How do make a current limiting knob using LM723 for Linear Reg power supplies?

[Gurumurthy]

Hi to all!!!

Currently I am trying to construct Laboratory type Linear Power Supply with LM723 IC. How do make a current limiting knob using LM723IC and my Specification Were 0-30V DC and 0-5A Current ? I am much concern about Current Adjust knob !!!!

[Kleinstein]

The current adjustment on the 723 circuit is usually not that great, but might be still acceptable. There are several options, depending on how accurate and how fast the current limiting should be. Depending on the circuit used, the adjustment range can be rather limited (e.g. 100 mA to 1 A). This is because the original circuit inside the 723 uses just the transistor base-emitter voltage as its reference. One can reduce the current setting by adding a part of a diode drop to the voltage, but this is limited, as the drift will be relatively high.
So the current limit regulation is more like an over current limit for protection, adjustable by choosing the right shunt.

There are a few circuits that use a shunt at the low side and an extra OP for the current limit - usually more accurate, but also slower.

[Gurumurthy]

Thank you sir. I need schematics to understand sir.

[Gyro]

You should find the original uA723 Application notes useful:

http://www.ve6aqo.com/Manuals/uA723_Application_Note.pdf

[David Hess]

A wide range variable current limit is best done by using an external error amplifier which pulls the frequency compensation node down; the output of the internal 723 error amplifier is designed to accommodate this.  I would keep the existing Vbe based current limiter and set it for maximum overload current but that is up to you.

[Gurumurthy]

Sir, It would be meaning full and I will get an insight if circuit is given.... Sorry sir i am new to power supply design.

[coromonadalix]

http://www.circuitdiagramworld.com/power_supply_circuit_diagram/0_to_40V_Lab_Power_Supply_LM723__2616.html

http://www.learningelectronics.net/VA3AVR/circ/ps3010/ps3010a.html

[oldway]

Look at the solution choosed by Philips in his triple power supply PE1542...(schematic fig 1001).

http://www.analogdesign.be/files/gear/Philips_PE1535-PE1545_Operating_Manual.pdf

[Gurumurthy]

Thank you sir. I need some practical circuit to build my power supply.

[oldway]

Thank you sir. I need some practical circuit to build my power supply.

Did you look the operating manual of the link I gave to you ?
At the end of this operating manual, you have the schematics with the values of all the components.
Is that not a practical circuit to build your power supply ? 

[Kleinstein]

The Phillips circuit looks very reasonably, and it is a tested circuit.

Just keep in mind that the pin numbers on the 723 are for the metal can. Today one will likely get the DIP version. So a little work finding the corresponding pins is still to do. The BDY20 used can be replaced by a more common 2N3055 or similar. The only thing a little confusing is finding out which pots are for setting the voltage / current by the user and which are just trimmers to set the voltage / current range.

[oldway]

The trimpots are indicated by a sign that looks like a hammer, the potentiometers do not have this mark.

[Gurumurthy]

Hi Sir,
        Can you please explain me this circuit. I cant understant Current Limiting Circuit.

[oldway]

Off topic....See videos of Dave about power supplies....

[BravoV]

Is that not a practical circuit to build your power supply ? 

Off topic....See videos of Dave about power supplies....

The OP is well known to spam the whole forum again and again with new thread on the same subject, when he didn't find any schematic that he likes, or no one answer him within a day. Looks like he doesn't like circuit that has too many components, or circuit that uses components that are not in his drawer.

Just read his past created threads. 

Also I did ask him a simple yet very important question on his 1st thread, but it seems like he ignored me. 

I'm guessing that what he want is only a circuit that consist of max 2 transistors and few capacitors and a LM317 to build an adjustable 0-30V and 0-5A complete with with CC and CV feature.

Watch for his upcoming newly created thread as he doesn't like to use the 723 anymore. 

[Gurumurthy]

 Hi bravo,
            Yes bravo , I want to  build power supply with integrated circuits than discrete devices. But I know that power handling capability will be small.

        I want the circuit which will give understanding , so that I can trouble shoot the circuit when problem arises.


         I need the simple circuit which will control the current. The circuit should teach me how to control current. I don't want exact circuit with lm317 or lm723. So that I will be knowing what is happening and what I am doing.

       Mr. Old way has referred nice manual. Sure I will go through it.

      When the problem is small, we can find the solution easily rather than blinking with huge chunk. Hope u will understand sir. Correct me if I am wrong...


     Lot of thanks to mr.bravo , mr old ways and all those good heart who replied to my post.

     

[oldway]

I think that Gurumurthy tries to do the squaring of the circle! 

Already, a linear supply of 0 to 30V with a max current of 5A, this represents a dissipation of 35V x 5a, that is 175W, it is totally unrealistic.   

You will have to use a way to reduce dissipation as a pre-regulator or switchover of transformer taps.

And in addition, he wants a simple circuit ....

I think he should open a topic in the "beginner" section and start by studying a little bit of electronics first.

It is clear that he does not know what he wants, nor if what he wants is possible or not ...   

There is no shame in being a beginner, we are here to help them, but at a minimum, the person should realize that he is a beginner and to have enough modesty to recognize it.

[Kleinstein]

Adding current limiting to a regulator like LM317 or LM350, that is not made for it is usually not a good idea. So explaining the function of those circuits does not make much sense as they don't work well anyway. Dave's video about building a supply around an LT3081 (or similar) has the same problem to start with. So if you want to build a lab supply, forget about LM317 and similar chips with integrated power stage.

The advantage with the 723 circuit is that understanding of the circuit is relatively easy, as there is no magic black box with not so well defined properties. The Phillips circuit is still relatively simple (not too many parts), though it has quite a lot of background behind it - so a good example to learn from. So it is really worth trying to understand that circuit.

A supply circuit to learn from should be more like 1 A - the parts a cheaper and less smoke to escape. For learning, the small one is just as good as a larger one.

[David Hess]

Can you please explain me this circuit. I cant understant Current Limiting Circuit.

The LM317/LM338/LM350 wants to maintain 1.25 volts between its output and adjustment terminals.  R1 could be made to be adjustable however potentiometers which will support 1.5 amps and more through their wiper connection and with the necessary power rating are very rare.  So instead, R2 is used to add a voltage to the voltage across R1 and the LM350 acts to keep the sum of the voltages equal to 1.25 volts.

The 2N5640 operates as a constant current sink holding the voltage across R2 at 1.25 volts.  As the potentiometer is adjusted downwards, the voltage across R2 replaces the voltage across R1 and the output current decreases.

[oldway]

It is not possible to use a LM317/LM338/LM350 for a 0-30V 0-5A labo power supply because it exceed the absolute max ratings of the device.
input-output differential voltage: +35V max.
Dropout voltage: 2.5V at 3A (-25°C)
Min input voltage at max current and min mains voltage: 32.5V
Mains voltage variation: +/-10%
At max mains voltage (+10%) full load: 32.5 +20% = 39V
At max mains voltage no load: above 40V....exceed the absolute max ratings of 35V

Max power dissipation at full load and max mains voltage with output in short circuit : 39V x 5A = 195W 

It far too high, you have to use a tap switching by relays or a preregulator.....

The Philips power supply is only 0-20V 0-1A.....

[Kleinstein]

The Phillips instructions include a view slightly different versions: one for 0-7 V and higher current and also one that is for 20 V and likely about 2 A. But it is relatively straight forward to change that.

For a high power at least transformer tap switching should be used - but high power should only be version 2 or 3.

[oldway]

Many engineers have been working on the project of a 0-30V 0-5A linear lab power supply and the result of their work is what is found on the market ....

Are they all fools or stupids ?
Do you think you can do simpler, cheaper, more reliable ?

[Ian.M]

The Fig. 1000 circuit (page 20 of Philips PSU manual, Oldway's Reply #7 link)  is about as simple as it gets fror a real PSU that wont let the holy smoke out the first time you connect it to a difficult load.

The specs on page 5 say: PE 1535-00  0V - 40V 0A - 0.5A.

Beefing it up to a 5A supply would be possible with a better pair of pass transistors and a preregulator feeding the pass transistors.  However that increases the complexity again . . . . .

[oldway]

These Philips power supplies were highly appreciated for their robustness and ease of use.

However, this kind of circuit only applies to limited output power.

For larger power (400W and 1000W), Philips used a thyristors pre-regulator, but their schematics had major design shortcomings and errors, so I do not recomend to use their circuits.

https://elektrotanya.com/philips_pe-1642_1644_1646_1648_0-20-40-75-150v%2C0-20-10-6-3a_400w_bench_psu_1987_sm.pdf/download.html

[Kleinstein]

The circuit diagram for the HY3002 has one obvious mistake in the drawing: the emitter of V26A should be at ground level.

To make it cheaper one could use an LM324 for all the OPs and maybe leave out the extra TL431 reference and use the 78xx instead. Chances are one could eliminate the -6V or -12 V internal supply. Still this would be only marginal savings.

The voltage regulation looks a little odd. The way of adjusting the feedback divider usually means the loop gain depends on the voltage setting and thus a compromise type performance. It might lead to instability for the combination of low voltage setting and highly capacitive load.

The phillips circuit for the 7 V/3 A version with two transistors in parallel could be a problem as show - just add emitter resistors and it should be fine. A slightly higher power version (like 30 V 2 A) should be OK -it just needs a lot of heat sink.

[Gurumurthy]

Dear oldway,

                  True sir, as you said i am a novoice to Practical Electronics. I want to learn power supply design by simple ICs like LM317 and 723 because they cost around 10 rupee, so i can afford to buy these Ics rather than LT308X series ICs. When some smoke comes from my breadboard, It is my money which is wasted. I have learnt certain basic things with LM317 and 723 and TIP142,2N3055,Mj4502 Transistors when things went wrong.This is my study for past 45 days.


                     Now i understand that i have to move to SMPS preregulator to save my Circuits. I am ready to learn things and apply that in practical circuit design. I am more inspired my the way you approach things . Hence please guide me. Give some reference book so that i am strong to handle big Circuit similar to Philips power supply circuits. Not based on university exam. Please help me!!!!!!!!

                         Forgive me if iam wrong...

[oldway]

If you don't want to waste your money with "magic smoke", you should start with low voltage, low current power supply projects, not a 30V 5A power supply.

For exemple, you could make a very reliable double linear power supply 0 to +/-25V with a LM317 and LM327.

Current limit will not be ajustable but will be safely limited under 1A.

Now, a question for the people who are looking at this topic.... 

I said there was something horribly wrong with the project of the Philips power supplies with pre-regulator as pe-1642....

The problem is with the pre-regulator....

Somebody found what's wrong with it ?

[MagicSmoker]

...
I said there was something horribly wrong with the project of the Philips power supplies with pre-regulator as pe-1642....

The problem is with the pre-regulator....

Somebody found what's wrong with it ?

No inductor in series with the output of the half-controlled bridge.

[oldway]

...
I said there was something horribly wrong with the project of the Philips power supplies with pre-regulator as pe-1642....

The problem is with the pre-regulator....

Somebody found what's wrong with it ?

No inductor in series with the output of the half-controlled bridge.

No, there is an inductor in the primary to limit the peak current....that's something more complicated....

[Ian.M]

I've just put together a LTspice sim of the regulator section of the Philips PE 1535-00  0V - 40V, 0A - 0.5A. single output PSU.

Some of the models for the discrete semiconductors are a little rough (heavy use of ako:), but its good enough to see what's going on with the current limit circuit.   I didn't bother with the transformer + bridge or the meter circuit.

Part numbers are as per the Fig. 1000 circuit (page 20 of Philips PSU manual, from Oldway's Reply #7 link). Edit: Link Rot! As of 2020 it can be found [here].

Controls and Adjustments

• R1 - Set output voltage
• R2 - Set current limit
• R104 - Trim max current limit (500mA)
• R108 - Trim zero current limit
• R123 - Trim zero output
• R127 - Trim max output (40V)

I've trimmed all the presets to sane values.  R108 is a little fussy, if you trim with the output unloaded it starts dropping the output voltage well before you reach zero current.  The practical lower limit for R2 will be a few tens of mA. 

I've parameterised the R1 and R2 pot positions as vset and iset respectively, for ease of sweeping them in the simulation.   All pots & presets take a parameter wiper which is the position as a number between 0 and 1.

Zipped LTspice simulation attached - Unpack it to a subfolder of your LTspice working folder and enjoy!    I *think* I've packaged all the custom models it needs - if anything is missing ask.

N.B. The dp() function defined on the schematic wont be seen by the .op data labels unless you copy it to your plot.defs file.  See https://www.eevblog.com/forum/blog/eevblog-516-ltspice-tutorial-dc-operating-point-analysis/msg1383565/#msg1383565 for details (and a better function for limiting displayed data precision).

[oldway]

Ok, i will try to explain the problem: that's all about phase control and transformers.

To vary the secondary voltage by varying primary voltage with phase control triacs or scr's, that's not recomended.
Why ? because a very little dyssimetry of primary voltage will produce a dc voltage, and this dc voltage will let a hight dc current flowing in the low resistance of the primary.

The transformer get saturated and a very high ac current flows in the primary....this will blows up the fuses and can even damage the transformer.

But you will answer:the scr's are in the secondary, then there can not have any problems....

But is this really a secondary phase control ?

[MagicSmoker]

...
Why ? because a very little dyssimetry of primary voltage will produce a dc voltage, and this dc voltage will let a hight dc current flowing in the low resistance of the primary.

Ah, yeah. Also the inevitable asymmetry in the SCR conduction times or their forward voltages can set up a DC bias in the transformer, too. I was so hung up on the current integration inductor not being in the right place I hadn't thought of any other problems.

[oldway]

But indeed the inductor is not at the right place ....It can't be in the primary because any asymetry of the secundary current will apply a dyssimetrical voltage on the primary of the transformer....And asymetrical current occur during transients, when load vary rapidly....

This is why this power supply is basically unstable and the pre-regulator has a tendency to oscillate like a hell, saturate the transformer, consume a primary current of 30 to 50A and blow up the fuses and circuit breakers.
With the years, it gets worse because the capacitor C27 loses its value and then there is no way to prevent it from oscillating

Wat would be the solution of this problem ?
There are two solutions:
- to place the inductance not in serie with primary of the transformer, but in serie with the secundary of the transformer.
- no need of inductance if you use a transformer with high short circuit impedance like MOT.

R108 is a little fussy, if you trim with the output unloaded it starts dropping the output voltage well before you reach zero current.  The practical lower limit for R2 will be a few tens of mA.

Current limitation of this power supply is very basic and not intended to work as a precision current source.
It only use a single diferential amplifier and gain is very low.
But it is very stable for this reason and can be very fast ....
The greatest problem of this power supply (0-20V and 0-7V) are the shunt resistors: they are made with thin wires and prone to failure.

[Gurumurthy]

Dear oldway,
                I requested yesterday for  nice reference book so that i can learn things and design. so please give me some advice to design circuit.

[oldway]

http://www.ebay.com/itm/Power-Supply-Board-Kit-PCB-Based-on-LM317-amp-LM337-IC-/260484989112

http://www.circuitstoday.com/few-lm317-voltage-regulator-circuits

[BravoV]

Lot of thanks to mr.bravo , mr old ways and all those good heart who replied to my post.   

Alot of those good hearts people helps which are so sincere got wasted  ,  just because you want an easy circuit and keep ignoring those good replies.

What happened to your old threads ? Why you abandoned it and create a new one every time there is no replies or replies that you didn't like ?

[Gurumurthy]

Dear Bravo,

               These replies Made me to learn electronics. For Sure i have not wasted the energy of all those who have given there knowledge of designing circuits.

                 But my knowledge is not sufficient to digest these stuff. So i want to take some time to learn and design the circuit. No new Post will come here after.

                   I want good literature survey on Regulated Power supply design and Good Books to study Practical electronics. I already have Electronics Principles By Albert paul Malvino. Guide me in this regard...

 You can all suggest me good Books to learn "The ART OF ELECTRONIC DESIGN". Thank you all.... Regret inconvenience..

[oldway]

Thousands of circuits: encyclopedia-of-electronic-circuits 7 volumes.....
For exemple, ....
http://docslide.us/documents/graf-encyclopedia-of-electronic-circuits-vol-6pdf.html

[Gurumurthy]

Dear sir, 

          Give me a book which could give nice description of principle of operations of discrete power supplies and integrated circuits linear supplies.

[oldway]

What about you to do something for yourself and to look on Internet after such a book? 

[Gurumurthy]

Dear sir,
            Internet may give data at any instant but a book has  a starting and ending. So mostly I should have hard copy of any book. Next time I should refer this to others.

[oldway]

Some basic informations on linear power supplies........

http://sound.whsites.net/power-supplies.htm

[Gurumurthy]

Dear All,

              here is the link to download the circuit book.
https://archive.org/details/EncyclopediaOfElectronicCircuits

www.onsemi.com/pub_link/Collateral/HB206-D.PDF

[MARSHALBSB]

Thanks man. I'm loving the lecture!