Linear 5V reg with adj.current limit

No wonder they're going nuts...

...

With those kind of voltages, you could do something like this with junkbox = parts. It's a simple highside current monitor driving a low RDS,ON MOSFET i= n foldback mode. Any 5V regulator will work. If you don't like the rheostat= then replace it with a DPST to select a 10R for 100mA or a 100R for 1A tri= p.=20 Please view in a fixed-width font such as Courier.

. . . 0-1A ADJ FOLDBACK 5V REGULATOR . . . .----------------[100K]------------------+--------. . | | | . | | | . | ----------- | | . | 5W | V.Reg | G | . +12V >-+-/\/\/\-+--+-|IN 5V|--+-----+-- D S---+--|->OUT . | 3.9R | | | | | | | | . | | =3D=3D=3D -----+----- | =3D=3D=3D IRF540 | | . [39K] | | | [39K] | | | . | | GND GND | GND | | . | 1N4148 | | | | . +--|>|---|----------. +------[390]------' | . | | | | [100] . | [39K] +12V | | | . | | | | LM358 | 0.1u | . | | |\| | | .----||----+ . | +---|-\ |/ | | | . | | | >-|PN3904 | | |\ | . +--------|---|+/ |> +--[10K]--|--|+\ | . | | |/| | .| | | >----' . | | UA | +--------|--[10K]--+--|-/ . | | | | | |/ . | | | | .----+ UB . | | | | | |_ . [39K] [39K] | [10K] | |/| . | | | | | [100] . | | | | | /| . GND>--+--------+-----+----+---|----+--------------------->GND . | . '------------> I_TRIP 1Volt/Amp . . adjust with output . terminals open . P =3D 1.3 W . FET,MAX . . . P =3D 3.2 W . REG,MAX . . .

parts. It's a simple highside current monitor driving a low RDS,ON MOSFET in foldback mode. Any 5V regulator will work. If you don't like the rheostat then replace it with a DPST to select a 10R for 100mA or a 100R for 1A trip.

Nice simple solution ;-) ...Jim Thompson

has

In a way it is really cool, the 723 is older than the 555 and still known to be very useful. Kinda like the uA741 and the LM101 as well.

?-)

As I've intimated before, I've had endless fun (not), getting 723s with added pass transistors, to stop oscillating at HF, in other people's (commercial) designs.

The 741 used to latch up.Maybe they've cured that problem in the intervening years. I stopped using them as soon as I could.

The 741's that latched were bad copies... the originally published schematic was of an early botched version, but was never changed... quick-to-second-source loons missed it ;-) ...Jim Thompson

Was that added pass transistor a PNP common emitter? Same problem as with the modern LDOs, you need to know a little bit about phase compensation. There is no stability problem with NPN darlington current boost and the 723.

The 723 was a second iteration of an earlier general purpose linear regulator with a less than ideal voltage reference, the NE550,

Commercial is not a guarantee for good designs. Often the circuit design is ok but the layout is horrid.

I've never had even a whiff of a problem with the 723, they always worked.

Back in its days discrete designs were often cheaper. So I rarely used it and if I did need an opamp I preferred the 709. But the bulk of my designs use the 324, it is much better.

I wonder what's the oldest IC still in production.

Maybe the SN5400/7400 (ca. 1965)- still shows as "active" on TI's website.

That's approaching the half-century mark, and starting only a few years after ICs were first commercialized.

Best regards, Spehro Pefhany

It's been a bit long ago but I think one reason I liked the 709 better was because it was not internally compensated so cuold be souped up. Of course, my class mates only souped up mopeds instead. One of them died that way :-(

Most likely a military chip where we'll never find out the introduction date unless it's declassified before we all are six feet under.

Sounds like it. The 741 was introduced in 1968, and the 555 in 1971.

On a sunny day (Fri, 06 Apr 2012 19:10:41 -0700) it happened John Larkin wrote in :

I made a nice audio mixer with some 741s, liked it much better than the 709, no need for that compensation capacitor. I still have some 741s in the box.

Putting aside that they need a rather higher bus rail voltage compared to today's standards, to be usable, I really don't see a reason why a

I still have a lot of them and still use them for generic applications where precise operation isn't required, which is the case in many applications these days.

I find that many of today's monolithic op-amps aren't much better or in some cases worst, than the old 741 and still they get the recognition, only because they're newer?

It's the same old argument with the 555 timer. Aside from the fact that they don't operate in the lower voltage world, they have proved to be something that is hard to kill. There was a 1 Cell 555 timer out there but for some reason no one is producing them at the moment.

I don't know, that's my take on it.

Jamie

Popcorn is rare these days, but shows up once in a while. We had some Analog Devices DACs that were terrible... until they discontinued them.

No.Discrete NPN Darlington arrangement, with multiple paralleled pass transistors.

Several similar designs from different manufacturers.

Same problem as with

I'd disagree. Experience back in the day, and some basic simulations, now, show that the 723, using Darlington current boost, and 100% feedback, is only conditionally stable. There's a problem area in the region of 3 to 10 MHz, where it is, exactly, depends on the ft of the particular transistors.

Reducing feedback, and adjusting the reference voltage fed to the noninverting input accordingly, makes the circuit unconditionally stable. Regulation *does* get degraded a bit, but is still around 1.5mV on a 12V regulator swinging from 1amp to 10 amps, using 50% feedback.

It's sort of a Nyquist thing.

I never had any problems with 709s.

I burned many-a-finger breadboarding 709s. The crappy white breadboards ("EL sockets", IIRC) made crappy contact and when they lost compensation they broke into smoke. I switched to 741s and got my projects (analog computer circuits) done. ;-)

An uncompensated 709 made a decent comparator. My first switching regulator, 24-to-5 volts, was a 709-based hysteretic buck. The 709 drove a 2N2905, then a 2N3055 emitter follower, which didn't saturate so stayed reasonably fast. I designed it to run around 23 KHz, because I could still hear 22K in those days.