1. Home
  2. Electronics Design
  3. SiC

SiC

Has anyone used SiC semiconductors before -- JFETs, MOSFETs or schottkies?

Specs look good on Cree's 20A, 1200V, 80mohm MOSFET -- better than silicon, and at that voltage, the $50 price tag isn't even that bad compared to the three $15 silicon MOSFETs you'd need to achieve the same conduction losses.

Switching losses in the datasheet aren't that great for high frequency use (>100kHz), but that's switching in hard recovery (i.e., another MOSFET's SiC P-N junction diode, which still recovers better than Si, though not by leaps and bounds), and a snubber and schottky will improve that dramatically.

What I'm most curious about, though, is why the gate "can't" go below -5V. Absolute maximum says +25/-5, appnote says -2 to -5 off, +20 on. Surely they use regular silicon dioxide gate insulation, which has no polarity? Is there a channel modulation effect when further reverse biased, like an increased reverse voltage drop?

I know the GaN power FET dies that are available, they don't really use the body diode; at Vgs = 0, the channel conducts before the junction, as long as you actually drive it at exactly 0 with a sufficiently low impedance.

Speaking of power GaN, anyone used them? Last I checked Digikey, they're only available up to 200V, solder bump dies, no power packages. That's not necessarily a bad thing, since the performance is enough to make a hard switcher at, let's say, 10MHz, with enough efficiency that you get pretty good current capacity from a couple of those dies, merely soaking the heat out with copper pours and gap pads. As power packages go, with all the GBW, you'd be lucky to even get a TO-247 to sit quietly, but there are lots of SMT packages that would at least give you some copper to heatsink from.

Tim

--
Deep Friar: a very philosophical monk.
Website: http://webpages.charter.net/dawill/tmoranwms
Reply to
Tim Williams
Loading thread data ...

The little I have heard about these (Silicon South parts from an IEEE meeting), indicate to me that the gate drive is too weird to use any of the parts at the present time. Maybe for a very high voltage, high power unit of some sort. Instead of having a D-S intrinsic diode, they're more like a J-fet and have G-S diodes that have to be dealt with.

boB

Reply to
boB

We use a lot of PHEMTS and some GaN parts.

The GaN fets have huge transconductances and low pinch-off voltages, low capacitances, so it's easy to drive their gates in switching applications. SiC used to need huge gate drives (20 volts, like a tube) but they're getting better. For RF, that doesn't matter, since you can use tuned matching networks to get gate drive.

We did destroy $1000 worth of Nitronex GaN fets one afternoon. I think they are better now.

We used some SiC schottkies, worked fine.

Most of these parts behave like jfets... no body diodes.

There are packaged GaN fets around. The trend is to make higher voltage parts, for RF amps.

--

John Larkin, President       Highland Technology Inc
www.highlandtechnology.com   jlarkin at highlandtechnology dot com   

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators
Custom timing and laser controllers
Photonics and fiberoptic TTL data links
VME  analog, thermocouple, LVDT, synchro, tachometer
Multichannel arbitrary waveform generators
Reply to
John Larkin

Most of these parts do behave like jfets. The GaN fets enhance to about 2x Idss if you push the gate positive maybe half a volt. Past that, it will start to conduct.

I thought I saw a super-enhancement, maybe bipolar, mode in some GaN fets as the gates start to conduct, but I didn't have time to explore that.

--

John Larkin, President       Highland Technology Inc
www.highlandtechnology.com   jlarkin at highlandtechnology dot com   

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators
Custom timing and laser controllers
Photonics and fiberoptic TTL data links
VME  analog, thermocouple, LVDT, synchro, tachometer
Multichannel arbitrary waveform generators
Reply to
John Larkin

I'm glad that someone is getting use out of the new GaN parts. I got real excited years ago for them and they finally started coming out. But so far the parts that are being made aren't useful for inverters and controllers. i.e. TO-220 packages with big die, etc. Although, regular low RDsOn FETs are so good these days that they really aren't the bottleneck they used to be. (250 volts and below)

I would like to see some 600 or 650V GaN parts in a TO-220 or TO-247 though. That could possibly be very useful. Let usknow if you find anything like that.

boB

Reply to
boB

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.