Around 30V Mosfet that can handle 30V Vgs

You're going to have to indicate current, Rds(on), mounting, and power dissipation if you want a useful reply.

I've seen a few 30Vgs FETs in my travels, but don't have any on the tip of my tongue. I always zener-clamp 'em, limit Vgs by design, or both.

Cheers, James Arthur

Vishay's parametric search lets you specify 30Vgs. Returns ~200 parts.

Cheers, James Arthur

Probably a good policy. Isn't Vgsmax +/-20V something of a standard? ...Jim Thompson

Yep, +/- 20Vgs is far and away the most common rating, though I've seen higher on occasion, and these days, lower.

Cheers, James Arthur

You really want a vacuum tube?

He wants P channel - P channel toobs have been out of production for decades.

Please see the thread "ping Win LND150 avalanche" from J. Larkin on this ng. I've learn few useful things ... IMHO.

Best regards, Habib.

I posted about that recently. The zener-protected parts seem to zener around 40 volts, and unprotected faets usually blow out around 70 volts.

A few years ago I reverse-biased an ordinary indicator LED, accidentally, in a proto.

Rated at ~4V, it took 24V forever and was none the worse for it, at least not brightness-wise. I should've measured reverse leakage afterwards...

Cheers, James

Rdson ~ 100m ohm ish SMD Max 100mW

I do the same, per design

This design will have a great benefit from a 30V Vgs part, will simplify a lot

Cheers

Klaus

That's great, just what I needed. Sadly very limited availability of devices for 30Vgs, mostly high voltage parts

Cheers

Klaus

That's good information. However, I will never be able to use that in a design review :-(

Cheers

Klaus

While looking for an LED that would break down it took me a good pile of assorted LEDs I have before I could find one that would break down at

Most that I have will handling my 40 Volt supply just fine..

Jamie

It's an interesting philosophical issue, whether to approach or exceed abs max ratings. A similar problem is whether to use measured values in a design, like gate leakage, or use the specified limits, which could easily be 10,000 times worse.

We exceed abs max sometimes when the actual parts seem to have margin, and when performance substantially benefits. Some people never do.

We use some schottky diodes at 2x their rated max reverse voltage. And we sometimes run resistors at several times their rated power. But only as tested, calculated risks, often only one critical part in a product.

I've reversed biased several (10-20) visible leds. most of 'em didn't go till they were above 100V. The ~25 V ones were/are interesting... very inefficient single photon detectors. There are some IR one's that go at ~10 V and are also photo sensitive.

George H.

A random walk through parts on my hard-drive turns up the FQP27P06. Fairchild, +/- 25Vgs, P-FET, Vds=60V, Ron=70milliohm, TO-220.

Imperfect, but might be a starting point for searching...

Okay, Fairchild has loads of 30Vgs-rated FETs.

Cheers, James Arthur

Better search spec, P-FETs: 33 parts

I think if something is important, then you better measure it (at least twice*) to be sure. And you should always try and blow things up.. push 'em to the limit. See where things melt and the smoke comes out. otherwise how do you know where the weak link is? (Not all my coworkers agree with me.)

I take "absolute limits" as good guide's, but not always so absolute. (I have this "36V opamp" that is powered from 48V (with regulators) but I was worried about what would happen if the regulators were shorted by students. I wanted to give them schematics and let them change things. I tested a bunch of opamps to ~50 V and they were all fine... (I think it was you who encouraged me to push 'em higher.)

George H.

• twice because one of Murphy's laws is that every prototype has one golden part.

Blow things up, absolutely! There's a huge difference between a 10% destruct margin and 200%, and it's crazy not to know.

We respect datasheet limits except when there is a big payoff for calibrated pushing past the datasheet limits.

Why did you stop at 50? That's no fun.

Lots of older analog process opamps, like 741s, would work OK around