Opinions on fixing a cheapo bench power supply I smoked

[Symbioquine]

Hi folks, first post here so be gentle...

I'm top balancing some LiFePO4 batteries and accidentally shorted -25v (@ possibly 260A or so) through a cheap 10A bench power supply I was using. I can easily write it off as a lost cause, but I don't like wasting things and would like to fix it if possible.

The device still powers on and responds to the dials, but now outputs about -2v @ 0.6A (with nothing connected to the leads) when set to the same settings that previously produced 3.64v @ 10A.

I've opened the power supply and don't see any obvious catastrophic damage other than signs of an arc between two pads - and possibly through one of the pads. What I'm looking for is some opinions about whether it's likely to be trivially fixable and - if so - what steps I should take.

My gut feeling is that the main thing is repairing the positive output pad in my second attached photo, but I'll readily admit I don't know enough about electronics to guess what else might have been fried.

Thanks in advance,
Symbioquine

[Symbioquine]

Here's a higher resolution photo of the pad I suspect to be the main point of damage.

[DaJMasta]

Could be worth an attempt - it's difficult for me to figure out at a glance with the stuff on the underside, but I would start with two things:
Cut out the carbonization between the pads and clean it up so there is either no board or at least no charred board, then add some solder to make sure the connections are good.
Check D30 (I think that's where those front panel leads go to) as it's likely the protection diode for the front panel connectors and it's probably been cooked.

If the diode is bad and with it replaced or just removed and the carbonization cleaned out and it seems to work, you're probably good and the frontend protection did its thing.  If not, it may be more challenging to figure out and may or may not be worth your time (sometimes even if a cheap replacement is easy to get, it can be worth it just to see how it works and test your troubleshooting skills, but it's all up to whether you think it's worth the time).

[bob91343]

On a slightly related subject, I purchased a Chinese power supply for $2 at a swap meet.  The seller said it was dead but for so little money I took a chance.

Two problems.  The first was obvious - the reverse polarity diode across the output was shorted.  The second, the supply then worked but the polarity was reversed.  It's laughable why.  The polarity was painted on the binding posts, the part you grab to fasten leads.  Someone had interchanged them so the red was the negative.

I have gotten a lot of use out of that little unit since repairing it.

One question remains.  The reverse polarity protection diode was rated at 3A and the power supply is rated at 2A.  Maybe that makes one assume it's a reasonable rating, but frankly speaking, the only time that diode would conduct would be if the supply were connected to another supply causing reverse current.  I am unsure about the philosophy of using a 3A diode.  I also see that it was rated at 600V but the supply only goes to 30 or 40V.

Oh well, it is a cheapie and I fixed it by replacing the diode with the same rating as the original.

[Berni]

Yeah this is the method of death for a lot of supplies.

I would try taking out the reverse polarity diode and checking the any of the power semiconductors near the output. This usually does not blow up everything so its worth a shot. Bench power supplies are one of the most DIY repairable pieces of equipment since they don't use much special chips, they are simple in design and mostly what blows from abuse is the power handling components.

Here at work i have like about 3 dead lab PSUs stacked in a corner, haven't got around to fixing them yet, all of them are probably a easy fix.

[Symbioquine]

Could be worth an attempt - it's difficult for me to figure out at a glance with the stuff on the underside, but I would start with two things:
Cut out the carbonization between the pads and clean it up so there is either no board or at least no charred board, then add some solder to make sure the connections are good.
Check D30 (I think that's where those front panel leads go to) as it's likely the protection diode for the front panel connectors and it's probably been cooked.

If the diode is bad and with it replaced or just removed and the carbonization cleaned out and it seems to work, you're probably good and the frontend protection did its thing.  If not, it may be more challenging to figure out and may or may not be worth your time (sometimes even if a cheap replacement is easy to get, it can be worth it just to see how it works and test your troubleshooting skills, but it's all up to whether you think it's worth the time).

This response was pure gold, thank you DaJMasta!!

I repaired the pad with some solder and tested the supply - no change.

Then pulled one pin of the diode and tested it with my meter; -0.1Ω one way and -0.2Ω the other way so it seems pretty clear the diode was fried.

With the diode pulled, the supply is working correctly again. I just need to find a suitable replacement diode now...

[DaJMasta]

I cooked one output of my SPD3303X once by accidentally connecting a car battery to it in reverse...

Similar damage without the carbonization, I actually heard the diode pop, while the front panel board was a little wedged in, I managed to get the diode out pulling it from the topside, then found a suitably beefy similarly sized and rated diode from the parts bin and swapped it in without even pulling the board.  Glad they build the basic protection into the frontends, makes the repair easy for a lot of user-error sort of failures.

[Kleinstein]

The reverse protection diode is mainly needed if one uses 2 supplies in series. The diode current rating would need to be high enough for the maximum current of the other supply.  A rating to something like 1.5 times the supply current and at least 3 A (as 3 A supplies are common)  makes sense. A larger diode usually does not hurt much.

It looks like the copper trace did it's job as a fuse.

[shakalnokturn]

The reverse protection diode is mainly needed if one uses 2 supplies in series. The diode current rating would need to be high enough for the maximum current of the other supply.  A rating to something like 1.5 times the supply current and at least 3 A (as 3 A supplies are common)  makes sense. A larger diode usually does not hurt much.

It looks like the copper trace did it's job as a fuse.

True but then the energy stored in the PS capacitors can change the outcome quite a bit.
One setup I had was for testing supposedly counterfeit power transistors for Vceo max avalanche voltage, my 120V 1A supply wasn't quite enough so in series with it had a dual 60V 3A supply, current limiting was at minimum. Before swapping transistors (with power supplies switched off) I shorted the outputs to be on the safe side. The capacitor in the 120V supply took one of the reverse protection diodes out in the dual 60V 3A unit.

[MathWizard]

Well it's a switching PSU, I'd say it looks like the row of caps on the left, are the output caps, then a filter inductor, and that great big diode is for reverse protection, right on the output.

Electrolytic caps hate reverse polarity, I'd check them for sure too. And the output of the main switching transformer and output rectifier, which feeds that row of caps.


I wish I had it here, I'd try for fun, but it would nice to get paid as well.

I was testing a transformer the other day, and by mistake shorted it, and wow a 1N5408, 3A diode, was toast in 1-2 seconds, the surge current rating is like 100-200A, but not for 1-2 seconds I guess. I actually had just put a 6.3A fuse on the primary side, and it got toasted too.

So check any diodes and semiconductors around there......if it's worth your time/effort.

[Kleinstein]

The white wires soldered to the bottom side look like an after-thought 110/220 V swtiching. The wires just soldered to the surface are unsafe, they may come lose and bring life voltage to the case.  Similar the capacitor added to the bottom, just under the transformer / isolation slots is not a good idea.

At least make sure the case is properly grounded and there are working GFIs when using this supply.

[Symbioquine]

The case is grounded (perhaps insufficiently) by way of the yellow wire in the lower right of the photo of the top of the board - the ring connector of which is screwed to one of the four standoffs which connect the board to the case.

In terms of surprising slightly sketchy things, I've also discovered that 2 of the three aluminum heatsinks on the lower left of that same photo are hot @ 120v AC. 

[Symbioquine]

I just wanted to provide an update here:

Towards the end of my LiFePO4 battery balancing, the power supply leads started sparking when connecting up different cells to balance them (this is with the power supply off). I measured with my multimeter and found that the resistance of between the positive and negative leads of the supply was around 160Ω compared to the 1.6kΩ of another (known good) supply that doesn't spark when used the same way.

At this point I'm going to write-off the supply in question as a brick unless folks have any further thoughts as to why the resistance would be so low or what tests/repairs I should attempt. I figure it can't be good/safe for the purpose of charging battery cells if it allows enough current to flow to cause big sparks connecting the turned-off supply.

Looking forward to folks thoughts and/or condolences on the dead supply...

[Berni]

The sparking is probably the battery filling the supplies internal capacitors. Tho the resistance across the output sounds a bit low for a dummy load resistor (Lab supplies have them to make them more stable with no load)

If you do a lot of battery stuff you can modify your supply to add an output diode and fuse. This makes it safer to use around batteries and will prevent the supply from draining the battery when off. I used that setup when i last needed a "programmable charger" as part of a automated system.