Switchmode Power supply.

Hello there, Im having some trouble with a PWM circuit i made to drive a lab power supply that i was making, But ive ran into some problems.

The power supply unit is just a basic computer power supply with all the rails (Ground,+5,+12) wired together (sepretley).

I decided to control the power output with a PWM device, A capacitor and a coil. The PWM device ranges from 12v to 24v at 10-20A.

So far ive had no luck, ive soldered the circuit up, but its been malfunctioning. Sometimes it outputs 10v from a 12v source, and sometimes its 0v, And on the occasion it would work, but vary it by 0.7v.

I re-soldered it, just incase i had dry joints, but that wasnt the case. I dont think any of the components would be damaged, I soldered the IC socket on first, then the mosfets then the caps followed by resistors. I dont see much potential for compnent failiure.

Thanks in advance, Nitrousoxide

I think you're going to need to post a schematic before anyone can help you.

I apologize for the inconvenience, but ill upload a schematic. They (jaycar staff), insist that its due to faulty components, but i think not.

Im having some trouble with a PWM circuit i made to drive a lab power supply that i was making,

That suggests you are making a power supply but that schematic is a PWM motor control. Then you go on to say:-

I decided to control the power output with a PWM device, A capacitor and a coil.

Not sure where the coil is on that diagram. If it is a power supply based on that circuit then it won't work as you will be trying to use shunt regulation with nothing to limit the load and chopping the ground.

Sometimes it outputs 10v from a 12v source,

How are you measuring this with a meter? If so you won't get any decent results because you are trying to measure pulses and your meter is not set up for that.

Have a look at:- http://www.thebox.myzen.co.uk/Tutorial/PWM.html

Thanks mike for the response, They just said i could use this circuit with a capacitor of 10,000micro farad across the positive and negative terminals to smooth it out, in addition with a coil for positive out (hooked onto the cap), so in turn, shouldnt i be able to read the voltage changes?

"If it is a power supply based on that circuit then it won't work as you will be trying to use shunt regulation with nothing to limit the load and chopping the ground." Could you please elaborate and explain that thought?

The staff at jaycar said that i could use a motor speed controller for this purpose, because it was pwm.

The staff at jaycar said that i could use a motor speed controller for this purpose

Then they are idiots.

That circuit works as a PWM motor control by sinking current from the motor in a PWM manor. That is turning it on an off rapidly.

To make a bench power supply from a PWM controller you have to have one that sources power to the motor in a PWM manor.

The control you have with this circuit is called shunt control, that is the current you don't want is shunted (or shorted) to ground. This is not only very inefficient but it means you have to have something to limit the shunting current.

Even with sourcing current PWM in and of itself is not a very good way of controlling a bench PSU because you have no voltage feedback. That means the load current will effect the voltage and so the supply will not be very well regulated.

So you saying that its not possible to use this pwm module to vary the power output of my psu. If thats the fact, may i take in some new suggestions for a control circuit, even if its modification of the pcb i currently have.

Ive fiddled with it for a bit, but, i think ive destroyed it even more, it now shorts out and triggers my computer power supplies safety cutoff function. even when the pwm module is hooked up to nothing, as soon as i connect it, it shorts (also the duty cycle doesnt affect if it shorts or not).

Heres a link for a set of images of my setup: http://img691.imageshack.us/gal.php?g=1258110924380f.jpg

So you saying that its not possible to use this pwm module to vary the power output of my psu

As it stands yes, it is not possible to use this circuit for that purpose.

may i take in some new suggestions for a control circuit, even if its modification of the pcb i currently have.

You would have to modify that PWM circuit so that is was a current sourced PWM. This would involve having a series FET (between the input voltage and the output) and driving that from the TL494.

However, even if you did that it would result in a very poor power supply with poor regulation. Regulation is how stable the output voltage is as varying current is drawn.

To improve the regulation you need to feed back the actual output voltage produced and compare that with a reference voltage of the output voltage you need. Then the difference between the two would be used to control the value of the PWM output. This would form a closed loop feedback system. It is probable that none of the components you have in your existing circuit would be suitable to give you the speed of response you need in that feedback loop. You are then in the realms of a switched mode power supply and those are difficult to design and implement requiring a good PCB layout and good electronics knowledge. What you need is a DC to DC switching converter, this isn't it and probably can't me made to be it. Sorry.

Oh wow, ok, looks like ill have to start all over again, Again. do you have any schematics you can recommend for my purpose? Because im quite lost here.

Say if i wanted to use the circuit as its labeld use. I keep having problems at even getting it to work. "Ive fiddled with it for a bit, but, i think ive destroyed it even more, it now shorts out and triggers my computer power supplies safety cutoff function. even when the pwm module is hooked up to nothing, as soon as i connect it, it shorts (also the duty cycle doesnt affect if it shorts or not). "

Please tell me that you wern't saying that this whole circuit is completley useless?

It is likely that Q3 and Q4 are blown short circuit.

The rest of the circuit probably works and could be made into a motor speed controller.