current limiter chip

Page 17, fig 39:

Those relatively high current monolithic lithium battery chargers use that dual MOSFET technique in which a second small scale MOSFET has the same VGS as the higher current charging path MOSFET.

Input voltage range? Precision? Bandwidth?

It takes an op amp with one-rail input, and voltage reference, and maybe pass transistor, to do that, using three pins. Low voltage range means only small current-sense resistance is acceptable, and precision will suffer when offset voltages dominate.

Some of the LED drive circuits are VERY available, and limit current... but do it with switch-off transients. And with the right power available, and into a low-Z load, a surface-mount resistor (0805) would work.

Mp5032 USB current limiter/switch

Cheers

Sure, I can design a circuit with parts. That one is not low dropout and does not thermal limit.

The curve for Rlimit goes down to 500ma. Is it less than 500ma?

Cheers

This seems appropriate

DigiKey claims to have stock. It's a sink, not a source, but... regulated for sure. You say 'limiter' which is a looser specification, and maybe a PTC resistor is just as good.

What I/O voltage?

"Set the current limit level. Place a resistor between ILIM and GND to achieve a high accuracy current limit. The current limit threshold varies versus the output voltage to maintain a fairly constant output power limit."

Sounds tricky. And maybe it needs to do a handshake with a USB device.

The current (no pun) problem is to protect a pulse output stage that makes 5 volts out, from a 5 volt supply. It normally has a 50 ohm series resistor from the driver to the connector, but some customers want us to be able to programmably short that resistor so they get 5 volts - not 2.5 - into their 50 ohm load.

That current limit should be a bit above 100 mA. But the current limit problem keeps coming up.

I don't understand that part. Does the mosfet switch hard to ground and PWM to tune the average current?

MAX4995 is close but it's $5 in small quantities.

There's no logic in the connections they show in the lame block diagram. Clearly, the pass MOSFET has to have its source lead connected to the Rext or 'typical application' won't work. I suspect the op amp output goes to the 'driver' input, also not-as-depicted.

It wouldn't be reasonable to PWM and not have any timing components.

Maybe there is someone that makes a high side switch with an adjustable current limit.

Cheers

NCV47722

Similar, but lower current.

Cheers

On a sunny day (Fri, 02 Sep 2022 16:20:36 -0700) it happened John Larkin <jlarkin@highland_atwork_technology.com> wrote in snipped-for-privacy@4ax.com:

Switch to a 10 V supply?

Hmm. My usual approach is to put a resistor in the supply lead, with a nice beefy bypass cap to let the output amp ride over transients. That won't work unless you can power the output amp from a bit taller supply.

When there's at least a bit of headroom available, and the current limit can be a bit sloppy, I've also used a low-sat PNP transistor with a resistor to ground from the base. This keeps the PNP in saturation in normal use, and rides over brief transients, even pretty large ones. Longer transients deplete the stored charge in the base, so the collector voltage collapses in a few microseconds.

(I realize you're looking for a specific integrated solution to a precise problem, but (a) I know of none, and (b) general on-topic discussion is always good.)

Cheers

Phil Hobbs

The part I want seems pretty obvious, but nobody seems to make it.

Somebody, National maybe, once sold a PNP transistor with a super tight beta spec, specifically as a current limiter. But it wouldn't thermal limit.

One alternate is an LM317 type reg with a big resistor from +12 volts to limit current. That's klunky.

Well, Martin R's ncv47722 device looks interesting.

Cheers

Phil Hobbs