neat paper on SiC history

Dan, I'm not sure there are any safe bets yet. There are a few folks dabbling, still working out bugs, and it's early days. That said, Cree might merit a look.

Don't forget GaN on the horizon, too--its fans think it'll eclipse SiC.

Cheers, James Arthur

Ooo, here's one-

Cheers, James Arthur

Worst of both worlds? Looks like the SiC is a depletion part, which is why the transfer curve flattens out.

Note the (?) zener that keeps the drain-source junction down.

It's not avalanche rated, which would be a concern.

SiC and GaN are just now getting good. The next few years should be fun. What's next? Doped diamond?

It doesn't look very attractive to me. I think it was your history of SiC link (OP) that explained they started off with SiC JFETs because they didn't think they could make MOSFETs, then cascoded a Si MOSFET to create an enhancement-mode hybrid.

I wish they'd hurry up with single-crystal diamond films. We need them.

Cheers, James Arthur

You can have any diamond film you want, as long as you only want intrinsic and P type.

N doping appears to be intractable. No shallow donors are known (meaning, anything that supplies useful carrier concentrations at room temperature), only deep traps. Theoretically, adjacent dopant pairs (giving more options as far as energy levels) could perhaps be _constructed_ (but likely not merely diffused into the crystal, more likely that would destroy the pairing), but who knows how to do that with something easy like CVD?

Unless it can be done with diamond films without substrate, it seems more likely another form of carbon will be the way to go. Graphene and nanotubes are fun but still "10 years out", except for a few special applications just getting practical today.

Tim

Somebosy makes a SiC bipolar transistor (Genesic?), but they can't make up their mind if the middle lead is "base" or "gate."

Am 05.02.2018 um 20:01 schrieb John Larkin: r

Do you mean these?

Cheers, Gerhard

Might be doable with ALD.

Cheers

Phil Hobbs

No, those are small-signal RF things. I meant these:

Am 05.02.2018 um 21:28 schrieb John Larkin:

Oh, that's more for robust people.

I'm just rewinding a Pulse CX2041 as a 10T:40T step up transformer under the microscope with 50u wire. Just found out that the primary has 0.7 Ohm. That eats into the noise budget of my new chopper amplifier :-(