[SOLVED] Problem with SMPS - No Output Volts

[kkeeley]

Thanks DXNewcastle for your support and ideas,

I'm not sure at this point what I can do with the transformer in the way of repairing it, I might be able to carefully remove the outside layers of tape and see if I can access the end of the damaged wire. At this time I'm going to hold off and wait and see what if any suggestion come along as to what to do. I don't which to damage or change the characteristics of the transformer as this may make matters worst.

I have got access to some old PC power supplys, so I will try to find one that uses the same style circuit and look to get the transformer out of that. The only problem may be that the setup of the transformer could be different which in turn will mean I have to redesign the circuit to work with the new transformer.

If I can get help on how to obtain all the information I need on the exisiting transformer, either by calculations or by the carefull disassembling of the transformer I might have a go at rebuilding it. I need to find out more about how to build my own trnasformers first though. Is it really that hard if you have the information you need?

Thanks for your help againg.
Kenneth

 

[FvM]

The transformer parameters are core size and material, air gap, winding turns and wire gauge, required layer and inter-windings isolation. You should be able to determine it even from a defective transformer, as longs it isn't fully burned.

I understood now about the relation of the two schematics, you're of course right, that the auxilary 5VSB converter from post #5 is exactly indentical to your circuit. You have also the windings polarities correctly marked. I also agree, that the replacement of low voltage transistor C945 isn't critical, any small NPN transistor (e.g. 2N3904) will fit. Using a slow 3A diode (1N5408) for the RCD snubber isn't that good, i fear. A fast recovery 1A rectifier should be used in this place.

A problem of the circuit is, that if the transformer feedback doesn't work, the output transistor will be overloaded and burn very quickly

 

[kkeeley]

Thanks FvM for the information required for the transformer.

Now all I did to do is work out how to disassemble the damaged transformer without destroying it completely. What I have found out so far is, most people break the core in the process, so I'm looking for a way to dismantle the transformer that doesn't break or damage any of the parts.

I might have to start another thread about this and see what help I can find.
Kenneth

 

[kkeeley]

Hi Guys,

For those of you that are still interested I have had a great outcome. After some time doing research I was able to dis-assemble the transformer without damiging it. From this I was then able to take notes of as much information about the construction as I could. From this information and some purchasing from Ebay I have been able to re-construct the transformer. I have now placed the re-built transformer into my circuit and to my surprise the circuit is now working. The fan is 12v rotating and the LED on the 5v is glowing. I have run the circuit for a couple of minutes before switching the mains off several times and there has been no changes to the performance of the circuit. Nothing has been getting warm yet.

My only last question is how long should I run the circuit for to really know if things are working correctly? Is 2-3 minutes enough time for any faults to show?

Thanks to all the people that have helped me to resolve this problem.
Kenneth

 

[DXNewcastle]

Well done!
and thanks for bringing the good news.
How long to test something for is another whole subject. Destructive testing can find the actual breaking points very precisely, but you no longer have a working device-under-test! Non-desructive testing can confirm that nothing breaks after some less stressful tests. You'll be wanting that approach, I'm sure.

Leaving a unit running overnight is common, lightly loaded and well ventilated. After that, I'd suggest some heavy load tests for 30 mins at a run, with poor ventilation.
If you have the means to perform the tests, I'd also recommend repeating those with variations to the mains supply of, say, plus 10% and minus 20%.

A celebratory cup of tea should follow next.

 

[kkeeley]

Hi DXNewcastle,

Thanks for your suggestions on how to go about testing the circuit. I realise that this opens a whole new can of worms.

I'm not too happy about the idea of leaving the circuit running over night at this stage as I don't wish to burn my house down in the event that some part of the circuit fails durning the night. However I think I could run tests for serval hours during the day at the weekend while I'm home to check on it from time to time this should allow me to make sure it's operating as expected. Overall as this circuit is a direct copy from a working power supply I feel that as long as the compents haven't been damaged in the early stages of my build/testing then the circuit should work fine. As this circuit is part of a bigger circuit that is still to be built and tested, I think I'll start building the next stage. The final product will be built from all new parts so any parts that have been used or damaged so far will not be used in the final build, so component failure should not be an issue in the final product.

Thanks again,
Kenneth