slick gate driver

Very cool in that it has an isolated 10V output supply !

Too bad that it is only 10V drive though.

Also too bad that you can't get them !

boB

And if you could get 'em, it has a 100A rating; how are you gonna attach the busbars? What do you use as a protective fuse?

Where is the data sheet is this 100A rating? I went through it and couldn't find it.

It does seem to be designed to drive big MOSFETs and can also drive SCRs and TRIACs. Those can handle a lot of current and might need bus-bars.

The data sheet does talk about Kelvin connections.

It's a gate diver in an SO8 package.

Repeat: GATE DRIVER.

I'd use them right now if we can be assured of delivery.

10 volts is enough for most mosfets these days.

The link (not spec sheet) in that URL says 100A, John.

It's actually something like 1.5A source and 2A sync.

10V gate drive is enough for a lot of applications but where you need absolute lowest RdsOn, still wants higher on voltage.

AND if you need negative gate drive too, (which I do), then 10V is not nearly enough to have full on and a few negative volts Vgs

boB

There is this one with 15V output:

John

how many of them with a build in isolated supply?

ok, so you didn't look at any of them ....

ofcourse not, it would prove you wrong ...

Then I suppose you won't be designing it into anything.

NOW we're talking ! Thanks !

Can't get these either though.

I put this in my datasheet directory of FET drivers.

boB

This TI part, if you could get them, includes the isolated secondary supply where none of the typical isolated drivers do.

This one(s) are definitely different than most.

boB

I wouldn't touch them with a 10 ft pole.

Why risk it? Not second-sourced, and may never be, considering the rate of data sheet revision . . . .

RL

Single sourced, true. But we have to use many single sourced parts. A lot of custom parts as well.

All of our microcontrollers are also single sourced, as most are.

For a single sourced gate driver, we can normally, easily design in a new one, and in this case, would have to add a way to power the secondary if need be if this part went away.

Designing in another microcontroller would be a MUCH bigger task and take many months, if not a year, possibly.

Not that I am thinking about using this gate driver, but it would actually work to replace a bunch of circuitry, especially where space is critical. But, can't get them anyway so a moot point.

Happy Christmas !

boB

The integrated secondary supply is cool, the downside is you trade off speed as you can only transfer so much energy thru that bitty on-chip silicon isolation barrier so quickly.

4000ns rise is pretty slow, without the need for transferring power the modulator/demodulator isolation topology can do more like 40.

correction: not the rise time, rather the propagation delay.

4 microsecond propagation delay ? That is quite a bit for a high-speed switcher but would be fine for a simple On-Off gate operated switch like I would have used it for.

In that case, I don't care about the propagation delay but I do care about having the HV side gate turned ON or OFF for a long time (relatively speaking)

boB

It seems there's a single magnetic circuit, and some kind of modulation signals the output to switch; so, many cycles of the energy-transfer are sensed before the 'on/off' signal is recognized and acted on. A high-speed SMPS would use feedback to stabilize the long-term output, so a 4us propogation delay wouldn't be a killer for frequencies up to 30 kHz. For something faster, you'd want a second (inductive or optical) channel for the logic signal.