LDO with current limit

tirsdag den 19. februar 2019 kl. 20.30.07 UTC+1 skrev John Larkin:

what current?

With +24 volts in, why restrict yourself to an LDO?

What current levels?

It will be a pulse generator and I'd like to make as much output swing as possible. 20 volts sounds good, 22 better.

I need to current limit in case some yahoo shorts my output and fries things.

Maybe a bit less, low end below 100 mA maybe. That would be a more versatile part to have in stock.

I was just wondering if anyone was using a part like that.

We have TPS26600PWPT in stock, but it's fairly complex and fairly expensive. I blew some up running off 48V so I don't entirely trust it.

My ideal part would be an LDO with resistor programmable current limit.

Trolling digikey the fist hit... (that have ~400 with current limits) is the ua723

GH

Jeesh I can't spell at all. Here's DK's list trimmed down some.

uA723 might be older than you are! But not a bad idea.

NT

tirsdag den 19. februar 2019 kl. 21.48.24 UTC+1 skrev John Larkin:

an ldo with a pnp and series resistor ppulling the feedback pin up?

I've never designed anything with the uA723, but it's used in a ~150V linear supply we use. And I had to hack the supply for lower voltage. (hack in this case means adding the right parallel resistor to the HV output sensing voltage divider.)

There's the LT3081.. higher current... I've used the LT3080

George H.

Maybe you can use a high-side pass transistor? GH

I'm using the LT3086 in a project at the moment, seems quite decent. Has monitoring pins for temperature and current, either of which can shut down the part if configured appropriately. Tolerates up to 40V in.

-- john, KE5FX

OK, fine, you know the drill. A P-channel MOSFET at the input, with small current-sense resistor at its source. A sensitive amplifier in loop for current limiting (needs foldback), or a comparator for an electronic fuse, both likely need special control smarts.

You could roll your own LDO with a PNP/PMOS pass element and TL431 reference, and use a sense resistor and second TL431 on the output as the limiter, kill drive to the pass when the sense drop hits 2.5 (assuming the goal is to current limit on fault condition where the output current tries to go >> nominal)

Looks like I'll make a current limiter with a sense resistor and a mosfet and an opamp. All parts that we have, and the voltage drop will be about 0.2

a pnp ought to do it with ~0.7 not much compared to 24

This ought to work:

It's slow and soft in time, which works for my application.

We rarely use bipolar transistors any more.

Earlier I suggested that a linear current-limiting circuit should include a foldback function. A 150mA current limit with 24-volts will cause 3.6 watts of dissipation in the MOSFET, which means you'll need a pretty serious heatsink. It'd be better to add one or two foldback resistors, and limit the FET dissipation to say 1.2 watts, easy to deal with.