Bench power supply with current limiter and voltage regulator

[Samyaza]

Dear friends. I want to build myself a bench power supply. I came across a nice power transformer that takes in 220v and put out 24v with a VA of 300. I need much amps because I want to use it also as a car battery charger. Also I want to put in a voltage meter and a current meter but my question is how can I make a variable voltage and a variable maximum current? I plan to use potentiometers. The majority of schematics I found are working with low amps. 300VA at 24 v means 12,5 Ampers and I didn't find any ic or other schematics dealing with such high current. Any ideas?

[Kleinstein]

For so much power one usually uses a regulator circuit / chip and external power devices. In the old days this would be something like a LM723 IC and several 2N3055 power transistors.

Usually chargers for 12 V car batteries are rather cheap today and the newer ones include a reasonably working control to stop charging when full. So little need to build such a thing.

If you really need so much current the question is whether a SMPS is the better choice. For so much power one would usually avoid a pure linear regulator or at least use a transformer with several taps to reduce the heat production.

One also has to take the power factor into account: the 300 VA mean 12.5 A of AC current. With a rectifier and large filter caps you get a DC voltage of about 31-33 V and a maximum DC current of about 7A without overloading the transformer. For more current one would need some kind of PFC correction or a switched mode supply.

[Floyo]

There are a couple of ways to go about making a high power linear supply.

First is just dissipating the excess energy into a heatsink, but at 300watts you are looking at a large heatsink and some fans. You also need a lot of output transistors in large packages to keep their
internal temperature reasonable. It's possible but not really a great option.

Then there is the thyristor pre regulator. This technique is a bit oldschool compared to modern switching regulators, but it has the potential to give very good performance if done right.
Forum member Blackdog made a design in this thread https://www.eevblog.com/forum/projects/very-low-noise-preregulator-for-benchtop-power-supply/msg626382/#msg626382 .
A commercial product that used this technique is the philips PE1540.

Lastly there is the switching regulator option, getting this to work right for a high performance bench supply at this power is tricky. The circuit board layout etc matters quite a bit.
Unless proper care is taken switching supplies also tend to be more noisy, but for charging a car battery that is not a big deal.

I'd go for the thyristor option in this case, you get lower power dissipation without extra transformer taps. This simplifies the thermals considerably, but the rest of the design is not as critical as a smps.

[Samyaza]

Thank you guys. I like the "Philips" implementation. But there is 20V 5Amps.. it will hold greater values?

[Kleinstein]

The regulator can be scaled up. The LT1083 might even work up to 7 A. Higher currents can be used with discrete transistors.
However the power factor with this pre-regulator is not good (don't know how bad it is at low voltages) - so the transformer might not like more than 7 A anyway, at least at higher voltage.

[Floyo]

If you design it to handle the dissipation. The most important part in modifying linear power supplies is making sure the output transistors stay cool enough and get sufficient drive current from the drive transistor. This depends on the beta of the output transistors you choose. Another important point is the transistor safe-operating-area, a device might specify being able to handle 20 amps and 60 volts, but not both at the same time, depending on time and temperature etc.

http://www.mikrocontroller.net/attachment/87663/37838_Netzteil_NT30V_10A_KM_um_1_.pdf
(the opamp symbols are rather peculiar in this schematic, beware)

This is a 30V 10A design from ELV, in German. They use darlington output transistors to keep driving all 6 of them simple. This design dissipates the full 300watts though, as evidenced by the large forced-air cooling tunnel on the pcb. If you'd combine this design with the pre-regulator you could scale back the amount of power transistors and end up with a nice supply.

There are lots of topics on this forum on how to design a powersupply, almost all variants have had some discussion at one point or another. So there is a wealth of information to be found already especially on loop compensation etc. If you build your own supply its likely it will oscillate at some point, so its good to know how to deal with that, amongst other things.

[Samyaza]

Ok I decided to use a SMPS power supply instead of a classic transformer. The parameters are 24V 10AMPS.
Now, how can I make a voltage regulator and a current regulator?
I found on the internet the LT1038 but they say is obsolete and is nowhere to be found.
What can I use instead?
Actually I need a voltage regulator schematic that can handle 10 Amps.. and another current regulator. Let's say.. "give maximum 2Amps" even if my source is capable of 10Amps. I found also the display to use in the circuit. Neat...
Those are my problems right now, voltage regulator and current (limiter) regulator..
I can't find any schematic. many of them are made for low amp.
Thanks.

How about this?
What do you think?

http://electronics-diy.com/schematics/1185/30V-10A-Variable-Bench-Power-Supply-circuit.jpg

[jeroen79]

More amps will be a matter of scaling up the rectifier diodes, thyristors and pass transistors.
And providing enough heatsinking.

You can copy older designs from Philips, Delta Elektronika (E030-3, E30-10) or HP.
AFAIK they all use the same basic system, just like the ELV above.
Elektrotanya.com will give you servicemanuals with schematics.

The LM723 supply you posted will get into trouble with the 500 ohm current adjustment pot.
Go ahead, calculate the voltage drop and powerdissipation when you put 10A through it.
And the the current limiter will kick in when the voltage between pins 2 and 3 exceeds .6V.

[Kleinstein]

For so much power IC regulators don't work well. So you will need separate power transistors and regulator circuit.
The typical circuits will have voltage regulation and current limiting. So you can adjust the maximum current to 2 A if you want.

The linked ELV supply above might be an option. Iit would still need an second small transformer.

[Samyaza]

Hello again.. i found this:
http://electronics-diy.com/electronic_schematic.php?id=981
Is using a Lm317 and can drive up to 10Amps. My switch mode power supply is 220V to 24V 10 Amps.. so I guess is perfect for voltage regulation.
Now I need one more idea for the Amp limiter..
What do you think about the design in the link? I think is amazing. Anybody tried this?

[rgawron]

What do you think about the design in the link?

It's Ok, but keep in mind:

It's output is 10A, 1-30V, let's assume that the input is 35V (needs to be a bit bigger).

You select output voltage to some low value, e.g. 1V. Input voltage is 35V, output is 1V, so the voltage on the collector-emiter of each transistor is 34V (35V=1V+34V). Since you want 10A current to your device, power loss on on transistors will be P=U*I=34V*10A = 340W, you have 3 of them in parallel so each will dissipate more than 100W (340W/3).

That's a lot of heat!

[jeroen79]

Now I need one more idea for the Amp limiter..

The design works because the LM317 sends current into the TIP41's bases.
Now if you could somehow divert that current...

[Kleinstein]

The linked LM317 / transistor circuit is not that great. I only regulates the output of the LM317 - so the output voltage does depend on the load current: No load could be about 0.5 V higher than at full load. Also the maximum output voltage is about 2 volts lower than the input.
At least it is reasonably easy to add current limiting: just a PNP single transistor that diverts the current from the LM317 directly to the output, if the voltage over the emitter resistors get higher than 0.6 V.

[Samyaza]

My input power would be 24 Volts.. not 35. So let's say 1V on the output. 24-1=23V  P=U*I=24V*10A=240W on 3 transistors means 80W. What if I add another 3 transistors in parallel? this would be like 40W each.
I know is not a great solution but at least it is a solution.

[Kleinstein]

40 W per transistor is still to much for the small TIP41 in TO220 case. So either use even more transistors of larger transistors like TIP3055.

[Samyaza]

I decided to use 6 transistors 2N3055 This will do!
Starting to make the montage.