Hello I have two links here with LM723 circuits:

http://schematics.circuitdiagram.net/viewer.php?id=klz1250398478f.jpg

This is the regular approach, like in the datsheet, and most of these circuits.

However, here is one weird circuit:

http://virtuatec.com.br/pages/Fonte-0V-%E1-50V-5A.html

where the output of the LM723 is routed via 100 KOhms to the VC+ input.

Can this work, and how??
I don't understand it. It looks backwards to me, at least.

I have been looking at these, and some other circuits, during a project developing a linear power supply (on request), we have now abandoned the LM723 approach.

So a solution is not urgent (anymore).

Would be happy though to see an explanation how and why this circuit can work, or is it wrong?

I would request a valid explanation (I did some hours guesswork myself)- maybe if you use or have used the LM723 yourself. I also have 2 of these chips here, have built this 2nd circuit, but then did not power it up.

 

That 100k resistor is the feedback path to tell the chip there is enough voltage.

Hang on...there's another one.

It looks like they are putting 238 uamps through the internal 6.2V zener, thus prebiasing it...
then current comes in pin11 as the 723 sends a "go" signal to the pass transistors...

Most complicated 723 circuit I've ever seen, and the datasheet is a PIB because the block diagram doesn't have the pin numbers of the 14 pin dip.

I'll have to redraw this before I can understand it.
Let's see...voltos de zeneros
Wish I could read Spanish.

OK. Main power comes in the pair of 4k7 resistors and is limited to 30 volts by a zener.
Pin6 is an internal voltage reference that powers a voltage divider made of a 10k over a 100k for 90.9 % of 7.15 volts = 6.5V
That 6.5 volts goes to the non-inverting input and the internal opamp goes positive, allowing current from the BD242A base to go through the 13V zener, the 4k7 resistor, the internal drive transistor, the internal 6.2V zener, to ground.
When enough current passes to the output, the voltage comes back in througn a 100k resistor to the inverting input (as in, "stop"), but adjusted by the 5k pot marked "VOLTOS SALIDA".

The pre-bias on the internal zener and the BC327 transistor next to the drive section must be about current limiting. The reset button on the left suggests it is a lockout current limiting circuit.

Is that enough to get you going?

 

hm...your writing is already a big improvement for my understanding. I spend hours really skipping back and forth through the PDF, due to different pin arrangement (for the older metal package I guess).

I use PNP myself in circuits but apparently did not get it for this one.

Maybe too hard for me to keep VCC+ and VC apart properly.

 

Well, there's your first lesson: Redraw it until it makes sense.
You don't have to put up with "mystery boxes" as long as you have a pencil.

 

yes I was considering this eventually.
But the LM723 is old technology.
I have researched many circuits for this project,
parts were purchased, proto circuits were made, and some 100s of web pages were read.

Currently a circuit using LM358 OpAmp is considered, having better (more precise) current limiter than the LM723.

Originally I was only involved to design/match microcontroller powered Volt/Amp meters for the project.

I was thinking this morning, try to clear up my questions, and once these LM723 are here, eventually use them for a circuit one day. Was an old regulation PCB by the way, from 1981, that contained 2 of these, plus related parts (63v capacitors), for 5 volts +/- 10% only. Unsoldered all the parts, and have put them in a drawer.

 

The pre-bias on the internal zener and the BC327 transistor next to the drive section must be about current limiting. The reset button on the left suggests it is a lockout current limiting circuit.

Is that enough to get you going?

Yes thank you very much. I have built the LM358 circuit during the past 3 days, and eventually want to power it up today.

The LM723 proto board is also still around, did not desolder it.

Yes the BC327 section on the left is some kind of flip-flop employed for current limiting, inhibiting the 723, and button-controlled reset.

I could test this in LTSpice if I wanted to.
Also I've downloaded the LM723 equtations for LTSpice, though did not yet enter this circuit (because consideration was changed to use a newer design/different circuit).

I am though very happy to find explanation here!

I could derive some of it on my own, but my main area are digital circuits/microcontrollers, so it takes me hours! I appreciate your explanation.

 

yes I was considering this eventually.
But the LM723 is old technology.

No problem. I'm an old person.

 

well if you don't mind, I have more questions about the LM723.

Is it common these days to build linear supplies using this IC,
for more than one Ampere? Most designs (for LM723 that I found during web searches) are for 3 Amps. max.

Originally, this IC was designed to be used on it's own, for some 150 mA current.

And for the LM723 circuit + the added symbols, I was still thinking NPN (somehow), did not get it how base current for the drive transistor can become established (during my verifying of the circuit/skipping through the LM723 PDF). It's obvious now, current flows out of the base, and through the internal zener diode.

 

In today's world the 723 is not very common. The 3 legged regulators have internal current limiting, overheat limiting, and require less board space. Even with the local capacitors attached, a 317L circuit is smaller than a 14 pin DIP.

On the other hand, the first time I checked out a power supply with a 3 legged chip, I attached the variable load, checked for current limiting, and had to wait for the smoke to clear after the rectifiers exploded. Current limiting didn't work that time! (Good thing it was somebody else's design.)

I particularly like the foldback circuit because IT WORKS! You can make a 723 regulator with Godnose how many pass transistors...I never found the limit...and the foldback circuit will save them. I did witness a laser power supply that had about a hundred 2N3055 pass transistors. (Apparently large MOSFETs had not been invented in 1970.)

I keep noticing you talking about PNP and NPN. The 723 has an NPN drive transistor in it. Better check your POV on that.

 

OK so there can be POV to use it for reason of ruggedness.

I was refering to the PNP driver in the linked circuit.
I am not completely unaware of the technology but apparently, mixing both (NPN and PNP) is enough to confuse me.

I don't always "see" how it works immediately. Well I did when I read your explanation!

 

When you get that "modern" circuit built, short the output and let me know if it survived.

 

Actually I work on this on behalf of a 3rd party.
I have run preliminary tests today.

1. I don't have a zero-Ohms voltage source.
2. The wire I use is not high diameter, the supplies are speaker wire, and on the PCB, bell wire in some places. There is also a shunt.
3. The circuit has 2 design features for current limiting.
4. It's MOSFET based, their Amps rating should be enough together with internal wire resistance, to survive 1000uF discharge. I don't know how long the second (fast) overcurrent protection takes to kick in, but there should be almost no delay to cut off the MOSFET drive.

No I haven't yet tested that.
And yes, I think it will withstand that.

 

If you want adjustable current limit I have a much more simple circuit. The circuit you posted seems to me somewhat overdone

 

The circuits posted have nothing to do with what is being built. We were merely examining a circuit that was so overdone that it was difficult to understand.

 

Yes the circuit from the first link is just an example, there are more of this type around, all having similar configuration.

For the second (spanish explanation on the website), I built a proto, but then did not power it up. One of the reasons was I did not fully understand how it works, but that was explained. Another reason is a better circuit was found.

With some moderate additional efforts, I could power up the 2n3055 proto as well.

I am searching a compromize between a total professional design, and a super-simple solution (only a few components). The circuit I have now completed, is somehow professional. I have to run more tests. Changing the design would mean, I have to go through all the testing again from zero.

One test is for nearly 60 volts DC input, so these input capacitors are not too common in my inventory. I use 3 now on the proto, one for instance for an auxillary rectifier (I want to test different supply for regulator amplifier/transistors), and one more is mounted on a 2n3055 assembly. This assembly contained 2xLM723 I removed most of the parts.

So, I have 2xLM723 here anyway, quite old, one from 1978. The 1000uF capacitors still fine (from 1980), giving powerful sparks, and storing power for quite a long while (I could not say if for 1/2 hour, or several hours).

I also wanted to get a general idea (from the replies), how much this IC is still being used. For the new circuit a new thread eventually would make sense, if I needed assistance/advice for that.

 

Hello again.

Today I made a new, larger proto PCB for the LM723.

-2 digit voltmeter, PIC controlled, 2x74hc164, firmware exist for other controller
-Voltage reference for PIC A/D, LM317L based with improved ripple rejection

-AUX supply for the PIC voltage display, based on LM317M + PNP transistor + reactor type coil from a small CFL. Using a wrong coil + smaller diode, the diode actually unsoldered itself.

It makes a whining noise now a little, 70mA/20V in, 200mA/6V out (added var. resistor to find optimal working point). 22 Ohm dummy load.

-LM723 + 2n3055

I want to build linear regulator circuit! The input voltage is about 20 volts (from transformer, DC). The circuit I mentioned earlier has 30V ZD!!

I want to change that reference to 15V, or replace altogether.

And also eventually, I want to add a Thyristor pre-regulator.

Have not yet decided if this eventually could be controlled by the PIC!

Any comments, ideas?? I don't critically need this circuit, it's for exercize/hobby only, LM723 = design from early 1970s.

Was a lot of effort actually to wire the 2-digit display.