AoE x-Chapters, 45 years of power MOSFETs

OK, depletion mode wasn't mentioned in the saga section, but it was plenty elsewhere, and cross- referenced. It's also reasonably-well covered in the main book, with Table 3.6 as well, where the IXTH20N50D is at the bottom of the table; the biggest, baddest part, the top of the heap. Piotr, have you used it for anything interesting?

Siliconix it was. I went to a conference talk by one of their apps guys in early 1982. They were really emphasizing the (then) good current sharing of parallelled FETs vs. bipolars.

Cheers

Phil Hobbs

It is impossible to overpraise these niche parts. Could you please just mention that their high-power variants do exist and let the reader google the details? ;-)

Once upon a time I badly wanted to use a pair of them as an AC current limiter, but they turned out to be expensive unobtainium back then. They would have had many advantages (e.g. they require no external power to start operation, while the enhancement-mode approach has the usual chicken-and-egg problem), but due to the empty stock I had to chose a more standard path.

I have 3 successfull applications you might find interesting enough:

1. Normally closed micropower "bleeder resistor" equivalent based on an IXTP08N100D2. This is to discharge the bus capacitor charged to 600V by a PFC stage. The current is set to the max level allowed by the thermal capabilities of TO220 under continuous operation. If the powered device (an FPGA, to be specific) fails for some reason, the bleeder steps in, otherwise it powers a tiny RF transformer wound on a 6mm core with a 1MHz square wave that produces the negative gate voltage, turning the bleeder off.
2. Active clamp transistor in a forward SMPS which operated at voltage just within the allowed VGS ratings. Extremely simple gate drive, the same PWM signal used to drive the main switch and the depletion-mode FET. Maybe patentable, I don't really care.
3. Phase-sensitive AC zero-crossing detector made of two LND150 and a coupled choke wound on an amorphic metal core. This is basically an AC current source powering a coil with 1mA, so nothing spectacular happens there. Unless the polarity reverses and the "current source" is no longer capable of providing constant current. Then the magnetic field collapses and bang! -- there is a very narrow voltage spike indicating the zero crossing. The second coil wound on the same core transforms this rapidly changing magnetic field into a desired voltage spike, say +/-2.5V. If the polarity information is a problem, just add a full-wave rectifier to the output. The choke can be wound with a TIW, so a substantial insulation level can be achieved. The LND150s in TO92 can be put together along the flat part of the casing and then inserted into the hole of the 12.7mm core. Then the entire isolated zero crossing detector would be composed of just 4 parts (2x the FET, a 1mA-setting shared source resistor and the "transformer" can fit within a circle of diameter ~14mm.

Best regards, Piotr