I've been looking through the forums for a while lately and I like what I see. Anyway, I am a EE student at UT Arlington and I like to take stuff apart to figure out how it works or in this case, why it went down in flames!

There have been a LOT of PC switching PSUs that have been just failing to hold rated load, overheating or even shooting out little bits of flaming shrapnel (bits of rectifier) out the back.

Too bad I don't have pics of the unit but here's what I found in a customer PC:

The PSU was labeled as a 450W PSU but only weighed about 2.5 pounds, I've seen 200W units with more substantial construction. The 12V rail said 30A continuous on the label.... the 12V rectifier had an absolute max rating of 15A! Needless to say, it was toasty, well very toasty, it had a nice chunk missing!

The 3V and 3.3V rectifiers were also similarly undersized but not toasty. The installed rectifiers were in TO220 packages... the correct one would have been in a much larger TO247 package.

No EMI/RFI filter was present at all, not even a cap. This explains why my radio reception went out the window when I turned the ailing PC on.

To add insult to injury, the line fuse was missing!!! On the PCB was a notice: RISK OF FIRE: Replace with same type and rating of fuse. The fuse holder was not present, what was present was jumper wire. This was not a DIY thing, all the soldering was factory! Yikes....

Other sins: 20AWG wire where 18 or even 16 would be called for, counterfeit capacitors, and a sleeve bearing fan.

So how did I find out all of this? Well, this PC came into the shop for a spontaneous reboot complaint (this is 90% PSU or capacitor plague related). I opened the case and saw the Powmax PSU and knew what was wrong but just to make sure, I decided to do a test with a multimeter. Well, when I fired it up, it decided to die right there. The rectifier shorted, the PSU overload protection circuit (switching transistor drive kill) was out to lunch and the smoke poured out. A trace on the PCB burned and cut the power.

The customer is extremely lucky the entire PC was not fried, the thing had a crowbar circuit that worked! The computer itself was put together by a now defunct PC shop in town that was infamous for using shoddy goods. I have fixed many of their POS rigs, I had been stealing their customers for a few years and so did every other honest shop until they folded.

I'll see if I can find the pics I took of a similar, but not quite as bad unit that failed in a friends computer. Somewhere I have pics of a Raidmax "420W" in contrast to the 400W Fortron unit in my PC. The difference is astounding.

Here is a PARTIAL list of known trash PSU makers:
Powmax/Leadman
Deer Computer (A.K.A Allied, Austin, Premier, L&C, etc)
Raidmax
Codegen
Young Year (Aspire/Apevia, older Ultra units)
Logisys
Coolmax
Sunbeam

Most trashy PSUs have at least a fuse, but many leave off the EMI/RFI filter and all of the ones on the list are grossly over rated pieces of Chinese junk. They are the Zinsco of the computer world.

What is sad is I know some class mates that STILL buy them for their own PCs. Oh well, school can't teach you everything. An ounce of experience and common sense is worth quite a bit more than a pound of theory. The EE professor almost had a stroke when he saw a similar board, the Physics prof had no clue why it was all wrong except for the missing fuse! :P

The FTC and CPSC need to step in on this, the BS has gone on too long. These units could cause a fire, at the very least they could (and have!) destroy a computer and the data within.

This is exactly the reason UL exists. The specific purpose of their standards -testing for public safety- is served by their testing products for these hazards.

If the products you describe are UL listed, the folks at UL would be very interested in any documentation / examples you might have.

If the products are not UL listed.. well, you save a buck or two. Your choice, your risk.

If the UL label is a fraud, again, UL would love to see it.

Finally, never underestimate the power of the marketplace. I can think of a few automobiles that met every standard, yet are history - because they were flawed. Some were even made by huge firms that "owned" the market. Sure they did.

Finally, have you written the companies involved? Some of them actually do seem to be in the business of selling quality products.

A call for governmental involvement should be the last recourse ... not the first. It has been said that, ultimately, government is expected to do three things: deliver the mail, pave the roads, and secure the borders .... once they do those jobs, THEN maybe we can trust them with another.

You will never find one of these units in a name brand PC, ever. This is a PC shop and DIY thing. Most PC shops have become aware of the problem (the one I got for FREE with a case never left my house, I take some pride in what I sell!) I still see them in PCs made by a shop in town that went under a few years ago, and almost always they come back to the customer with a new UL listed, heavy as heck PSU.

I don't think the sample I had from my friend's PC was UL listed I tossed the unit a while ago so I can't check. Of course, that's the first clue that they are junk. The second clue, and one that should give it away immediately is how flimsy the product feels. They weigh about half what a good one would of the same rated output.

As for UL or CSA listings, the product description on newegg.com says they are listed, however on a few units when you look at the label on the unit itself, no such mark can be found. Rumor has it that the importers of these POS units will send one unit to test and import another.

I have complained to newegg.com, but they don't care, the Powmax (the worst of the brands by far) units sell like hotcakes to idiots that measure quality based on how flashy something looks or how good of a "bargain" it is. Newegg is owned by Amazon.com and something tells me they aren't going to consider an email from one person in deciding to suddenly stop carrying a product. Maybe they should look at the reviews, others are not pleased. The ones that are pleased have systems that don't even put a 50% load on these. A PC with one HDD, a DVD drive and a basic Celeron CPU doesn't draw even 200W.

The new ones that are UL listed and DO have a fuse and most of the time, a basic EMI filter (one cap, one choke in hot and neutral), however, they still cannot hold rated load or even close. So they may not burn the house but they are still perfectly capable of taking out a motherboard or CPU. Worst of all, the hard disk usually gets damaged when the things go boom and customers don't usually back up regularly.

As for market forces, they exist on price alone. One of these POS special "550W" units costs about $25, a good 550W unit is around $85-$100. If it's too good to be true, it IS!

Some of the PC shops that used these have gone under due to huge warranty costs and a bad reputation. Now that DIY PC assembly has taken off, these crap PSUs have an even larger market. The average PC DIY with a how to book doesn't know jack about electricity. They don't understand how a PSU with an input rating of 115V @ 2.5A couldn't possibly even come close to putting out 500W. The DIY or uneducated customer is thinking this when I try to explain: conservation of energy? huh? what? Just because they can put a card in a slot doesn't mean they know what they are doing.... kinda like a DIY with a wiring book from Lowes doesn't know how to safely wire his house!

This is all very interesting HWG. Perhaps you should post a reference to these findings, and introduce yourself, on the "Computers and the Internet" forum. By the way... WELCOME TO ECN!!!
Joe

Thanks for the welcome! I will post in the computer section shortly.

Well, I have dug up some more info on the the latest Powmax/Leadman Technologies PSUs.

They claim UL and TUV certification but on the unit I cannot find the UL Registered Component mark (heck I cant even find the normal mark!) nor can I find the TUV mark. I did find the D Mark which is provided by UL-Demko but I have reason to believe that this is bogus.... Europe is even more anal about stuff than the US.

It also claims CE but the the logo is WRONG.

This thing is a fraud, pure and simple and every PSU Powmax has is BS.

With the evidence I have and the customer reviews from Newegg.com, I may fire off another eMail to them. This PSU needs to disappear.

I have a friend that works as a tech at Frys Electronics in Dallas, so maybe he can get me some more data as they are sold there as well. (they NEVER use them in custom builds despite the cost difference...that should tell volumes!)

I found something interesting in a friend's junk pile tonight: a CompUSA branded Powmax LP6100D 400W PSU! What is radically different about this older model is the fact that the internals are actually OK! There is a fuse, a good line filter and heavier components throughout. The heat sinks are MUCH larger as is the transformer. It is based on a board that is 99% identical to the newer POS PSUs.

It appears that they got the certification for the old unit, then pulled a fast one and put the POS guts in sometime later.

There is another virtually identical unit with a different model that comes in cases and it has no UL mark. The only difference between the LP6100D and the one that comes with the case is the case version has a single 80mm fan instead of a dual fan and the wattage rating is higher.... 450W supposedly! Yeah right!

Oh I found the pics that tell the difference between a quality PSU (Fortron Source 400W) and a shoddy unit, a Raidmax 420W. The difference is STUNNING. I understand the mod wants the images on a server, so just tell me where to eMail them.

I will take pics of the old version of the Leadman/Powmax LP6100D and the new one that came with a friend's case. That friend was educated about the PSU before it saw the wall socket...I probably saved the computer!

See the POS for sale: http://www.newegg.com/Product/Product.aspx?Item=N82E16817163013
Note the way the 4 pin 12V CPU connector is wired.... not cool!

http://www.powmax.com/SUPPORT_PAGE.htm
This is their support page and they supposedly have testing certificates here. I'm not too familiar with the appearance of a legit certificate, so if anyone knows what a legit one looks like, It would be great. I suspect the UL is bogus, a search for the company yields no results on UL's site.

One or two little details about UL labels....

First, there is no requirement that a manufacturer have all of his product carry the UL label. Since UL charges 'per label,' it is not unknown for a manufacturer to offer the same product, either with or without the label. Still, if there is no UL tag ON the product, there is no reason to infer that it's the same as the listed version.

Another fact of commerce is that the actual manufacturer may not want his identity known. For example, it may be preferred that the product carry the name of the vendor, or the distributor. For example, Sears make no products - but try to find the manufacturer's name on any.
In those circumstances, the UL label will have a "File Number" on it, that will allow UL to identify the maker.

hmmm... just a thought.... would it be possible, that some PSU's are plagiated by chinese facturies?
Here in germany, there are a lot of electronical Products with counterfeited Security signs like CEE etc. and trademarks issued.

I can imagine, that the use of less materials in transformers and heatsinks and leaving out fuses and protection circuits would boost the profit margin decently high... at the cost of the end user for sure

Regards

Herrmann

Also beware of the "Name Brand" Power Supplies especially Antec (Except the Seasonic built units)as they are made to be quiet and therefore have inadequate cooling for the caps used and can fail early I have had one fail early and take out the Motherboard and Video card and Ram Luckily my Hard drive survived. Plus Antec was nice enough to replace those components with new ones(I build my own custom computers)

I've had some power supplies with no NRTL marks. Definitely lower quality. No solder mask on the PCB and no strain relief for the wiring (tugging on the wires pulls directly on the solder joint). Parts teepeed because of change to design without redesign of the PCB. But this isn't too suprising since any company too cheap to get their products listed or recognized probably cut other corners as well.

Hardwareguy, you made one statement that I find to be backwards. European product approval is MUCH easier than the United States and Canada. In fact, its a joke in comparison. Even Mexico is tougher.

To apply an NRTL mark in the US, you must submit samples for testing to a Nationaly Recognized Test Lab such as UL, Intertek/ETL, CSA, Curtis Straus, MET labs, etc. The NRTL must be authorized as an NRTL by OSHA and operated and audited under ISO 17025 (defining appropriate laboratory procedures). After product approval, the NRTL will make regular visits to your factory to make sure you are still building your product with the appropriate components - the ones that were identified as critical during the initial product evaluation. The safety standard in the US and Canada was adopted from the international IEC version, but adds requirements, sometimes rather significant. For example, Europe does not have a "power cross" test for telecom circuits to ensure safety from fire in the event a tip or ring pair comes into contact with outdoor power lines. Also, in the US a user-accessible exposed SELV (low voltage <60vdc) circuit must remain within SELV limits. In Europe, under fault conditions, it is allowed to jump to the maximum voltage allowed for a telecom circuit.

In Europe you only need to apply the CE mark to get your product to the store shelves. The CE mark means "this product complies with all applicable compliance criteria, and I can supply data to back up this claim". The CE mark is a "self declaration". There is no requirement for third party verification, and no requirement to audit manufacturing facilities. You can apply the CE mark based on design data if that suits you (good luck on that lawsuit).

To me, the no fuse issue would raise an eyebrow, but it does not mean the product "fails UL". Computer power supplies are currently Listed to ANSI/UL 60950-1 and/or CAN/CSA C22.2 No. 60950-1 (themselves a variation of IEC 60950-1). This harmonized safety standard specifies safety under normal operation and under single-fault conditions. As with many standards, there is no specific requirement for a fuse. If failure analysis indicates no fuse is needed, then you don't need one. That said, under single fault conditions, primary to secondary insulation damage is likely to occur if there is no electrical fuse and there is no thermal fuse. But, protection may be provided by other active circuits, or by Listed fuseable wire in the primary circuit...or something else.

OSHA requires use of Listed equipment in the workplace, and your insurance company may fight you on a claim if you have a fire as a result of using non-Listed equipment (that being the whole purpose for creating Underwriters Laboratories). A reputable company that only sells to US/Canadian customers should not be selling non-Listed or CE-only electrical equipment to the general public. You probably won't find this stuff at a North American Walmart or Sears.

Do the sayings:

"You get what you pay for;" and,

"Let the customer beware" apply here?

Reno: those sayings should apply. I am concerned that we, as Americans, have fallen into a mindset that imported items should be low cost and high quality (because many have been). Retail now tends to be all about price competition. I guess we all know that one major electronic retailer fired all its knowledgable salespeople to cut costs. Even if you know enough to question products, there may not be anyone who can advise you. I feel privileged to have access to this forum, where such things are up for discussion.

HWG,
I had a 400W Raidmax PSU nearly burn my house down last year, I was at work when the fire started and the quick actions of a fellow FF contained the fire to one room, plus smoke damage.
What annoyed me the most was the fact that the PSU was being used nowhere near it's output capacity, in-fact it had an old 386 computer running Mandrake Linux on it for my weather station.
I had the idea that bigger was better, when 200W probably would have done the job.
I lost a lot of radio gear in that fire and it turned out upon investigation that the fuse in the PSU was rated at 3 times what it should have been, this was out of the factory.
Also what I found, was that the MCB at the panel had blown, but the PSU (According to my mate) continued to burn "with bright yellow flames and thick acrid smoke exiting the rear fan aperture".
Now, one thing that you guys have to remember, is that brands mean nothing anymore.
We've had parallel importing here for close on 10 years now, this is where cheaply made brand name goods are imported without the consumers awareness, but are sold as the real thing.
These goods are made as cheaply as possible and have huge mark-ups once they make it to the retail stores.
All that aside, I really don't see how you can tell the quality of a SMPS based on wieght, maybe I'm missing something here?

hardwareguy,
You can do some checking of the UL information here.
You type in the UL category code (everything before the "." on the powermax page that you linked to) This will give you a list of every manufacturer that has a UL listing under that product code. Note that "QQFU2" is the only valid category code listed on that page and Leadman Electronic Company does not have a valid listing in that category. You could send a note to UL with that page link and ask them to check on it.
Don

It isn't a foolproof measure by any means, but more weight generally means larger heatsinks on internal components, larger transformers/inductors, and bigger capacitors, all of which can help improve longevity.

Mike, Thanks for that story. Many of my friends and family think I'm daft switching off the power to my PC when it's not in use. After hearing of your experience, I intend to carry on doing just that!

A point I often find myself making when people ask about the "best" brand of domestic TV/VCR/DVD etc. Frankly, at the typical consumer-product end of the market, I don't think there's anything to choose between them these days.

Thanks for the PSU blacklist. I checked Microcenter, none of them except for the Coolmax were sold there.

It used to be all but fool proof, but the newer high efficiency designs are proving difficult to measure by mass alone. Sometimes it is necessary to peer through the vents. I now looks for some distinguishing features of a good PSU:

Decent finish and thickness of the enclosure but overly flashy may be a sign of trouble.
Fan type (ball bearing, magnetic or hydrodynamic are best)
Wire size (no 20AWG wires except for sensor or control leads)
Board material, better PSUs use fiberglass PCBs, cheap ones use phenolic.
Component brand choices, especially on capacitors!
Heatsink size is not a huge factor with some modern designs with efficiency ratings over 80%, small can sometimes be OK.
The last tell tale sign: warranty

Wit that said, even a quality modern 80+ efficiency model will still have some beef to compared to the cheap crap.

http://www.jonnyguru.com/modules.php?name=NDReviews&op=Story&reid=71

Here's a review of some bottom of the barrel power supplies that promise the moon for little or no money. The results are as entertaining as the violation photos section!

I say no money because these pieces of junk are often supplied with a PC case, and said case is $20 while the minimum market price for a case happens to be about $20, with or without the crappy PSU. Oh yeah, 20 dollars does not get you a decent case either! Good cases start around $40.... without a PSU. Bad cases tend to be so flimsy that the motherboard flexes when cards, memory modules or in some cases, cables are attached. They are also too flimsy to keep vibrations damped, they tend to be quite noisy.

Don't know how I missed this thread, but don't get me started on consumer electronics! It's all the same cheap crap anymore; in a world where marketing trumps all and salesmen know they can take advantage of people who think more expensive = better, price doesn't seem to matter squat. How many times can you open up a $100 box and a $500 box and find the exact same stuff underneath? Oh, for an economy ruled by the engineer, when you can actually make an informed decision base on price, features and quality, and you actually get what you paid for!

I've had so many hard drives and PSUs die on me, I've become quite jaded. After having a 300W Deere die on me, I was determined not to make that mistake again- but the 3 400W+ Antecs that followed it didn't last any longer. I have another expensive 400W+ PSU in my present PC, but cost really doesn't mean anything anymore. It was highly rated on Newegg and is about 3 years old right now, so that's good, I suppose...

You all not only regularly back-up your critical files, but keep an extra copy offsite somewhere else so that you don't lose all your digital photos, tax records, etc, in a fire right? My parents get a freshly updated DVD-R with all our photos every time I see them, and I've got another DVD at work with everything else. People think I'm paranoid when I meantion off-site backup for my personal files, but it's not like it's all that difficult, and I know I'm covered in case of catastrophe.