1.4kW; is this crazy?

[davygrvy]

Was playing around in SPICE combining some different ideas and came up with this hybrid. Output is circlotron and I'm aiming to play around with those SemiSouth SJEP120R100A silicon carbide static induction transistors (modern power jFET). The INXS mos as shown is just a placeholder.

Bias feed will need some more stability, and a required power on delay relay, etc.. but I think the topology looks usable for the high powered bass amp I'm going for.

Anyone have any thoughts on this before I breadboard it?

[davygrvy]

Nobody?

[geraldfryjr]

it looks cool !!!
I am always up for highpower designs.

jer

[kenpeter]

A real static induction transistor would have voltage amplification
factor "Mu", and the curveset of a triode. This one looks a really
big enhancement mode JFET to me. With much as 1A of gate to
source forward bias current. You will never drive that sort of A2
with 100K "gate leak" bias...

Maybe you linked us the wrong specsheet?

[davygrvy]

Yes, it'll draw some gate current. Good point. I can't wait to find out how much

[kenpeter]

Also note reverse transfer (Miller) capacitance varies from 60pF to 600pF
over the 170V-10V range of drain voltage I expect you are shooting for...

In a lot of ways, this part is far harder to drive than its worth. Its not
voltage linear (not a SIT) , nor is it current linear, nor is its capacitance
stable, demands gobs of forward gate current (worse than 2n3055).
What was the point again?

You need to forget SIC enhancement JFET and look you to IGBTs.

[davygrvy]

we'll see, non believer.

At a forward voltage of 3Vgs, it has 220mA of gate current. Drain current is off the graphs.. might be 20A to 30A there.

Time to warm-up the Tek 575 and get some graphs above 6V

[davygrvy]

IGBTs? Those are probably the worst parts to ever be used in linear amplifier applications. They are commonly found for high voltage pulsed applications like power inverters. IGBTs go thermal rather easy from my experience

[Andrew Eckhardt]

I guess it could work if you could keep the case at 25C. You will probably have to use refrigeration or more parts to get that much power. Use full face shield.

[kenpeter]

IXYS SEMICONDUCTOR|IXGH48N60B3C1|GenX3 SILICON CARBIDE (SiC) IGBT | Newark.com

At least Miller capacitance stays below 100pF down to about 13V... And no forward gate current.
Why you not using quadrature feedback to shape output currents and prevent thermal runaway?
You probably need serious current control anyway, just to keep from redlining wires in your wall.