The importance of eyes fingers and mains

[Shock]

Story alert.

Working on a little SMPS that was given to me as dead (no output) and I plugged it in before opening it up and couldn't get any AC/DC voltage readings.

The device the SMPS powered I confirmed was ok by using the bench supply initially current limited, found normal operation at about 1A and since the SMPS was 3A, thought that checks out and it's back to crack open the SMPS.

The SMPS pcb was a bit of an eye opener. It had two tiny wires going to the 240V mains socket, one broke off or had already been broken before I got in, the earth connector on the mains socket went nowhere so that was a waste of time. The worst problem was lots of shitty soldering and a couple sets of SMD resistor bodges on the output. Convenient lack of protection and filtering on the mains input.

Anyway I tested the primary and the mains input through the fuse, bridge rectifier, every transistor and diodes both ways and any type of weirdness, most caps for shorts, low ohm resistors for opens. Hooked it up and measured the voltage of the filter cap that was good. Discharged it and took a quick check of the transformer and then moved onto the secondary.

Secondary looked like two diodes in a half wave setup, one filtering cap, a more elaborate control circuit than anything. Looking at the bodges one SMD resistor on the output was reading weird so I removed it and found it had a friend added underneath, added up to 1K so thought fair enough and put them back in. Checked optocoupler nothing shorted or open, checked output and not shorted.

I did find one free floating SMD resistor embedded in the solder on the positive of the output, wondering if it was stuck on the assemblers iron and they added it in. I tacked that open on another spot for later retrieval. Reworked a lot of the PCB as it was a mess and the joints looked really bad, too much solder on some of the bodges and a lot of flux before I got there.

Time for a test flight, hooked up the mains and measured the main filter cap - good. Checked for DC on the output and across the output cap - nothing.

Thought ok I'll load it down 10mA and see if it starts up because at this stage the secondary was dead and everything else looked good to go. Touched the load on the output a couple of times for a second or so it didn't kick in nothing seemed to be melting. Then left it on for a few seconds.

BANG!

Well she started up alright and this is the point and moral of the story, always wear eye protection. The fuse while it didn't explode in my face it was brutal and left a nice couple of inches of carbon flash at ground zero. I'm glad my fingers were down at the output end away from the action as well.

The next thing is I should have really used the dimbulb tester circuit for some current limiting, but things were looking so promising , it could have saved me having to now additionally find and order some micro sized fuses for it.

So you got this far to the end of the story? What could be the problem with the SMPS? What did or didn't happen and what did I forget? I have my own ideas of course but it makes for discussion.

[Shock]

Here is photo I took before I realized I should have taken a better photo. It's mounted in my Aldi cheap bench vice in case anyone is wondering. I'm excavating excess solder and just dug out the spare resistor. Primary at the top and secondary at the bottom.

(image removed)

[julian1]

Is it possible the SMPS wasn't actually switching/oscillating - leaving a short through the primary?

[AndyC_772]

Is it possible that it's just an unmitigated piece of crap not worthy of your time?

[Gregg]

You now have the perfect reason to toss it.  Maybe use a hammer to ensure all of the angry pixies have escaped and to give yourself some sense of satisfaction [be sure to wear your safety glasses or it may bite again].
I often find that the case for junk equipment can be the most valuable part when re-purposed.

[Frankje]

Power supplies can be nasty for faultfinding. I've spent about 20 years of my career repairing them.
What I normally do after the checks you did, is using a variable transformer ( which also separates mains from earth ). I use a dummy load on the output ( lamp bulb of proper wattage ). Then I gently apply mains voltage and use an oscilloscope to check if the primary oscillates. I also measure the output voltage to see if this regulates.
Now, for your problem, capacitors in the feedback loop might have caused the problem. With high frequencies, they tend to conduct when defective. So, they can shut down the regulation and BANG !
Sometimes tiny cracks in the copper lanes cause malfunction too.
Eye protection is really important when working on electronic circuits under high voltage...

[SL4P]

awk!
That photo looks like someone threw a  bucket of solder at the board as they  were walking past

[Frankje]

Hi,
I usually start from 0 Volts on the mains input of the smps. I connect the scope on the current measuring resistor at the primary mosfet to check oscillations when cranking up the variac until the oscillations start. Pay attention to ground the scope at live ground !
If they don't start at about 1/4 of the mains voltage, time to check the primary for faulty parts.
When oscillating occurs, I check one of the secondary voltages to see if this stabiles.
If not, check for faulty parts in the feedback loop.
Cracks in solder pads can occur with transformers, heat sink and power transistors.
They might not be visible and you can't measure them. But when high currents flow, they can act as open circuits.
Greetings