230VAC to 1.5V-24V adjustable 0-16A current. Which chips do I need?

[king.oslo]

hi,
If you are into designing circuits of various types, I would suggest downloading the free LTspice simulator.

In that way you can try out different circuits under different conditions and temperatures.

There are number of LTS user on ETO who would be happy to help.

Thanks for that information. I will do what you suggested. However, I thought you insinuated that the circuit i posted wasn't ideal, so I wanted to ask how it may be improved. Or at least in what area or respect it needs improvement. Thanks.M

 

[ericgibbs]

Thanks for that information. I will do what you suggested. However, I thought you insinuated that the circuit i posted wasn't ideal, so I wanted to ask how it may be improved. Or at least in what area or respect it needs improvement. Thanks.M

hi,
Look at this pdf for options.

EDIT:
Skip that, misread the spec....

 

[KeepItSimpleStupid]

The 723 circuit I posted should become a good basis for a design. The one thing I dislike about it and about most power supplies is th inability to set the current without shorting the leads. Increasing from 10 to 16 amps would require a little tweaking. Mainly, the number of paralleled transistors need to be nearly doubled. Some tweaking in the current limit section and possibly the additionof course ans fine voltage and current controls.

LTSpice MAY be able to simulate the circuit.

 

[Nigel Goodwin]

The 723 circuit I posted should become a good basis for a design.

I would agree, it's a fairly standard 723 based design - funny how the 723 has never been replaced by anything more modern though?.

 

[KeepItSimpleStupid]

I looked through that article and it's impressive if you look the other pages.

Two Hammond 165S30 transformers will do: http://www.hammondmfg.com/165.htm They are 30 V 10 A units. Put two in parallel to get 20 Amps.

They made the 723 surface mount, so it looks like it will stay for a while.

I bought a +5, -15, +15 supply at about $15 that's 723 based. It wss part of something. It gets upset (dies) when a reverse voltage is applied. Some commercial power supplies will tell you to put reverse protection diodes on the output because they are not included in the design.

I need to find that junk place again, 30 miles away, and pay a visit. Time to use the GPS.

 

[king.oslo]

Hello again,

I've been looking at the datasheet for the 723 and the circuit **broken link removed**.

I am beginning to understand how current-limiting work using wikipedia, but using the 723 datasheet and circuit I fail to understand how current protection works in the circuit above.

I understand that the input on pin 2 on 723 goes to the base of a transistor inside the 723 and takes away current from the series pass transistor that ultimately lowers the voltage of the big 2N3772 transistors, but I cannot see how this happens, as I would expect current to flow through D3 power diode and not to R12 when the current gets very high.

I was hoping somebody could explain this!

Thanks lots!! M

 

[carbonzit]

I understand that the input on pin 2 on 723 goes to the base of a transistor inside the 723 and takes away current from the series pass transistor that ultimately lowers the voltage of the big 2N3772 transistors, but I cannot see how this happens, as I would expect current to flow through D3 power diode and not to R12 when the current gets very high.

Look at the schematic again. Current does flow through D3, but it also flows through R12 (in series with R13, R14, and R21 and the zener D2. Most flows through the diode (that path has much lower resistance), but a sample flows through the voltage divider R12, thereby developing a voltage into the 723 transistor base. You don't want much current there, but it has to be a fraction of the output current.

 

[KeepItSimpleStupid]

D3 is there for reverse polarity protection. i.e. connecting the PS to a powered circuit.

 

[carmusic]

first of all this type of power supply need switching for efficiency, you can use transformer to get 24-30vdc then use a switching controller as the lm25116 or other similar. it will be very hard to get 16A at 1.2V with this type of controller since peak amp from pwm will be extremely high (over 200-300amp!)
other option would be multiple switching regulators (one for low voltage, one for medium and one for high)

or if you want you may buy this type of power supply (already use them)
https://www.aliexpress.com/product-...-watt-switching-power-supply-wholesalers.html
those power supply can vary output from 4V to 24V at 33A!

 

[king.oslo]

Switching for efficiency? Remember you are dealing with a newbie here. Surely the voltage regulator only consumes a fraction of a watt? how can the efficiency be critical, if it is say only 50% efficient?

And why will peak current be 300A? I only need 16A. If I have 10 transistors, will they not be 1.6A each? And which pwm are you talking about. Will the 10 transistors not constantly conduct current, and not in frequencies?

I am probably all wrong because I am new at this. But when you explain these things to me, I will not be able to learn your great skills unless you teach them like you would to a baby

Thanks for your time!M

 

[KeepItSimpleStupid]

What carmusic is trying to say is that switching power supplies can be from 85 to 98% efficient. The LM323 based supply is low noise, linear. Depending on what's it's asked to deliver, determines it's efficiency. Switching regulators typically use the collapse of inductors to regulate. The inductors are pulsed with a varying in time DC signal, thus trying to not waste energy.

If you think about a 31 V unregulated supply asked to provide 1 V at 10 Amps, (31 V - 1V) * 10 is 300 Watts. This is going as heat. If the same supply is asked to deliver 0.1 A, then (31 V - 1V) * 0.1 A; then 30 W is wasted as heat. Typically a little less, but you get the idea.

For a bench supply, linear isn't necessarily bad, but it will be BIG.

 

[king.oslo]

Thanks for that.

But surely the 85-95 per cent efficiency is only for the regulation IC of a few mA, and not the transistors, which actually deliver the 10A?

Thanks.

Kind regards,
Marius

 

[KeepItSimpleStupid]

Nope, the whole package. Efficiency is Pout/Pin * 100 Pin is AC Mains Power. Pout is DC Delivered Power.

You asked for it: https://www.electro-tech-online.com/custompdfs/2011/10/HB206-DPDF.pdf

 

[ronsimpson]

Switching power supply board from National Semiconductor. Should work with a change to the feedback circuit.

 

[king.oslo]

Nope, the whole package. Efficiency is Pout/Pin * 100 Pin is AC Mains Power. Pout is DC Delivered Power.

You asked for it: https://www.electro-tech-online.com/custompdfs/2011/10/HB206-DPDF-1.pdf

Thank you, I read through it this evening. However I could not find a straight answer to my question, perhaps I overlooked it?

Anyway, I think I may have been able to cook up an answer, and I want you to correct me if I am wrong.

Is it correct that my entire switch mode power supply will be more efficient using a switch mode controller because the transistors it is driving will be either on or off (due to the square wave output), and not half open. This reduces heat (and efficiency loss?)

Please correct me if I am wrong. I really want to learn this!

Thanks.

Kind regards,
Marius

PS: RONSIMPSON, I havn't looked at your paper yet, but I will now. Thanks for taking the time to help me.

 

[ronsimpson]

Yes you understand a switch mode supply is on or off. There is voltage at 0 current and current at 0 volts there fore the power loss is small.

 

[carbonzit]

Is it correct that my entire switch mode power supply will be more efficient using a switch mode controller because the transistors it is driving will be either on or off (due to the square wave output), and not half open. This reduces heat (and efficiency loss?)

As Ron indicated, you've hit on the essential difference between switch-mode and linear supplies. Switching at (as near as possible to) zero voltage and zero current means that the switching devices (transistors) aren't dissipating power, but only transferring it between the supply and the load. In the real world, we take high efficiency to mean anywhere in the range 80-95%. Much better than a linear supply, which must throw off any current it needs to keep from the load in the form of heat.

 

[king.oslo]

Thats great! I feel great to finally understand! How great! M

 

[carbonzit]

Yeah, it's like a great, non-addictive drug. (Actually, it is addictive ...)