simple current limiters

And if you need more current, power or energy handling, do it with a buck converter.

That's basically what my current limiter box does (though at higher voltage, lower idle current, and bidirectionally and needing more components, than COTS regulators).

Tim

For higher currents you can use a "load switch". The ones designed for USB are extremely cheap and quite convenient, e.g. AP22802 ~ $0.05. 2.7-5.5V, protected with respect to over- current & temperature with a controlled slew-rate.

You can get them for other voltages too.

The USB 3.0 switches are quite useful, having a 20V 2A rating or something like that (I forget, it's been a little while since I looked).

Tim

Don't put yourself out with any kind of brain damaging cerebral effort there....

We need a minimalist current limiter in the 50-100 mA ballpark. Higher current more complex things are easy.

Polyfuses are too quirky for this app. They could take out a wall wart before they trip.

Yup, polyfuses open in 100s ms, transistors die in 100s us. Wall warts of either type (lin/SMPS) will easily collapse under the load of a polyfuse trying to open.

Tim

A flyback will just drop out of regulation and so long as the rectifier survives will produce extra current at a lower voltage unless the controller goes into foldback limiting. I stuck a 20W PMDC motor load on a 15W wall-wart, it was still speeding up after 20s

A wall-wart that can't survive a dead short was junk before it failed.

Typical flyback SMPS has a maximum power point just before it goes into hiccup mode.

The primary side only stays active until aux power runs out, then the controller stops and waits. I haven't seen an SMPS without hiccup mode in a long time, they're very common.

Tim

raw=1

Would something so simple, by itself, be suitable for driving delicate laser diodes, perchance?

-- This message may be freely reproduced without limit or charge only via the Usenet protocol. Reproduction in whole or part through other protocols, whether for profit or not, is conditional upon a charge of GBP10.00 per reproduction. Publication in this manner via non-Usenet protocols constitutes acceptance of this condition.

An opamp CSs is simple, but has a dreadful tolerance.

You should check out the huge variety of linear low-dropout LED current sources offered by ON Semi. E.g., these types stored in my computer. Most are 15%, but some types with A in the suffix are 10%.

NSI50010, 10mA in 0.46-watt sod-123, last two digits 10, 15, 20.

NSIC2020B, 20mA in 3-watt SMB, last two digits 20, 30, 50.

They also offer higher-power versions, NSI45030Z, 30mA in 1.4-watt sot-223; last two digits 20, 25, 30, 35, 60.

Adjustable J versions in 1.5-watt sot-223. NSI45020JZ, 20-40mA NSI45035JZ, 35-70mA NSI45090JZ, 35-70mA

Adjustable power versions in 2.7 to 4-watt DPak. NSI45035J, 60 to 100mA, NSI45060J, 60 to 100mA, NSI45090J, 90 to 160mA, NSI45090D, 90 to 160mA, NSI50150J, 150 to 350mA.

Unlikely. The short circuit current limit of opamps and LDOs is not normally well specified. You would do better than that just feeding them via a resistor from the supply.

We'd assume that an opamp or an LDO can furnish what the data sheet says. Actual ICs from a given manufacturer are actually pretty consistent. The ICs will thermal limit once they come off the rail, and will recover whenever they can. Most IC thermal limits actually cycle on/off, so are a nice full-current auto-retry fuse function as opposed to a foldback.

Those seem to be jfets or depletion mosfets; I've experimented with them. They are cool but their voltage drop is high.

We have a 5-volt supply and want to pass some +5 out to our customer on a connector pin, so he can power up a bit of his stuff. We want to be close to zero drop in our current limiter.

A polyfuse would take down a 5-volt wall-wart, transiently or permanently, if the customer shorted his side.

Probably not. The current wouldn't be very well defined.

An LM317 or LM1117 and a resistor makes a good constant-current source, but costs a few volts of drop.

Not many LDOs will thermally limit in the 50 to 100mA region. Very few opamps have thermal limiting. And most develop horrible voltage drops while still far away from so-called current limiting.* It'll be a matter of finding the perfect unusual part.

• I'm just now poring through 100s of datasheets, in order to add any current-limit and thermal-limit info I can find to the high-speed opamp table for the AoE x-Chapter book. It's very slim pickings.

BTW, most are not cheap, and they show excessive voltage drops above 25mA. Some parts you might check: OPA354, OPA357 >100mA, LTC6268 60-100mA?

Where does this "resistor" you mention (great economy of language on your part there, John) :-) figure in your suggestion?

-- This message may be freely reproduced without limit or charge only via the Usenet protocol. Reproduction in whole or part through other protocols, whether for profit or not, is conditional upon a charge of GBP10.00 per reproduction. Publication in this manner via non-Usenet protocols constitutes acceptance of this condition.

If they do go constant-current without thermal limiting, that's fine too. But a small part, SOT-23 size, probably will thermal limit at half a watt or so.

Most will survive a shorted output. Those are the ones I'd use.

LDOs are interesting, too. LP2980 maybe.

We have one new application where we will use two 317's in series. The first one is a current limiter like above, and the second is a DAC-programmable voltage regulator.

I use these.

1. A simple, widely used current limiter consists of a current sensor (usually a low-value sense resistor), a control circuit, and a pass transistor.

I wonder would this work with switching type 7805s. Like this.