Power dissipation limiter

[Paul_J]

Wishing to build a variable DC power supply with current limiting I came across this problem: when the output is short circuited to ground and with the current limit set to maximum the series pass element (in this case a transistor) has to dissipate all the power so it requires a big heatsink. But I don't have room for such a large piece of aluminium in the case in which I'm going to build the power supply.

For example, my supply will deliver a voltage up to 24V at a maximum current of about 2.5A so if the output is shorted to ground the power dissipated in the series pass transistor will be 60W.

I need a circuit that limits the voltage drop across the series pass transistor.
Can anyone help me with this?  :-\

 

[audioguru2]

You can change the circuit to "foldback current limiting". When an overload tries to draw more than max current then the circuit reduces the current. For a shorted output the current can be very low. At any current below its max setting, the circuit performs normally.
Look in Google for Foldback Current Limiting Circuit.

Or you can add a high velocity fan to the existing heatsink.

 

[ante1]

Hi Paul,

Welcome to our forum.

The heatsink you need shouldn’t be inside the case anyway to be most effective! To help you more in detail we need the schematic to see what could be changed.

 

[Paul_J]

Hi guys,

Foldback current limiting would be a solution...
I know this method of current limiting from a book about voltage regulators. It has all the formulas for calculating the output short citcuit current...
How can I adapt it to this circuit http://electronics-lab.com/projects/power/011/index.html?
Will it work with a variable voltage power supply? ???
I doubt it.

If the output voltage is 2V and the current limit set to 2.2A and the circuit draws 2A the current limiting circuit won't be active so the dissipation will be 40W.
This situation is unlikely to be achieved but it can happen. 40W is still too much.

Another soultion would be to keep the voltage drop across the series pass transistor constant so the dissipation would be constant too at a given current.

One way to do this is to use a SCR after the rectifier but before the filter capacitor. The SCR will conduct only when the voltage across the series pass transistor falls under a certain value set by a Zener diode. Besides the SCR and the Zener diode the circuit has a low power transistor and three more resistors. I like the way the circuit works but I don't like the frequency at which it works - 100Hz.
I already thought of something better...  ;D

The heatsink you need shouldn

 

[audioguru2]

Your circuit doesn't have a voltage reference. Then the base-emitter voltage of Q1 is a thermometer and Q2 and Q4 boost its current!

I don't think the output voltage is regulated so it changes with temperature, input voltage changes and output current changes.

You could look at a datasheet for the LM723 regulator that I think has a circuit with foldback current limiting.

 

[Paul_J]

The LM723 does have a circuit with foldback current limiting...

How about using the other way to limit the dissipated power - the one with the SCR (thyristor)?

 

[audioguru2]

How is an SCR going to reduce the voltage or current? By shorting the unregulated supply?

I have seen power supply circuits that automatically switch taps on the transformer to lower voltages when the output voltage is reduced.

 

[Paul_J]

How is an SCR going to reduce the voltage or current? By shorting the unregulated supply?

No.
If the voltage across the transistor increases, the SCR, which is placed between the bridge rectifier and the filter capacitor, will stop conducting until the voltage decreases under a value set by a Zener diode. The variable voltage linear power supply will be connected after the SCR.
Regardless of what supply I'm going to use (regulated or unregulated) it's going to have an adjustable output voltage. :
I'll have to find a way to post the schematic of the circuit...

 

[audioguru2]

If the voltage across the transistor increases, the SCR, which is placed between the bridge rectifier and the filter capacitor, will stop conducting until the voltage decreases under a value set by a Zener diode. The variable voltage linear power supply will be connected after the SCR.

I see, good idea. ;D
The SCR will gate the pulses from the bridge rectifier. The frequency doesn't matter as long as it reduces the power dissipation.

I'll have to find a way to post the schematic of the circuit...

I use Microsoft Paint program to sketch schematics or to copy and paste whole schematics, pieces from schematics or symbols from other schematics or datasheets. Then I attach it here to my reply.

 

[Paul_J]

Here's the circuit.
I know it will work but I'm developing a better version which I'm not sure how well it works...

View attachment 39737

 

[audioguru2]

It looks like it will work! Good idea. ;D

 

[ante1]

How about Q2s base? Just left open? How does this work?

 

[audioguru2]

Q2 is the series pass transistor in our "Variable DC Power Supply" project, the one that doesn't have a voltage reference and so I call it a thermometer.

 

[ante1]

OK, so you mean this : http://electronics-lab.com/projects/power/011/index.html
That’s an interesting approach; I remember I have tried something similar once. I’ll see if I can find the schematic.

 

[ante1]

I found it, this one worked very nicely:

View attachment 39749

 

[audioguru2]

Hi Ante,
The circuit is complicated.
I bleached away most of the dirty paper towel that you had it printed on.

View attachment 39753

 

[ante1]

Do you think so; well it’s a complete PSU not just the pre-regulator!
It isn’t a paper towel, it was my scanner set on color on a gray scale pic! ;D

 

[smithy666]

Elektor published a dissipation limited back in 1991 - similar principle to the SCR circuit. I had built the Elektor Lab Power supply from 1990 (dual 0-40V, 0-5A) which employed a pre-regulator circuit on the primary side to reduce dissipation. I got the supply to work fine, but the transformers make a hell of a noise when a load is applied. I'm looking to replace the pre-reg circuit with a dissipation limiter on the secondary side.
I've posted the schematic - the links are so it can be changed between +ve & -ve versions (with transistor changes)

View attachment 39840