The automotive guys can't get enough of them. Apparently they don't blow them up, or dump EMI down the wiring harness. You must've been using them wrong...
TPS1H100 and suchlike are pretty amazing, too. Hard to kill.
Tim
First thing out of college, I did it at 400kHz. Seemed to have enough microseconds to spare that it would work at 600, maybe 800, if the drivers were up to it.
The RF equivalent is simply tracking the negative-going amplitude. What's impractical about that?
Tim
Depends on the definition of high frequency.
That's a complex IC with 8 pins. I was referring to 3-terminal fets. They need to power the protection circuits from available gate drive. Maybe the ones that I tried didn't like my gate drive.
Other people here have said discouraging things about self-protecting fets. There may be a good one somewhere.
While that's acceptable work at low voltages, bipolar switches, like a humble MMBT4418, are nicely ON at well under 1V.
I'm not comfortable with the lowest MOS threshold devices, don't they leak? Assuming the drive can pull down to ground, a bipolar makes a good switch ON and a better one OFF.
That's a fine point, too: low-Vce(sat) types are very competitive with MOSFETs, especially for low drive voltages.
Tim
MMBT4418 ??
How about MMBT4401W and 4403W, in SC-70 packages?
I'm far from a FET expert. I tried some of the three pin protected fets (checks part drawer vnp10n07, auips1031, bts133bksa1) I can't recall exactly what I didn't like, but I ended up building in upstream protection.
I was hoping for something like the LM395.
George H.
Doesn't a bipolar transistor need medication to act sane?
Yeah, I fumbled at the keyboard; I just brought up the highest stocking transistors on DigiKey's search, and it was MMBT4401 that I intended to name as 'humble' based on the nice low price.
SM packages just confuse me, though. For my hand-wired needs, bigger is better, anyhow.
TO-92, easy. But SOT-23, 1.9 mm spacing, not so bad either. SC-70 with 1.3 mm spacing, surprisingly easy for hand soldering. These are all 3-pin parts.
But 6-pin SOT-23-6, now 0.95mm spacing, getting tough, and SC-70-6, 0.65 mm, painful with a soldering iron.
8-pin SOT-23, OK, throw in the towel, fire up the solder paste dispenser and hot-air pencils, etc.** ** Or draw a bead of solder across all the pins, go back with solder wick and clean between pins, examine with magnifier and hope for the best.
Ahh, those are fine. They still have leads, and a steady hand or a bit of poking gets things lined up. SOT723 is rough to do by hand though, need to leave a lot of toe fillet to even get a radiused iron tip in against the pad.
Tim
Haven't tried them, you're struggling with the tiny 0.2mm pad sticking out from the package?
It'd be useful to have a 3-terminal "FET" that included thermal cutoff protection, in case it switches a short. OK, give it four terminals, allowing for a fault signal. OK, give it five terminals, to power the fault circuit.
There are several around. I tried a few and didn't like them.
Some telegraph shutdown through the gate lead.
That would probably work better.
Or just bring out a chip temperature signal, like many FPGAs do.
Bingo! I'd love to get BJTs, MOSFETs, etc., with integral diodes or BJTs for temp meas.
SOT-23s seem to preferentially fall out of the tape upside down. Very annoying.
Do the smaller ones do the same thing?
--Spehro Pefhany
Nice! It's unusual to get a decent Vds with Rds(on) guaranteed at 1.8V or 2.5V.
--Spehro Pefhany
It's called an SC-70 package, OK, but actually no, it's one of these bottom-only pads packages with some exposed unplated copper on the sides, not suited for reliable hand soldering and hard to inspect. The distributor images are wrong.
Have something to add? Share your thoughts — no account required.
Ask the community — no account required