Repairing Meanwell Power Supplies

heathyoung

100 kW
Joined
May 27, 2009
Messages
1,545
Location
Newcastle, Australia
OK - I'm sure there are a few of you who have meanwell 350W power supplies that have died (or you have killed) :roll:

I've just repaired 3 - all of them popped fuses violently on power up. Thanks to Hyena I had three dead ducks to fix.

Fortunatly, as far as SMPS's go, they are pretty simple beasts, with off-the-shelf components.

The first two units: Blew fuses, no crispy bits, the last one - ouch! But we will get to that later...

The most common fault is that they kill their high-side push-pull transistors (the big pair bolted to the end of the case) - they usually fail open or shorted. When they fail shorted, they also take out the bridge rectifier as well. Trust me on this one - it tests OK, but at high voltages they short out (they start to behave like Zener diodes - had one that was fine on the diode tester, but suddenly shorted at about 15V, and the other at 40V).

If the push-pull transistors are shorted, replace them - they are cheap and readily available from old PC power supplies if you want a $0 fix like I usually do. So are the bridge rectifiers. You also need to check R13 and R14 (0.47 Ohm resistors) as they become fuses as well :)

Now - the moment of truth. RULE number one in SMPS repair - don't just chuck a fuse in it and hope for the best! Put a 60W light bulb IN SERIES with the active (or HOT for those in the US of A) - If the light bulb flashes brightly for a second and decays (caps charging up) and the power LED lights up - you are in business. IF the light stays at full brightness, keep troubleshooting sunshine :)

Funny story the last power supply: It had burnt tracks around the common-mode filter choke, and the neutral line had no continuity - it appears that there was some bad soldering from the factory, but the lead let go and arc-ed badly - the resulting spikes killed the bridge rectifier, R13, R14 and high-side transistors - Replacing the parts (along with a salvaged common-mode choke from a computer PSU) and the supply was producing a voltage, but sounded VERY unhappy. Drat (or words to that effect). 25V was getting produced, albeit very noisily. Hmmm... This means there is a problem in the regulation section OR the push-pull driver section was only pushing (or pulling)

Some more poking and prodding rules out the transistors driving the pulse transformer - and revealed a dead 33V zener diode - ZD3. I didn't have a 33V one, but a 20V one did in a pinch (they are only to limit the flyback spikes to the transistors) and restored it to its full 48V output.
 
Interesting post, i just bought two 350w meanwell power supplies. Both are said to be working.

You've got 3 dead ones.... now i wonder about these :(
 
Great post - I've repaired a few where the blocking diode (the bidirectional one) was fried in both directions and replacing was all it took to fix them... never went beyond that, they are so low cost they are nearly disposable :)
-Mike
 
i guess i did not follow this. can you post up a picture of the push pull transistors and where they are in the circuit? is the 'blocking' diode the schottky diode that keeps the current in the back end?

pictures of all these parts would help and where they fit in the circuit would be valuable to learn from.
 
I received a 12V DOA off Ebay. Powering it up it just buzzes faintly. Power light doesn't come on.

The seller sent me a new one no questions asked which was great.

Obviously I checked the fuse, but that's the extent of my know-how.
I always wondered if I'd be able to fix the old one.

Edit: it was a cold solder joint on the main transformer, re flowed the solder and the thing has run fine ever since.
 
Yep - 3 leg bidirectional diode = Blocking diode :)
-Mike
 
You mean the shottky diode on the output?
 
Yes :) I had hooked the power backwards when I pulled that component to test it (since it was first in line) I realized that it was shorted in both directions - so I replaced with somthing heftier a bit (all I could get) and whalla - instant Power Supply :)

-Mike
 
I like your write-up of the steps; they're easier than the ways I've been doing it (on any SMPS). :) Especially the lightbulb in series with hot/active. :idea:
 
heathyoung said:
RULE number one in SMPS repair - don't just chuck a fuse in it and hope for the best!

How about soldering a few strand of fine wire across the fuse after you've blown all your replacements ? :oops:

Good work on getting them going again!
I just threw out 4 working computer PSUs on the weekend, I would have given you those too if I knew you could use them.

I'll post a link to this thread over in the main meanwell charger / mod thread
 
i just blew up my meanwell clone....

stupid buzzing noise i tried to raise the current limit a bit to get rid of it, 20 seconds of screwing around with the POT and all of a sudden BZZZZ!!! both of those "push pull" thingy shorted...and blew the fuse.

so what else did u say blows with this? picture of the other damaged part?
 
Hyena said:
heathyoung said:
RULE number one in SMPS repair - don't just chuck a fuse in it and hope for the best!

How about soldering a few strand of fine wire across the fuse after you've blown all your replacements ? :oops:

Good work on getting them going again!
I just threw out 4 working computer PSUs on the weekend, I would have given you those too if I knew you could use them.

I'll post a link to this thread over in the main meanwell charger / mod thread

Man I used a piece of solder to bridge the fuse holder. I think thats the only way Im running the unit well past its rated output with success.
 
Red_Liner740 said:
i just blew up my meanwell clone....

stupid buzzing noise i tried to raise the current limit a bit to get rid of it, 20 seconds of screwing around with the POT and all of a sudden BZZZZ!!! both of those "push pull" thingy shorted...and blew the fuse.

so what else did u say blows with this? picture of the other damaged part?

Bridge rectifier goes sometimes as well. Its the other part on its own heatsink, 4 legs, near the mains caps. The buzzing sound is the PSU getting unstable, increasing the current would only make it worse.

Personally, I wouldn't be running these things much past their factory ratings, they are a little fragile. If you were to up spec the power transistors, you could probably pull 500W, thats the weak link in the design. You will come to a point where you will saturate the inductors though.
 
I've been running mine at 420 watts (17 amps @ 6S charging voltage) for over a year, often charging twice daily and frequently forgetting about it and leaving it on for 12hrs or more and it's still working perfectly. I've also blown a number out of the box too due to over current issues (pulling 500w+) so I would recommend keeping them to 400w for longevity.
 
teslanv said:
have to see if I can resurrect it :)
Well you managed to successfully resurect this 3 year old thread so I have faith in you :p
 
Pics and mods MeanWell S-350-48

S-350-48_1.JPG
 
teslanv said:
I've got a SE-600-48 that I managed to blow. have to see if I can resurrect it :)

post up some pictures inside and we can walk you through the components. they have axial fuses on the board that can blow also and you can check continuity on them easily with the DVM.
 
.
 
i usually look for output on the schottky diodes first. if there is voltage there then the problem is in the back end and if no voltage present then the problem is in the front end.

check the voltage across the push/pull switching transistors on the front end first. they are the two on the left below the two big input caps at the lower right. the other is a 5V regulator i think.

the voltage from the collector of the high side npn transistor to the emitter of the low side npn transistor is gonna be about 340V DC. if no voltage then the rectifier bridge may be the problem.

the pulse generator heath mentioned is that small transformer in the back on right and it takes the current from the two little npn transistors next to it that are controlled by pins 6 and 8 i think on the TL494. the current from the secondary of that little transformer drives the bases of the two switching transistors in the front end and on some of these types of power supplies there is a thermal fuse in the wire that connects the output of the pulse generator to the big switchers in the front end. the thermal fuse is usually physically attached to the large transformer because that is where the heat load starts.
 
Ran into a problem with my pair of USP-500s (http://endless-sphere.com/forums/viewtopic.php?f=14&t=54854)

The SPW20N60C3 blew for sure, taking the fuse an ICR with them.
So Mouser shipped, I replaced the 3 above - thanks to the lightbulb hack above I didn't blow anything else but now know something else is still amiss...
I think I heard a "click" the first time I powered it on, and the bulb dimmed for a brief moment beforehand.
The other, "working" one flashes, dims and then goes on and off like an indicator (probably PFC at work?).

So what's the next thing to check, the relay (because of the "click")?
 
Just managed to get my S-400-48 MW clone unit working again after stupidly connecting to a battery with the current limiter set to max. Saw the current on the wattmeter rocket to well over 20 amps and rising, made a grab for the mains switch but the pyrotechnic display quickly told me I was too late.

Anyway, just for info, the following had all died following this little adventure:- Q1 ,Q2, Q3, Q4, R10, R13, R14, D1, D2, replacing these seems to have restored normal operation. Thought maybe this info might help someone else who's as daft as me, though to be honest the cost of the various components eventually ended up being not far short of a replacement clone unit. Considering my limited electronics abilities however, it was quite satisfying to rescue it from the 'parts only' bin.... :)
 
Postby marcexec » Fri Jan 09, 2015 8:02 am

Well done!
How did you figure out which to replace? Just visually?

....I had the lid off when the thing went pop, it was fairly obvious that the fireworks had come from the Q1/Q2 area and quick visuals around there immediately condemned these, plus D1 and D2 with various splits/charring/lumps missing.
Further meter checks in this area of the circuit also showed R13/R14/R10 to be open cct.

Q3 and Q4 were a bit more challenging for me - replacing the above mentioned items showed that I had a decent rectified mains supply at Q1/Q2 but no signs of any switching going on. At this point I was reduced to random prodding around doing DC checks with a multimeter, fortunately I had another supply to use as a comparison. Eventually found discrepancies on pins 11 and 12 of U1 (resistive check to ground), this led me back to Q3/Q4, measuring these out of circuit showed up partial shorts in all directions, replacements doing the trick.

I got lucky really, most of these were fairly easy to find with a bit of basic knowledge, a bit of persistence always helps too I suppose. Whether the time spent is justified when these things can be had on the bay for £23 a pop is less certain though... :D
 
Back
Top