MeanWell power supply quality control issues!

[AlanR]

I just wanted to give a heads up to anyone that purchases power supplies from MeanWell should always look inside them for issues. I recently purchased a 24 volt DC at 63 amps, Model RSP-1500-24 that I paid about $379 for. I wanted a high power, power supply to characterize 1200 watt or lower DC motors that are 24 volts.


Well, I always look inside things I buy to see if I am getting quality. So, inside the unit I found an 8-pin DIP style chip labeled as TOP244P switcher that was barely soldered into the board!!! I mean a damn snow sled could slide down the thing, it was sticking up so high on one side. On top of that, the solder joints looked like they did not reach proper temperature because they looked like dull colored boulders instead of the shiny mountain peak look of a good solder joint and I could see lots of surface mount devices having almost completed solder bridges.  Another issue I have found that I will need to correct is that for some reason they figured that they could just run the 24 volt plus and minus copper buss bars very close to the aluminum frame inside the power supply enclosure with no insulation shielding. Short circuits anyone? It looks like one simply has to press down on top of the enclosure to make the buss bars short out against the aluminum frame! So it looks like I will be hand soldering all the cold joints I see, seating that 8 pin DIP chip, and buying some Kapton tape to put over the buss bars. Geez man. I did not even power this sucker up yet, and will not do so till I fix these issues. Hopefully the magic smoke will not be released once I do power it up. I'll let you guys know if it does!

[BravoV]

Are you sure its not a MeanWell "clone" ? This particular brand is widely copied among Chinese cheap knock-off manufacturers.

I've seen with my own eyes, they copied the label,box,sticker, in/out terminals, the adjustment pot and the aluminium frame so perfectly that you just couldn't tell which is which before opening it up.

[AlanR]

It is the real deal. Even has their logos silk screened on the boards and the transformers. All parts are also prime spec except for like three capacitors I found to be the dreaded CapXon. Rest of them are all Rubycons. It is a two board design that uses a data harness and power harness to provide data and power to the other board. On the filtering, line conditiong and power factor correcting board I see the following info: RSP-1500A-R6 and 2009/08/25 and of course the MeanWell logo. What I found peculiar though, is a I see name in the lower left corner of the board, "Jim". Don't know who this Jim fellow is. Maybe the designer? Lastly it shows made in Taiwan.

[bookaboo]

Meanwell has to be one of the most pirated brands out there silkscreens are easy to copy, it's reputation is pretty good. Where did you buy the unit from?
I remember being told by our PS supplier that we could check if a meanwell product was geniune on their website by inputting a S/N.
I never actually did it but I think this is it http://www.meanwell.com/webnet/service/service_px.aspx

[Mr Simpleton]

Not sure you are supposed to see shiny blobs of solder, with RoHS and lead free solder, they tend to look pretty dull and the ability to wet is not as good as with lead solder.
 
And if you rehead lead-free be sure to crank up the temperature too!

[David_AVD]

It would be interesting to see if this is a genuine unit and pictures of the offending areas.

[AlanR]

I got a file called Lot No. : W1204A212 PDF from that site when I type in my serial number and model number. Some lady named Michelle Yu gave the final okay. I guess they cannot see chips sticking up and incomplete soldering. Kind of disappointing. Seems like my power supply was  rush job. Maybe a Monday or Friday made.



Pictures of the issues:

 Cold solder joints of that 8 pin DIP chip.


http://imageshack.us/f/593/a06p.jpg/



More cold soldering joints.



http://imageshack.us/photo/my-images/819/i4e2.jpg/



The chip that sticks up. It is a lot worse than this picture portrays.


http://imageshack.us/photo/my-images/12/qegp.jpg/



CapXon capacitors!!!


http://imageshack.us/photo/my-images/94/qlbg.jpg/



Exposed DC buss bars.

http://imageshack.us/photo/my-images/266/tzkx.jpg/




This is what sits on top of that board above. There is less than 1/8" clearance of the buss bars and the heat sinks. No paper was put on this surface.


http://imageshack.us/photo/my-images/89/36bg.jpg/

[wraper]

There are no cold solder joints, that's just how lead free looks like. No problems with that dip, this is dual sided board, so solder flows inside holes.

[AlanR]

Except for the fact that a pin was almost completely free from the board on that chip. Also a good lead free solder joint should still not ball up like that. It seems they have temperature control problem with lead free solder. Makes since because of higher temps required. That still does not excuse them from not keeping those buss bars and that heatsink which is used to get power from the diodes (the positive buss bar attaches to the heatsink) from being so close to the metal enclosure without some form of insulation barrier. It would not take much to cause the top of the enclosure frame to bend down to the point of touching the buss bars when a force is applied to the enclosure. Finally, there should be no reason to use CapXon capacitors at all!! The rest of the capacitors are Rubycon, but they figured that it was okay to skimp on a few of them by using CapXon? That is not excusable either. Go all out with prime spec parts or throw it in the bin as I like to say, if you plan to use garbage parts with prime spec parts. I plan to replace all the CapXon capacitors also.

[David_AVD]

 

[amyk]

It seems they left out the plastic sheet that's usually put between live components and the casing.

Mistakes do happen. Order another one if you dare, and see if it's the same.

[Psi]

WTF, who calls a company "meanwell".

[bookaboo]

I don't see anything unusual about that soldering, not much you can do with lead free to make it look like the old days. Also Capxon caps are everywhere, Ive even seen them in samsung and siemens gear. Not the best but sounds like they have identified where they need the good quality caps and where cheapos will do.
Dip component sitting up a bit... don't like it but no way id get out the iron for any of that. You're more likely to cause trouble than prevent it according to murphys law.

The busbar thing would worry me though, how hard do you have to press to make it short?
Also, why not complain to the seller rather than try to do their job for them?

[Len]

Also Capxon caps are everywhere, Ive even seen them in samsung and siemens gear. Not the best but sounds like they have identified where they need the good quality caps and where cheapos will do

CapXon caps are everywhere, and they fail everywhere. It's one of those brands that show up a lot in reports of failed electrolytics. IMO CapXon caps are a good indicator of a product that was carefully engineered to last just past its warranty period. Here's my example.

[wraper]

http://imageshack.us/photo/my-images/593/a06p.jpg/
You say that pin is almost free of the board but I can clearly see terminals on the other side of the board. Solder blob is not pretty, but doesn't affect reliability at all. I doubt that they have any soldering temperature problems because this is wave soldering, not reflow. As for bus bars it would be better to take a photo from the side so we can see how tall and close they are to metal enclosure. Yes capxons are crap but they are everywhere nowdays. BTW meanwell uses only Japanese caps on medical grade and high reliability PSU's. Those capxons are not used on high power rail so should not be an issue.

[AlanR]

http://imageshack.us/photo/my-images/593/a06p.jpg/
You say that pin is almost free of the board but I can clearly see terminals on the other side of the board. Solder blob is not pretty, but doesn't affect reliability at all. I doubt that they have any soldering temperature problems because this is wave soldering, not reflow. As for bus bars it would be better to take a photo from the side so we can see how tall and close they are to metal enclosure. Yes capxons are crap but they are everywhere nowdays. BTW meanwell uses only Japanese caps on medical grade and high reliability PSU's. Those capxons are not used on high power rail so should not be an issue.

Here you go:

http://imageshack.us/photo/my-images/59/0c94.jpg/


I don't see any CapXons in my computer power supply that cost me $200 less than this turd. All of them are Chemi-con. Anyways, I plan to polish this MeanWell turd so that it can hopefully perform well. My hopes aren't so great from the poor design choices, some component selections, and no inspection after soldering. It is worth a shot though. Let this be a warning to anyone else that needs high power, pay to play or pay for it in a bad way and be dismayed...




Well everyone,  I should have bought a TDK-Lambda Alpha1500 series. Sure it is $600, but no turd polishing needed as Lambda is known as a premium quality power supply.  Unless of course, any of you seen quality issues from them??  I'll let you guys know if this turd can pass power on with no overshoot and can take load upon power up, and other tests after I do all the polishing work. Don't expect miracles with this piece, though.....

[wraper]

Forgot to say, cheaper low power models, like PS series are 100% crapxons ond only 1 year warranty, but more expensive models with japanese caps have 3 years warranty. Took a look into datasheet, they offer 5 years warranty for your PSU, strange that they put capxons there.

[Sar]

Throw names about if you will, but Meanwell has been a good power supply supplier for quite a long time. It's possible that they have reduced their quality control, but it's also possible that you just got a very very rare turd. I have hundreds of 600 Watt/24v Meanwells running in the field and probably another 100+ 350Watt/5v running in the field and they are the most reliable piece of equipment compared to everything else. I've -never- replaced one in over 11 years, but I have replaced a lot of cruddy circuit boards designed or assembled by the less technically inclined.

I -do- think it's good you brought this up because it could signify a shift in their production quality, it's just that I would be totally shocked that they dumped their quality production and reliability in exchange for cheap production so quickly.

[free_electron]

i don't see any problems. soldering looks fine. this is lead free wave soldering.

[poorchava]

I should have bought a TDK-Lambda Alpha1500 series.

In few days I will most likely have a one like that (24V, 70A) for a diagnosis and repair. It overheated. Customer said, he made a mistake and obstructed air flow by accident, but I'm wondering why it didn't trip OTP... Anyway, client sad that there was a magic smoke .

As for good caps and CapXons in one place, there is one good explanation. A phrase hated by engineers around the world: "cost reduction". They probably designed it with Rubycons all the way, and then some penny pincher came and said something like: "ok, this is a capacitor, that CapXon is also a capacitor. They even have similar capacitance value. And they cost 1/10th of Rubycon. How stupid of those engineers". I'm almost sure anyone working engineering field experienced that

[nukie]

They are not stupid. It's a clever time bomb that will last the warranty period. CapXon are not the crappiest there are worst out there. They will last pretty long if operating in cool environment, perfect for open frame power supplies.

[bookaboo]

Was just discussing this one with another engineer, we both work on LED signs (he has been at it 25 years). All LED signs have loads of PS, usually 5v at 30-60v.
Neither of us can recall a warranty meanwell unit (i.e. <3 years), I cant recall ever replacing a meanwell and my colleague says he's only ever done a few. Both of us have replaced dozens upon dozens of other brands, mostly genuine crap no-name ones.


Take that fwiw, companies change.

[metalphreak]

They also only qc a select few units from each batch. You can see in the reports it's like 15 out of 1000.

[ovnr]

Oh, this looks fine. Can't say much about the busbars; it's not clear from the picture.

And as for CapXon - it's horrid, but as long as they stay cool and don't see any real load it should be fine.

[amyk]

They are not stupid. It's a clever time bomb that will last the warranty period. CapXon are not the crappiest there are worst out there. They will last pretty long if operating in cool environment, perfect for open frame power supplies.

I have a bunch of LCD monitors (5) with CapXons in them still going after 10 years... and they're left on 24/7. They get inspected and cleaned yearly, no bulges so far. Many others have not been so lucky, but these ones are kept in a controlled temperature room at 20C.

[mos6502]

You can't say that the CapXons are bad in this PSU unless you have the schematic.

The life expectancy of an electrolytic cap is primarily a function of the ripple current that it has to take.

If the CapXons aren't subjected to heavy ripple currents, but the Nippon Chemi-Cons are, it's very well possible that the Nippon Chemi-Cons will die before the CapXons.

[free_electron]

You can't say that the CapXons are bad in this PSU unless you have the schematic.

The life expectancy of an electrolytic cap is primarily a function of the ripple current that it has to take.

If the CapXons aren't subjected to heavy ripple currents, but the Nippon Chemi-Cons are, it's very well possible that the Nippon Chemi-Cons will die before the CapXons.

Exactly.
People are too eager to shot capacitors in power supplies.

And the coment about ther eonly beeing so many space between bus bar and cover... don;t press on the damn cover. i'm sure the datasheet will tell you how much warp there is allowed to be.
For the same reasons that harddisk covers have big warning stickers saying : dont drop , don't press here.

I don't see a single thing wrong with that power supply.

[metalphreak]

CapXon got a really bad name in the industry for a relatively short period of time where they were just making utter garbage. Now they're probably just as good (as bad?) as all the other non-Tier-1 suppliers.

The fact they aren't using CapXon all round for everything kind of indicates they've only specced them for non-critical parts of the design.

The busbar clearance issue seems odd. Maybe they did somehow forget an insulating sheet on yours. Is the outer casing quite rigid/thick aluminium? If it's within the clearances required for that level of DC voltage, and it's a fairly solid casing, it shouldn't be an issue. A supply like that isn't meant to be used in an environment where it gets bashed around.

If you hadn't opened it, would you have ever known if it continued to work fine? Just because it looks poorly designed doesn't mean it is. Mean Well make good quality products, but they are never marketed as laboratory grade super premium top end supplies or anything.


If you want to see poor QA, check out my thread where I discovered why my chinese no-name supply started smoking. They put in a much smaller value resistor for the output load than what it should have been, putting the poor 5W resistor past it's rating.

[free_electron]

what people dont realise is that capxon is one of the largest manufacturers of electrode material....
Many a capacitor manufacturer doesnt make that themselves ! So your rubycons may very well have capxon electrode laminates in them ....

this whole bad capacitor things stems from 2 misconceptions

1) in the early 2000's there was a stolen chemical formula for a dielectric . Mitsubishi chemicals had  an ew electrolyte yielding a much higher capacitance density. this formula was stolen by some Chinese outfit. only they did not realize they had an incomplete formula. a key ingredient stabilizing the electrolyte long-term and over temperature was missing... the Chinese outfit sold the electrolyte to many capacitor manufacturers and they sold oodles because it was cheaper than MCM electrolyte. The electrolyte hit the fan when it turned out that after a few months to a year under heave load the electrolyte started disintegrating , giving off hydrogen gas giving burst capacitors...

2) many capacitors are wrongly applied. inexperienced designers and beancounter malpractices pick the cheapest possible part. and that may , form a technical perspective, be the wrong part. Pulse loading , esr , and other parameters are overlooked yielding prematurely dying capacitors.
especially switch mode power supply tank capacitors live a VERY hard life and are subject to pulse current with high crest factors. combine dielectric losses with esr losses and you are boiling off the capacitor over time.

1) has been solved a long time ago.

2) will only get worse. Consumer stuff has, profit wise, hit rock bottom so the beancounters are pulling harder and harder and the quality of the design 'engineers' that rolls out of the education institutes is getting weaker and weaker when it comes to understanding components. They may be very good at coding and theoretical stuff but have no understanding of part fundamentals. Having to listen to 10 lectures on capacitors and 10 on resistors is boring.... besides, the old farts that knew that stuff have long since retired and been replaced by young whippersnappers that want to teach the theoretical stuff..

[Kjelt]

1) this formula was stolen by some Chinese outfit. only they did not realize they had an incomplete formula. a key ingredient stabilizing the electrolyte long-term and over temperature was missing... the Chinese outfit sold the electrolyte to many capacitor manufacturers and they sold oodles because it was cheaper than MCM electrolyte. The electrolyte hit the fan when it turned out that after a few months to a year under heave load the electrolyte started disintegrating , giving off hydrogen gas giving burst capacitors...

If a company really has any form of QA in their ranks they should test a new (part of) product before putting it in their own product.
So this says more about the companies that did this namely that they can not guarantee ANY of their products and it could happen just as easily tomorrow due some other production failure or cheap/fake supplier. So the big question is if your company would trust such a company?
Then there is this shortage of electrolytic products due to the disaster in Japan a few years ago and products couldn't be made due to the fact that the parts were not in stock. A lot of companies then switched to other brands, and this was not always a succes (and then I put it very mildly).

[electronac]

Hello;
Seeing that you get several photos of the Power Supply RSP-1500-24, by chance you have pictures of the resistors R25 and R90?
by a short circuit has been burned and I can not identify their values.
Thank you.

[AmmoJammo]

Old thread (I didn't dig this one up!) but it looks like there's a plastic insulation sheet inside the lid anyway

[AlanR]

Hello;
Seeing that you get several photos of the Power Supply RSP-1500-24, by chance you have pictures of the resistors R25 and R90?
by a short circuit has been burned and I can not identify their values.
Thank you.

I am not going to help you. I am not trying to be mean, but the power supply is embedded into a test rack as a power source for 1/4 to 3/4 horsepower 24 volt DC motor characterization testing and motor speed controller design testing, and I am not going to through all the hassle of having to disconnect everything and then having to open the power supply up for two resistors you need to look at. Just contact Meanwell for the values or ask someone else that does not have their power supply inside a test rack setup.

[LabSpokane]

If the power supply is for play/fun - maybe it's worth your time fixing it.  If this power supply actually needs to be dependable, just scrap the Meanwell and get a decent power supply.  We banned Meanwell power supplies in a previous life after we had nearly a 100% failure rate with *dozens* of Meanwell power supplies.  These were all ordered through legit distribution channels, so I believe they were "genuine."

After the tens of thousands in warranty costs, Meanwells were not cheaper than a quality power supply. 

[electronac]

Alan R: well thanks for responding me. I agree with you, neither do i would disarm it for to take 2 photos.
The idea was to see if between the photos you already  took when you had disarm the power supply, it was  the resistance I need.
i already contacted with MeanWell , I was told that I contact the local representative.
It does not work, the place has no idea because they do not open the power supplies, they only buy and sell.
Anyway, thank you.
I hope you understand, because my English is not very good.