• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Solid State Cathodyne Phase Splitter Device?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I'm working on an amplifier with what appears to be a solid state cathodyne phase splitter. It has +300VDC on one leg and -300VDC on the other, with the signal coming from a triode (12AX7) gain stage. The solid state device cannot be read. It's a TO-220 package or similar. What sort of solid state device would be appropriate in this application? Any recommendations for one to try?
 
Last edited:
A 1 KV. rated MOSFET seems to be indicated. You want a part with a low reverse transfer capacitance. At 9 pF., this looks like a reasonable candidate. However, you did say the driving element will be a 'X7 section. I don't know if that low gm/high RP triode has enough "stones". Frankly, the only MOSFET I'm comfortable with in combination with 'X7s is the ZVN0545A and that part can't handle the voltage described.
 
Thank you all for your helpful responses.

The resistors are 23k's.

I had not thought of an IGBT. Does anyone have experience with them in an audio signal path application?

The other options look good too (and I'm always interested in more). The fun thing with solid state components is they're inexpensive and have only three leads, so I can try them all, if I'm careful.
 
Obtainium $1.78 qty 1 (1200V 6pf reverse)
Digi-Key - 497-4353-5-ND (Manufacturer - STGF3NC120HD)

Nelson Pass built a very nice amp of IGBTs.
Complimentary pair was once available.
Not sure why P-IGBT was dropped?

12,000V P-Unobtainium?!? Silicon carbide: Go figure...
http://www.scientific.net/MSF.600-603.1187

For your circuit:
600V/(23K+23K) = 13mA max. (probably biased at half this amount)
300V*6.5mA = 1.95W max (and probably biased to idle on this peak).

Wonder what blew the original? Is there a Zenier Diode nearby?
 
Last edited:
The gate to source capacitance is canceled out cause they track same voltage.
The gate to drain (Miller) capacitance is double cause volts go equal and opposite
in your Concertina circuit.

Zenier without stripe might be two in the same package back to back, no direction.
10V is usual Zenier value. 1200V drain, still can't take more than 20V gate to source.
 
Last edited:
I have some of the the Fairchild FQU1N80 devices in the I-pak case. These are just a TO-220 with the bolt tab trimmed down to a stub. You just have to use a little spring clip (made for these and available) to bolt them to a heatsink. No need for the insulated pak either since the mounting bolt does not touch the Mosfet tab. Just a mica insulator underneath does it. Or there are insulating sleeves that you slide over the whole thing.

The reverse device capacitance (output back to gate) gets multiplied by the device gain (2X here). This has to be driven by the pre-amp tube, setting a bandwidth or slew rate. 12AX7 is about as wimpy as they come.
 
Last edited:
12AX7 Plate to Grid adds another 1.7pF to that equation.

80K plate resistance into 100pF would yield 20KHz cutoff
Anything substantially less than 100pF is potentially usable
in a circuit with no global feedback loop.

Still the question other poles of phase shift in a global loop?
Its best we keep this pole high, and let the transformer be
the dominant low pass. Minimal capacitance here the better.

If we got two low pass filters near same cutoff frequency
in a loop, it can become an oscillator. Make them all be high
cutoff except one. That low pass will attenuate the loop to a
safe gain below unity before other phase shifts roll into play.
Do we all agree that one should be the output transformer?

I'm just saying we ought not to settle for 20K rolloff here.
We want to shoot as high (low capacitance) as possible.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.