Is a Meanwell power supply any good? What actually fails, and how to check yours is real

Ask around and you get two answers, both delivered with total confidence. One shop has run the same units on a packaging line since 2014. The next one wrote a blanket ban into their spec after a single bad batch. Neither is lying. The split comes down to heat, sizing, and whether the unit was genuine to begin with. So: what actually fails in a Meanwell power supply, what the published reliability numbers mean, and how to work out in two minutes whether the one on your bench came from the factory whose name is on it.

What actually kills a Meanwell power supply

Electrolytic capacitors and the fan. For anything that dies of old age rather than an install mistake, that is the whole list.

Capacitor life runs on chemistry, and the chemistry only moves one way. Every 10 degrees C of extra ambient roughly halves it, every 10 you take away roughly doubles it. Meanwell rates its own product against that curve. The numbers get uncomfortable fast. Full load in a 55 degrees C cabinet estimates out around 35,000 hours. Back off to 75 percent at the same ambient and it stretches by about 1.6 times. Hold that load and cool the cabinet to 45 degrees C and you are near 114,000 hours. Same part number, same hardware. The only things that changed were heat and how hard you leaned on it.

There is long-term evidence behind that curve. Meanwell pulled back a batch of its own 150 W LED drivers built in 2008, units that had spent ten years in street lights, and measured them against the original spec. Efficiency came back at 90.1 percent against an 89 percent rating, ripple at 109 mV against a 150 mV limit, hold-up at 56 ms where the spec asked for 50. The electrolytics inside were Nippon Chemi-Con and Rubycon parts, and capacitance, ESR and dissipation factor were all still in tolerance after a decade outdoors.

Those were potted LED drivers, not an enclosed unit in a hot cabinet, so do not read it as a promise about your RSP. It does show the parts selection is good for ten years when the thermal conditions are sane. Which puts the weight back on the install.

Worth knowing for diagnosis: every unit gets a full burn-in before it ships, which screens out the infant mortality end of the curve. A supply that dies in week three is not wearing out. Look at the install, the load, or where it came from.

Estimated service life against ambient temperature at three load levels. The gap between the marked points is the whole argument about heat.

 

The published numbers, and what they do not tell you

MTBF is the figure quoted in every Meanwell power supply argument I have been dragged into, and almost nobody reads it right. Meanwell calculates it by stress analysis to MIL-HDBK-217F, and what it means is that at the MTBF hours, the probability a given unit is still working is 36.8 percent. At twice that figure, 13.5 percent. It describes a population, not your cabinet.

Service life is a separate calculation, worked back from capacitor temperature rise. On an RSP-750-12 the MTBF is 109,100 hours, while the life cycle figure for its main capacitor is 213,000 hours at 50 degrees C. Same unit, two numbers, nearly a factor of two apart.

Published Meanwell power supply figures across the RSP enclosed family. Same series, same warranty, and a wider MTBF spread than most people expect.

The RSP-2000 row is the one to sit with. Same family, same five year warranty, under half the MTBF of the 750 W unit. Push 2000 W through the same 1U footprint and more parts work harder in the same volume of air. If you were about to buy one 2000 W supply where two 1000 W units in current sharing would do, that is the trade, and nothing in the headline numbers tells you about it.

Do not use MTBF across brands either. Meanwell says so itself, which I respect. The figure moves with the standard chosen, the stress assumptions and the component count, so a bigger number can just mean a friendlier spreadsheet.

The install mistakes that make a good Meanwell power supply look bad

Sizing causes the most grief. Meanwell's own recommendation is a unit rated 30 percent above what the load actually draws, so a 100 W load wants a 130 W supply. Most of the Meanwell power supply units I have seen cook themselves were sized to the load and nothing more, then bolted into a sealed cabinet next to a drive.

Orientation costs you too. A fanless unit mounted vertically instead of horizontally gives up around 5 degrees C of headroom by Meanwell's own example, and by the rule above that is service life thrown away for free.

Multi-output units have a minimum load on each channel. Leave a rail with nothing on it and its voltage drifts out of tolerance, then somebody spends an afternoon chasing a fault that was never there. Put a bleed load on the unused rail or stop trusting the reading.

Overload behavior is not consistent across the family. An RSP-750 goes into constant current limiting and comes back by itself once the fault clears. Do the same thing to an RSP-1500 and the output drops after 5 seconds, then stays down until you cycle the input. Assume the first on a unit that does the second and you will write it off as dead.

Then inrush. An RSP-1500 pulls 60 A on a 230 V start, an RSP-2000 draws 50 A cold. That is why the recommendation is a type C or type D breaker, and why several supplies on one feed need their startup staggered or the breaker trips on a perfectly good day. Paralleling has its own rules: four to six units maximum, output voltages matched inside 0.2 V, and about 90 percent of the combined rating rather than all of it.

How to tell a genuine Meanwell power supply from a fake

This is the half of the reputation problem nobody brings up. Meanwell states outright that counterfeit product is in the market, and the copies stopped being obvious a while back. A counterfeit Meanwell power supply used to announce itself with crude assembly and a smeared logo. Plenty of them photograph fine now, and the tells are consistency tells.

The most useful check costs nothing: does the model make sense inside its own series? One fake going around was an LPV-200-12 in a metal case. The LPV series is plastic bodied and has no 200 W model in it, so the part number was fiction before anybody opened the box.

Then the wordmark. Genuine product carries MEAN WELL as two words, worth knowing precisely because everybody writes it as one. Copies get the letterforms or the spacing subtly wrong, and once you have seen a real label the fake stops looking right.

After that, the serial number. Meanwell runs a warranty period check on its own site, and the logic here is blunt: a serial that does not match their record is not their unit. When a safety case rides on it, ask the local Meanwell office directly.

Run them in this order. Any single check failing is enough to stop and ask questions.

 

Five checks, about two minutes with the Meanwell power supply in your hand.

What a fake costs you is not just the failure rate. No real safety approvals behind it whatever the markings claim, no warranty path, nobody to call when it takes a rack with it. That is a different category of risk from sizing a genuine unit badly.

Things people ask me about this

Is a Meanwell power supply good enough for industrial use?

For most panel work, yes, provided you spec it like a thermal component and not a commodity brick. Size it 30 percent over the load, keep the ambient down, respect the derating curve, and buy through a channel you can trace to the factory. The ones that earn the brand a bad name run at 100 percent in a hot enclosure.

How long does a Meanwell power supply last?

It depends almost entirely on how hot the capacitors run. At 75 percent load the published estimates land near 57,000 hours at 55 degrees C ambient and 114,000 hours at 45 degrees C. Six and a half years against thirteen, same supply.

Why does my Meanwell power supply keep shutting down and then coming back?

That is protection working, usually overload or over temperature. Both drop the output and both clear when the condition goes, except some models restart on their own and others need the input cycled. Measure the real load current and the cabinet temperature before blaming the supply.

How do I check that a Meanwell power supply is genuine?

Model number against the series, then the wordmark and the printed markings, then the serial through the manufacturer's own check. If one of the three does not line up, treat it as suspect however respectable the invoice looks.

If you are speccing the supply rather than diagnosing one

If you're speccing a supply rather than diagnosing one already in the panel, the shelf here runs Allen-Bradley, Siemens, Phoenix Contact, Omron and Sola, DIN rail and enclosed, tested, with a two year IQElectro warranty. An Allen-Bradley 1606-XL that finally gave up is the one I get asked about most: Rockwell discontinued that line and points you at the XLS, and we carry the 1606-XLS120E in new open box condition. Send the part number off the dead unit and we'll tell you what actually crosses.