIO (1.5Kv) -> 6Hy oil choke -> 10uF (1Kv) PIO . Almost no hum. Almost )

The output of my Source-o-dyne PI (2 phases) are RC coupled (FT3's) to two Wolfspeed MOSFET Source Followers (textbook Tubelab) , then DC onto the grids of my DHT's.

The input of the Source-o-dyne Phase Inverter is DC coupled to the prior gain stage (in my case , a 7b4)

Wolfspeed C2M1000170D MOSFET Source-o-dyne Voltages:

Power Supply: 960 VDC

Source and Drain Resistors: 33K Caddock thick film

Drain: 676 VDC
Source: 275 VDC
Gate: 280 VDC

All in all a great outcome, and special thanks to DiyAudio member tikiroo for his massive help ))

PS forgive my sloppy skills (I just had to get it working)

 

[stocktrader200]

Bias the gate at about 102v should be sbout perfect. 20k source and drain resistors.
Works really well. I use it at 150v however.
2.2m and 1m is excellent with DN2450 Fet