SMPS repair - breaker tripping but parts intact?

[de_light]

I'm trying to repair a cheap Chinese SMPS 12VDC 30A that I was using to power a radio.
It is similar to, but not completely the same as the schematics posted at http://imajeenyus.com/electronics/20151028_smps_variable_voltage/index.shtml

One day, on turning it on, the device failed. The glass 6.3A fuse had exploded. One of the 2SC2625 transistors had exploded, charring the housing and taking out a trace. An associated large diode near this destroyed transistor had also failed and the bridge rectifier package had also failed.

I replaced all of the failed components and repaired the trace with some large gauge wire. None of the capacitors had failed but the primary input capacitors (x2 200V/680uF) had a fairly high ESR so I replaced them anyway with slightly larger ones - 200V 800uF with a low ESR. I added in the missing thermal switch and soldered in a mounted 6.3A slow-blow fuse. All of the transistors were mounted properly with mica spacers and heat paste - they were just flapping in the breeze, not mounted to the case before.


I connected the supply up in series with a dim bulb tester. Initial output was promising - a smooth 13.8V (which is what I had originally set the voltage to using the adjustable pot). The dim bulb tester lit up briefly as expected for an inrush current but then went dim. Within about five seconds, I could hear some....fizzing and then a high-pitched whine that rapidly increased in frequency until my breaker tripped.

I disconnected the supply and opened it up. Nothing has appeared to be broken. The fuse is intact. The caps are not bulging and are measuring a normal capacitance - I haven't measured the ESR. I desoldered the new transistors and checked them and they are working fine. There is no smell, no charring, nothing.

Does anyone have any thoughts on how to diagnose it next?? The whining sounded like that of a transformer or an inductor - I'm not sure how the breaker tripped without any sudden increase in current? The bulb never lit up anymore, the fuse didn't blow. Quite confused!

[TheMG]

Is the breaker in question a GFCI (RCD in some countries) and/or AFCI?

If yes, then there likely was a ground fault somewhere within the power supply. Obviously with a light bulb in series with the power supply, there's no way it would have caused an over-current trip.

Possible suspects include the insulation between the primary switching transistors and heatsink/chassis (don't forget the special insulating washers on TO-220 type parts!), the interference suppression capacitors from primary to secondary could also be suspect in a cheap Chinese power supply (sometimes they are not proper "Y" safety rated caps, which could fail short). Lastly, the transformer itself could have an insulation breakdown between primary and secondary (often the transformers are not wound properly taking care to have proper spacing and insulation between windings).

[de_light]

Thanks. Yes they are RCDs.

I'll check the insulation between the transistors - that could be a problem.

Any thoughts on what would cause that odd fizzing sound and transformer whine rising in pitch shortly before the RCD tripped?

[themadhippy]

did you clean off all traces of skid marks on the board and case?

[de_light]

Is the breaker in question a GFCI (RCD in some countries) and/or AFCI?

If yes, then there likely was a ground fault somewhere within the power supply. Obviously with a light bulb in series with the power supply, there's no way it would have caused an over-current trip.

Possible suspects include the insulation between the primary switching transistors and heatsink/chassis (don't forget the special insulating washers on TO-220 type parts!), the interference suppression capacitors from primary to secondary could also be suspect in a cheap Chinese power supply (sometimes they are not proper "Y" safety rated caps, which could fail short). Lastly, the transformer itself could have an insulation breakdown between primary and secondary (often the transformers are not wound properly taking care to have proper spacing and insulation between windings).

Small update. I have rechecked all the insulation - I had missed off an insulating bushing off one of the switching transformers. It survived a dim bulb test with no evidence of a short and the breakers did not trip. However, it continues to have a fault condition that persists both on DBT and mains.

The output is LOW - 2VDC output (at least there's something, I guess). There is a loud, ticking sound that comes from the transformer. This seems to speed up and down at times - I guess this is coming from the PWM from the IC? A google search suggests that capacitor failure could be a cause. Do you have any other suggestions or tips for troubleshooting this?

[TheMG]

Could be going into hiccup mode (short/overload protection) due to a fault on secondary side. Could be a bad capacitor in the primary side too, often a small cap providing the initial charge to fire up the SMPS IC is the culprit.

What does the SMPS use for a control IC?

Was the power supply ever used under heavy load for some time? I've heard of some of these having aluminum wire in the transformer, getting too hot under load and the insulation breaking down between turns, consequently blowing the switching transistors.

Checked the output rectifier diodes?

[de_light]

Could be going into hiccup mode (short/overload protection) due to a fault on secondary side. Could be a bad capacitor in the primary side too, often a small cap providing the initial charge to fire up the SMPS IC is the culprit.

What does the SMPS use for a control IC?

Was the power supply ever used under heavy load for some time? I've heard of some of these having aluminum wire in the transformer, getting too hot under load and the insulation breaking down between turns, consequently blowing the switching transistors.

Checked the output rectifier diodes?

Well, this is totally bizarre. I've done a lot of troubleshooting today but ultimately, have arrived back at the same problem!
To answer your question, I would say it had a fair amount of load - up to 20A at 13.8VDC with a 50% duty cycle for a couple of hours at at time

I removed the dual Schottky rectifier packages and tested them - fine, not shorted, not open.
Removed the main transformer - fine, normal low DC resistance and not shorted between primary to secondary and neither to core
Removed ?base transformer - as above

Checked ESR of main filter caps, low. Capacitance in expected range given tolerances.
Checked all fusible resistors and diodes - none open, none shorted.
Checked the controller IC which is (probably fake) Fairchild KA7500B - this outputted a sine wave at 28kHz on pin 8 when 15VDC was supplied to the VCC pin.

Put it all back together, dim bulb tester. More odd fizzing - nothing. IC not receiving any voltage. Thought I had fried it. Turned out the second chopper transistor has failed short-circuit! I'm missing something here and can't for the life of me figure it out.

Excuse the condition of the board - it is getting tired now given the amount of soldering and desoldering that has occurred!

[perieanuo]

hi, I didn't hear my fav song, 'all eletrolytics were replaced' or at least verified
did i mentioned, i saw once in some TV power supply one electrolytic (the one filtering the pwm control IC supply voltage), when I moved a little bit his legs, the smps worked? so sometimes multimeter check is not so obvious, replace those "important" ones, we'll talk after
 in another life, a 32A disjunctor (without differential) tripped cause an emi filter caused (from time to time) short L-GND. the owner begged me to stay and wait for it to happen, after 30 mins I 'broke' the door to get home, BANG, 250Vca cap exploded, disjunctor tripped
in your case, the smps seems to misbehave from some evil capacitor i'd guess, but don't eliminate any possibility