fast, small low-capacitance zener to protect MOSFET gates

Series inductive peaking will get you about sqrt(2) bandwidth improvement with a flat response. A constant-resistance T coil will get you a factor of 2.8, at the price of an extra lead.

Cheers

Phil Hobbs

to

L may improve f response, but you now have L in series with your D: how doe s that bear on its protective functionality?

NT

There's L in the source of the MOSFET too, at much lower R. As long as the gate TC isn't too much faster than the source, it works fine.

Win did this a few years back when testing Tom McEwen's patent on it, after a discussion here.

Clifford Heath.

Ahh, interesting, and unsurprising. The I layer has to re-establish, of course. I suppose the same thing happens with rectifier diodes. In the case of diodes turning on to protect a MOSFET gate, perhaps the turn-off delay wouldn't be a problem however.

Clifford Heath.

Using the reverse breakdown of an EB-junction, with CB shorted? Often that is about 6..7V, low capacitance and very low leakage. Fluke used that as DMM input protection.

Regards, Arie de Muijnck

Yes, that should work well, and was actually my first plan. I didn't have room for an SOT-23, and tilted towards a zener diode, but an SC70 or SOT-23 vs two diodes may be more attractive.

related,

Thanks, If I'm reading the fluke 87 schematic right (at ~5:30) there is 1k (fusible) + 1.5 k ohms of series resistance. Seems like that would be a lot of current at 1000 V_in... there must be some other voltage clamp?

George H.