Capacitor explodes on PSU

[Sparkymark]

Have you got hold of an incandescent bulb?.

Yes I have managed to get a bulb. I have checked that and no short. Lamp lights bright and goes off. O was thinking maybe something shorting sending voltage back along the negative and that's why the cap explodes but not sure why the diode is shorting as this is before the cap on the positive side literally just after the bridge rectifier.

 

[Sparkymark]

Remove the main switching device(s) - probably the ones bolted to the big heatsink, replace all the missing/defective components and power up. You should be able to measure the correct DC voltages which will 'prove' the main input side of the circuit is operational.

I have removed the heatsink and the 3 devices. 1 is a triac BTA06-600C and 2 mosfets ITA9N50A. I tested the triac on resistance mode and seems to test OK apart from when I put meter in 2M ohm setting I'm getting a reading but it's in the k ohms 270k across A1 to A2 and 270k across A2 to gate. Mosfets seem to test OK. I will remove the 2 opto's and test those too.

 

[Sparkymark]

OK so I have removed and tested the opto's, both are EL817V and both are measuring 27k ohms across the transistor side when led side is activated. I believe this is not right. Please correct me if I'm wrong. Does this mean both opto's have blown?

 

[kellys_eye]

Please correct me if I'm wrong. Does this mean both opto's have blown?

You are wrong.

You can't just measure across the output transistor. You need to connect its collector to a supply via a resistor and measure the voltage across the c-e junction, looking for a 'switching' capability.

 

[Delta Prime]

Hi all, I am a qualified electrician and recently decided to get into electronics

You got to slow down sparky!
You possess an in-depth knowledge of electrical building codes read blueprints and schematics & design electrical systems .
Component datasheets are your friend! Below is Page 8 of the datasheet. enjoy.

 

[Sparkymark]

You are wrong.

You can't just measure across the output transistor. You need to connect its collector to a supply via a resistor and measure the voltage across the c-e junction, looking for a 'switching' capability.

Sorry, I just thought you put diode test across 1 and 2 and the diode will light the photo transistor to activate it therefore closing the switch. I will do some more learning and come back.
Thank you

 

[Sparkymark]

I did have a reply on here a short while ago but now disappeared. It was a chap say about checking this and that. One of them was about checking the capacitor for leaks. I have attached a pic of the capacitor which happens every time I put a new one in which I have only done twice because there is no point doing it again until I find the fault. As you can see the capacitor just ejects its insides with an enormous bang. It also takes the diode with it.

 

[Martaine2005]

You obviously have a short circuit.
Isolate the secondary of the transformer to eliminate downstream circuitry. This can normally be achieved by removing the bridge rectifier. Then power it up.

 

[Sparkymark]

You obviously have a short circuit.
Isolate the secondary of the transformer to eliminate downstream circuitry. This can normally be achieved by removing the bridge rectifier. Then power it up.

Sorry I don't quite understand. There are 2 bridge rectifiers. A small SMD type that goes + to pin 1 of optocoupler and - to pin 2 of optocoupler. The main rectifier goes to 1st filter cap which is in series with 2nd one then off to a small inductor then the diode that shorts and the cap that explodes. If I remove the main rectifier it will isolate the entire thing apart from the ac that goes to the small board and the triac.

 

[kellys_eye]

the diode that shorts and the cap that explodes.

Clearly there's an issue with the main bridge rectifier - the big black 'slab-like' thing AND, quite probably, shorted switching transistor(s).

 

[kellys_eye]

and the triac

What triac?

 

[Sparkymark]

What triac?

BTA06-600C
Its something to do with the small board. And the wires that lead to the transformer. I have removed the rectifier and it seems to test OK. Both the mosfets test OK on a component tester. Both come up as N E MOS showing gate drain and source. Triac also tests OK on tester.

 

[Martaine2005]

This is why you should isolate primary and secondary.
Or HOT and COLD on some boards.
It eliminates upstream and downstream. Same as you fault finding a constant tripping RCD etc. you check the devices connected on that circuit or unplug ‘all appliances’ on that circuit first. Then check inside the sockets for loose or disconnected, shorting wires. It’s simply a process of elimination.
But nowadays, you stick your Fluke In circuit and just press buttons

 

[Sparkymark]

So how would I isolate the secondary without also isolating the primary. As I said the bridge rectifier goes to the primary of the transformer via the components that go faulty. Sorry if I seem clueless but just need to understand.

 

[Sparkymark]

Clearly there's an issue with the main bridge rectifier - the big black 'slab-like' thing AND, quite probably, shorted switching transistor(s).

I do have new bridge rectifiers so could swap it out and see what happens. Just don't want to go thru a ton of caps and diodes trying different things. I was thinking about powering on and checking the voltage at the point where the anode of the diode would be. Definitely not a short before that otherwise my current limiter would light up.

 

[Martaine2005]

That device is certainly a triac. But what is it controlling?.

bridge rectifier goes to the primary of the transformer

Transformers are AC, rectification via diodes produces dirty DC and then on to filtering capacitors.
You should have X2 capacitors too on the mains input. If they fail short, your symptoms will be the same.

 

[Sparkymark]

That device is certainly a triac. But what is it controlling?.

Transformers are AC, rectification via diodes produces dirty DC and then on to filtering capacitors.
You should have X2 capacitors too on the mains input. If they fail short, your symptoms will be the same.

There is 2 main capacitors 470uF 200v in series. First 1 directly from + on bridge rectifier and the thru a small coil/inductor then the shorting diode and exploding cap. The main caps test OK bridge rectifier tests OK. AC goes from the bridge rectifier to the small board and then to the triac and to the 2 wires that go to the transformer. The small board is removed in the 2nd photo. AC comes in on the right. Pic 3 is from the other side so AC in on left, 2 wires to transformer in centre and a track to the triac. Pin on right doesn't seem to go anywhere but not 100% sure. The cap that explodes is 400v 4.7uf.

 

[kellys_eye]

Sorry if I seem clueless but just need to understand.

There are plenty of online reference articles that explain the workings of SMPS devices and some that offer advice on repairing them. Your current knowledge is lacking and your pictures/descriptions don't help either.

From FIRST PRINCIPLES - the power comes in via a fuse and filter(s) to a bridge rectifier and smoothing capacitor arrangement when the DC that is required to be 'chopped' is derived. If you isolate EVERYTHING after the smoothing capacitors you should be able to measure around 310V DC across those capacitors.

Anything that 'explodes' after this 310V is delivered is as a result of the main switching component(s) being shorted.

The devices on the big heatsink seem to be only 'plugged in' to the circuit board - are they IN (and soldered extremely badly) or are they OUT?

Lastly, do you know what 'landscape' means (in terms of photos)?

 

[Sparkymark]

There are plenty of online reference articles that explain the workings of SMPS devices and some that offer advice on repairing them. Your current knowledge is lacking and your pictures/descriptions don't help either.

From FIRST PRINCIPLES - the power comes in via a fuse and filter(s) to a bridge rectifier and smoothing capacitor arrangement when the DC that is required to be 'chopped' is derived. If you isolate EVERYTHING after the smoothing capacitors you should be able to measure around 310V DC across those capacitors.

Anything that 'explodes' after this 310V is delivered is as a result of the main switching component(s) being shorted.

The devices on the big heatsink seem to be only 'plugged in' to the circuit board - are they IN (and soldered extremely badly) or are they OUT?

Lastly, do you know what 'landscape' means (in terms of photos)?

Yes thanks, I do understand what you have described about the incoming AC and all that. I have new mosfets coming but they take forever to get here so have just put the heatsink back on the board with them attached and not soldered as I intend to replace them when they arrive. I still have to replace the diode and 400v cap too. I am learning all the time thru reading and watching others but at the moment like you said my knowledge is lacking. That's why I am here. I don't want to give up. Thanks for the input.

 

[Sparkymark]

Lastly, do you know what 'landscape' means (in terms of photos)?

Not sure what your point is, as long as the picture shows what I'm talking about I don't think it matters too much does it.