Slashdot Mirror
Illicit Leaky Capacitors Killing Motherboards
mcd7756 writes "The IEEE Spectrum magazine has an article about how capacitors made with a stolen formula for the electrolyte are leaking and causing motherboards to fail. Some computer manufacturers are admitting to the problem; others are hiding it."
I have a feeling that this is the case with many Abit motherboards. I have been the proud failure of two electrolyte leaking boards.
http://slashdot.org/article.pl?sid=02/11/04/175121 0&mode=flat
"Indeed, those who have repaired the damaged boards say that they have encountered crippled motherboards from Micro-Star International, ASUSTek Computer, Gigabyte Technology, and others."
;)
Well shit, and I thought the woolly jumper I wore when I built my box killed it. Now I don't have to feel so bad about sending our broken mobos back and claiming it arrived that way! (I'm joking, I'm joking! Nobody tell Dell or the Cowboy gets it.
Good thing my computer is from 97.
HTTP/1.1 400
That was fun, all Abit. The caps develop a bulge
:-)
on top and after a while they leak the stuff out.
Spontaneous reboots, blue screens and all sorts of fun.
I'm just glad it wasn't me doing the replacing
To quote: "The large volumes of passive content in any electronic device means that you have that many more chances for a product to fail". I can see that motherboards that pop are going to be a major pain in the backside, particularly for us small system builders, but what other devices use these suspect capacitors? aircraft? cars? lifts? phones? routers? Anyone care to speculate?
other than this is just further proof of the lenghts corporations will go to in order to make more money. Theft, lies, deceit, are all perfectly acceptable business practices these days, especially in east Asia.
This story has been circulating around for a long time, but this article is a good update on what's going on. I was very surprised to read that manufacturers actually threatened that guy who put a list of problem boards on his website.
You know, this is an all-too-disturbing trend. If you look at the behavior of media-giants, RIAA, MPAA, and now computer hardware makers - they'd all like to see us just locked in our homes, doing what they want us to do, seeing only what they want us to see, and not having any communication with anyone else... because if we can communicate with other people (i.e. by publishing a list of boards that are prone to failure), we'll realize just how badly we're being taken. That would eat into profits, and therefore should be made illegal. Heaven forbid consumers are allowed to make informed decisions..
Starting to sound like Soviet Russia?
This shows that quality comes at a cost. If you truly want to get good quality goods, don't expect to keep forcing the market to make cheaper and cheaper products.
Why would a company steal a formula such as this? so they ddn't have to pay as much for the 'real deal' and then henceforth could sell at a cheaper price and undercut others. When this happens quality suffers.
It has happened in many other industries and frank, I'm surprised it hasn't yet happened in something as stressed and pushed-cheaper as the motherboard and other componentry markets.
Rampant commercialism is causing problems like this
Tech Support : Is there a capacitor leaking on your motherboard ?
Customer : What ???? How dare you talk about my moma like that, tell me right now, who's that #@&*%$ capacitor ?
getSexySig();
It's starting to sound rather like America, and exactly the sort of capitalistic tyrany the founding fathers were afraid their republic could turn into.
You see, the difference is, in Soviet Russia the government owned the means of production.
In America the means of production are in private hands ( the very definition of capitalism) but own the government.
A subtle difference to the man on the street perhaps. After all, at serf level tyrany is tryany, but it isn't fair to slander it with the label of the great "evil empire." It's pure laissez-faire capitalism and a "free" wage slave is still a slave.
KFG
From the article:
"Zogbi cites tests by Japanese manufacturers that indicate the capacitor's lifetimes are half or less of the 4000 hours of continuous ripple current they are rated for."
4000 hours for the good capacitors? That's like 6 months of continuous usage. Surely shome mistake?
The batch of SunRay 1 terminals (first generation?) that Sun recalled due to a power supply problem have the faulty capacitors. They were the same brand of capacitors as some dual-CPU MSI motherboards we had that died.
At the time, I also wondered what other devices may have these faulty aluminium electrolytic capacitors, but it appears the answer is not many; probably due to their cost, they seem to be restricted to high frequency switched mode power supplies.
ESR capacitors
I read this as: 'ESR capacitated and leaking all over my motherboard'. Trying to get that picture out of my head. Failed..Going home, to take a nap.
The article speculates that a scientist stole the recipe for the electrolyte and sold it (probably for a lot of money) to the competitor.
How difficult would it be to buy just one capacitor from your competitor (for $0.05), open it up and do a chemical analysis on it?
karma police: arrest this man, he talks in maths; he buzzes like a fridge, he's like a detuned radio. [radiohead]
This is old news! (Sep. 2002) I submitted this to /. some time ago, but it was rejected...WTF?
Links to original (and informative) articles are Faulty capacitors and Passive Component Industry report.
In fairness most of the products affected carry a "Made in Tiawan" warning!
I checked over 3 epox boards I have here. They
appear to be fine. No suspect looking capacitors.
YMMV. I've been using an 8k5a2+ for a while and
I'm really liking the thing. Using it now. The
onboard sound is a PITA to get working with
surround and linux though. When you do give up
and get the OSS commercial driver, you'll notice a 30ms lag
in every game until you give up again and get
something cheap that works great with kernel
drivers like an es1371. Damn The Strokes are a
good band.
The most important thing any republican needs to know.
In the early 90s some "clever" manufacturer made a line of 486 mainboards with n Kbytes of "Virtual Cache" on board. The cache memory chips were fake, just empty plastic chips soldered to the board. The BIOS was also hacked to show an unexisting cache and of course soldered to the board to discourage analysis/reprogramming.
Didn't they mean to say that "Leaking Capacitors Muck up Fotherboards"?
only works if the govt. lets it work
Notice that mp3 trading is a classic example of the free market applying pressure to the suppliers and yet instead of heralding it as a success of capitalism in action the US and others are applying protectionism to the music industry.
CDs are too expensive. The only analysis I need is that people are putting effort into copying them for free. The market wants to see a CD album for somewhere in the $3.99 region.
The CD producers have been prosecuted for running cartels and still they whine.
There are places where the networks are not touching,and there are places where they are-Boeing's Lori Gunter
Basically anything that uses a switching power supply or a switching voltage regulator is at risk IMO. That covers most consumer and commercial equipment other than stuff that's physically too small to have one of these caps, or that handles only audio/radio.
Here we go again! Yet another base compound which was poorly made, put into a product, sold to others, put in their products, sold to OEMs, built into PCs, bursting and leaking, in the house Jack built. In this instance, instead of having one big company to point the finger at, we've got loads of little ones all over Asia. Fun!
I am just curious if the Asus or Abit motherboards that I own might be affected ??
All the failed capacators seem to be from the company JPCON. On my motherboard (that also failed) they are branded JACKCON.
So if you board has these capacators, I'd keep an eye on them.
It's been well established that the ECUs in the '90-'94 turbo Eclipses and Talons (DSMs) were made with substandard capacitors which would leak after several years causing the exact problems outlined here. Traces on the board would be destroyed and teh things would be left useless.
The difference? Mitsubish *never* acknowledged the problem. They just fixed it under the covers in '95 and never told anyone about it. Nice, huh?
Thomas Dorris
My old EP-8KTA+ suffered from this problem. I gave it to a friend who's an electronics freak and he checked out out.. said he's gonna try and replace the capacitors. I wonder if he'll get it to work again.
The effects were that first, I couldn't get it to run with my old power supply. Somehow the mainboard failed to switch it on. I had to use one that would switch on when u switched the power button on the backside. Then it just failed to work alltogether...Ah well.
If a train station is a place where a train stops, what's a workstation?
I've already lost a MOBO to these leaky capacitors. It really pissed me off, since it wasn't even a year old. And yes, it was an Abit board. I will never cheap out again.
I've got a bad attitude and karma to burn. Go ahead. Mod me down.
Power Supplies also use low ESR electrolytic capacitors. I'll bet some of the bad capacitors turn up in power supplies too.
Wansu, th' chinese sailor
While this is pretty bad for the home consumer, its not as bad as it is for a large company who buys in bulk. Many of the larger companies, and even some smaller, will buy say 150 of the exact same type of system for their workers for various reasons. If this is an issue with that board, thats 150 computers that could die out for a known issue.
How many companies do you know have checked what mother board is in the, say, Dell Dimension 1234x Desktop Model? What they look at is they can buy it in bulk under a business license, it has a certain size hdd, and its easy to replace/repair. The Mobo on a business class machine like this isnt even taken into consideration unless you are looking for onboard something.
This is a lanadmins worse nightmare now.
We've had el-cheapo board with sub-standard components around for years. Back in 1998 when I was working for a major UK PC builder/retailer, we had a very large batch of motherboards with this same problem. It was the biggest capacitors on the board which used to fail, sometimes blowing a chunk out of the PCB!
Like tinyurl, but one letter less! http://qurl.co.uk/
now the drip?
What the hell is the next STD for motherboards?
digital clymidia?
However, I'm glad that this is happening to boards that end up in the hands of tech-savvy individuals that can spot the problem. People who buy ABIT, Asus, etc... boards expect a lot from the product that they recieve and are usuially knowledgeable about the equipment that they run.
I could only imagine if this happened to a major computer company, how it would be swept under the rug (which it may already have been). I see that IBM is named in the article, so at least they are willing to accept the failures. IBM is one of the only computer makers that I trust anymore after the way that they handled their hard drive failure issues. Yes, they tried to fix the problem by changing the uptime specs, but in the end, they got the problem worked out without too much hassle to customers (hardware zealots excluded).
I would like to know if this problem has been documented by any users that aren't using products from the manufacturers listed in the article and their expierence with the equipment, service and support.
I don't keep a lid on my coffee so when I walk around I look busy -me
High-Reliability motherboards don't use electrolytic capacitors for their onboard power supplies anyway. You'll only find these types on the typical built like crap (found in the majority of PC's built today)high volume motherboards. 5000 hour lifetime is what you'll find for typical ratings on low esr and long life electrolytic capacitors. This is only 200 days of continuous use.
High-Rel designs use very low esr tantalum caps. like these T530 Series
Quidquid latine dictum sit altum viditur
Think of how many decisions in business don't even look good on paper anymore. Companies shedding devisions that, while making money, aren't making BOOKOO money. All the games like 37.5 hour work weeks on your pay stub. And all of those assine hoops they jump through for tax reasons.
Now if I go to company A and say, hey, for yor next data center upgrade I can save you 80% of the cost by going with Linux I would be laughed out of the meeting. If I turn around and say I can save 10% of your next round of computer upgrades by skipping the floppy, they might buy it. If I say that I will save them a fraction of a penny on a penny component by going with a noname manufacturer, I'd get promoted.
"Learning is not compulsory... neither is survival."
--Dr.W.Edwards Deming
they're a little older @ 2000 so they may be safe from this, how far back does this go...the article mentioned that problems started showing in 2001
:)
Good question. I don't remember exactly when I built that machine. The 800Mhz Pentium III was fairly new when I got it, and 'orribly expensive. I'd guess it was 2001.
I also remember being annoyed at these new 1Ghz CPU's that made my machine obsolete
The issue with electrolytic capacitors is this:
All capacitors have what is called an equivelent series resistance, or ESR (great: now we have 2 TLAs that are overloaded in context: RMS and ESR... )
The ESR only matters when there is a current flow through the device - a static voltage does not create (much of ) a static current flow - that is rather the definition of a capacitor. So if you are putting a DC voltage across the device all is well.
However, if what you are putting across the device is NOT DC, but rather DC with an AC component on top of it, then there will be a current flow as the capacitor tries to hold the voltage constant (again, that's rather the point).
However, due to the ESR, some of that current will cause heating of the device (power = I*I*R).
In caps with the good electrolytic, nothing much happens. In caps with the bad electrolytic, the electrolytic breaks down into hydrogen and oxygen, as well as boiling into steam. Pressure builds, and eventually the cap leaks. Since the other stuff in the electrolytic is caustic, your PC board traces rot away.
Now, at low ripple currents, this does not happen very fast, and any cap will have a long lifespan. However, as you approach the limit of the cap, the heating becomes the dominant factor, and the cap will cook itself fairly quickly.
That's where that 4000 hours comes from - that is not the cap running with a few tens of milliamps of current ripple across it, that is the cap getting amperes of current rammed down its throat, and running very hot.
That's also why you use tantalum caps wherever possible - tant's don't have an electrolyte, they use a very spongy tantalum slug with lots of surface area. They don't have quite the capacitance per unit volume that electrolytics have, but they don't leak, either. (but they do blow up real good (sic) when you exceed their rated voltage!)
www.eFax.com are spammers
I had a simalar experience with Compaq. It was on an older system (PII/300mhz) a few years back. I was on my computer one night, about to call it quits and I hear a loud POP! About 20 seconds later the computer crashed and went black. It did that often though, so I didn't think anything oddly of it. I assumed the pop was a tree branch or something.
So I come back the next day to turn on the computer. Nothing. It's dead Jim. So of course I open it up to see what's wrong. When I open it up, I saw the outside casing of a capacitor laying on the floor of the case, and noticed an odd indention in the case. Aparently a capcitor near the CPU on the MOBO just blew itself off the night before and put a dent in the case! I found on the MOBO where the capacitor was, and it had greatly expanded.
Now it get funny. So I call compaq, thinking that I can ask for a new capacitor, because it seemed to be an odd one that I didn't have a spare of. I call up and I say, I need to get a new part for my motherboard. He asks me what the problem is, not listening to my question. I said the computer won't turn on because of a problem with the MOBO. He took that as a cue to run me through making sure all the cables were plugged in, etc... At one point he thought I had it in sleep mode. Moron.
Anyway, he is like, well I want you to do this- and I said "well I can't really do that I know the problem is a capacitor on the motherboard gone bad". He tells me to hit the sleep button again, making sure everything is plugged in. I say I can't do that because the computer is taken apart on my desk. (to get to anything on those you had to take half of the Chasis apart to get to the mobo, which was now laying on the desk). He seemed rather taken back by the fact that I had even opened up the case, let alone taken out the mobo.
So he says to me "didn't you see the warrentee stickers" I said yea, but how else was I supposed to put in a Voodoo 2 instead of your crappy Rage card, and more memory. He seemed to think I was supposed to send it in for that. Anyway, he wouldn't send me a capacitor. And told me to get my warrenty stuff in order and then call them.
Morons at compaq. When I build systems for people, I expect them to take them apart. It's not like I overclocked it or anything.
Tibbon
tibbon.com
The early graphite-colored AirPort Base Stations had a very widespread nasty problem of failing because of two capacitors failing - two Lelon capacitors. Apple acknowledged the problem within a certain serial number range (PW940XXXXXXX through PW952XXXXXX), and fixed the problem for customers out of warranty. However, Apple continued using the same apparently-faulty Lelon caps, and graphite base stations continute to fail. All that's needed is to replace these two capacitors; see this site for background information, and a picture of the Lelon capacitor compared to others. It's significantly smaller, considering the voltage and capacitance ratings, than other similar capacitors. Since it's a Lelon (one of the brands mentioned in the article), and because one of the supposed benefits of the secret electrolyte was to be able to make the caps smaller, it now seems clear what went on here. Apple hasn't made the graphite base stations for some time now, but they kept using the same faulty Lelon caps until the very end.
For anyone who wants to repair their base station - the symptom is all red lights, continuing to power cycle, and perhaps even a faint hissing noise coming from the failed capacitors - I recommend Radio Shack 35V 220uF electrolytics (272-1029). They're small enough that the base station doesn't have to be modified (the hardest part of some of the suggested replacements), and seem to work fine. I've replaced a few with these now, and they've all been working like a charm.
It's a pretty simple fix though. Pull the ECU, remove the dead caps, clean board really well, $2 worth of good quality caps from Digikey and back in business. You can find detailed instructions on how to fix it here
Beauty is in the eye of the beerholder.
I'm willing to bet some of these capacitors ended up in cheap power supplies. I can't tell you how many ATX power supplies i've replaced in the last couple of years. Almost all of them smell like burning electrolyte when they die.
The old adage applies; you get what you pay for. I've since stopped buying $29.00 power supplies and sub $100.00 motherboards. Now I pretty much only use Intel server and workstation boards (unless i'm building an Athlon machine).
-ted
I brought it back to the place I got it http://www.nanosys1.com/, (I don't work for them, BTW) and they not only took my board back and RMA'd it, they sold me a newer one at 10% over cost. I think I would be greeted with confused looks and laughter if I went to Best Buy or Computer "Go-Round".
I now have my motherboard back (they explained that it's better to RMA through them, since they have a relationship with the board makers; more chance of MSI fixing/replacing it.) and my new board is chugging along nicely. I've been buying stuff from them since '97 and now that I've moved out of the Mpls area, I drive an hour and a half to do business with these guys. Happily.
A couple of years ago when i used to work on tech support for Packard Bell machines, we started getting people phoning up saying "my computer smells of fish" - most of them also complained that their computer was no longer working. Turned out the capacitors on the power supply were leaking - and for some reason the electrolyte smelled of fish.
I think maybe they just used cod liver oil or something.
--
cHris
We have replaced 25 Gateway E-3400 733Mhz motherboards. This is about half of the number that we have of this particular motherboard and model. Several capacitors are bulged and black stuff is leaking out the top of all these. We have several other models and speeds of the same model but it only is affecting these 733's of a certain age.
Kemet Electronics is near where I live. They make a considerable chunk of electronics and computer related capiciotors. I have heard they started cutting costs through "dynamic engineering restructure" (what exactly does that spin mean) - ever since they have had to lay off lots of workers because they are getting less and less orders due to quality problems.
Yell & scream & rant & rave... it's no use... you need a shaaaave ~ Bugs Bunny
Then you wouldn't be a "wage slave" in those terms. You would have risen to the status of "the help."
While there is a difference in class status between the scullery maid and the corporate manager their *state* is the same.
I used the term "wage slave" in the sense of ability to personally control ones working conditions or not. The classical form of wage slavery that you refer to still exists in a form though, but it's slavery to the interest charged by lending institutions rather than the company store.
If you need to borrow money to own a car so that you can get to work, you may well be a wage slave, even in the classical sense. It's just less obvious.
I fully admit that I'm an odd bird in today's world. By *my* standards the guy who barely scrapes by with his own pool cleaning business is of a higher status than a corporation president, no matter how much he "makes."
The classical American Dream is *independence,* not income.
KFG
Oh pulleaze. People are bad. When the right sequence of events occurs, evil can propogate from bad person, to bad person, magnifying itself. This happens regardless of the legal framework. Some frameworks are better than others. Soviet style socialism was probably more corrupt than global corporatism, and far more secretive. "In Soviet Russia..." this story would not have appeared in any of the state controlled news outlets. In fact, in the real Soviet Russia exploding TV sets were a leading cause of fire because the tubes were bad.
Exactly what is the conspiracy here? Are you trying to tell me that the Japanese engineer who stole the formula had a meeting with the contractors who cut corners, who then agreed to threaten the manufacturers so they wouldn't say whether or not they were guilty? And since I've got the Simpsons on my mind now, I want to know exactly how the saucer people were involved in all this.
No, this is not starting to sound like Soviet Russia at all. Don't you see the irony in complaining that we "can't communicate" by "publishing a list"? First, we are communicating on Slashdot. Second, the IEEE article already contains some preliminary investigative work that can be used to develop such a list.
No, it won't be easy to track down all the bad boards. Nobody ever said freedom was easy, but at least it's possible. So kwitcherbitchin, open your case, check the caps, and start asking questions.
For all intensive purposes, "whom" is no longer a word. That begs the question, "who cares"?
Paul.
You are lost in a twisty maze of little standards, all different.
This is the lowest thing Microsoft has ever done in order to sell additional licenses.
I'll see your senator, and I'll raise you two judges.
The USAF had, in the 1970s and 1980s, a field electronics reliability assessment program. About 1% of the electronics boxes in Air Force inventory were marked with a stencil that said "If this unit fails, send it back to ... for analysis." An Air Force unit tore the failed components apart and found out what went failed. And why. Components were pulled apart and examined in detail, using electron microscopes and other analysis tools to figure out exactly why the thing failed.
When the USAF found something, articles would appear in Aviation Week and other trade magazines, with company names, part numbers, pictures of failed components, and detailed explainations of exactly how the manufacturer had screwed up. This was very effective in tightening up quality control. It is, in fact, one of the main reasons minor components are far more reliable than they used to be.
This stopped during the Reagan years.
In the heyday of 3.5" diskettes, Sony had a unit in Japan analyzing failed diskettes sent in for warranty replacement. They discovered that the main cause of failure was scrapings from the shutter getting onto the recording surface. A redesign of the shutter cut failures way down.
In general, I think there is way too much emphasis on high performance and low cost these days, with nowhere near enough emphasis on stability and quality. This capacitor issue is just one symptom of quality generally going down the toilet.
Over the past year, I've personally had three motherboards become unstable or fail outright due to "exploding capacitor syndrome". The three boards were all different brands (MSI, Abit, and FIC). On the MSI, several of the capacitors literally exploded, and the board wouldn't boot any more. The Abit and FIC just went slowly downhill in terms of stability, as the capacitors became bloated and started to leak. I also scavenged a fourth dead board (another MSI with exploded capacitors) from one of my clients, who was throwing it out.
If you've got a little experience with soldering, it's not too difficult to replace the caps; this will generally restore the board to working order, provided none of the capacitors shorted out and took other components with them when they died. All you need are replacement low ESR capacitors (Panasonic FC series are good, you can mail-order them from places like Digi-Key); a soldering iron and desoldering tool (from your friendly neighborhood Radio Shack); some wire cutters (for trimming the capacitor leads after you've soldered them in place); and a steady hand.
If you get particularly unlucky, you may also have a fried switching regulator (MOSFET), or even burned PCB traces. It may still be possible to salvage the board even in extreme cases like this, but you'll need to buy replacement MOSFETs as well, and get a little creative with soldering some wires to the board to bypass the burned traces.
I've already repaired both MSI boards and the Abit. The MSI from my client turned out to be more trouble than I expected though (one of the VIO regulators was toasted). The FIC is next up on the operating table...
I opened up my machine sometime in December in order to inspect it for this very problem. Because I'd read in the previous slashdot article that Abit motherboards had experienced this problem, and my motherboard was (is) an Abit KT7-Raid (non -A flavor), I was particularly curious. Sure enough, two capacitors had clearly bulged open and were leaking paste, and three more were on their way towards failing. Interestingly, I hadn't had any observable symptoms whatsoever; I just checked the board on a whim.
I consulted the Abit website, and at the time they required the original 'invoice' from the motherboard if you wanted to have them repair the problem for free. If you have that paperwork, RMA'ing the board should not be too troublesome. I really didn't think I could find my old paperwork for this board. Abit offers to repair motherboards without original invoices for a charge of (as I recall) $25.00US. I think you have to pay shipping one way.
I considered using the services of the guy linked to in the previous slashdot article, but his prices were about in line with Abit's. That didn't really help any -- for the amount of money he wanted, I could just have just had Abit do it. I could also have just as easily replaced the board for $45.00 plus shipping on Ebay, but it probably would have just failed all over again.
I felt I could replace the capacitors myself, and as it turns out, I was right. Here's my advice to anyone who wants to try to do this repair themselves:
Good luck. Don't blame me if you scew something up, burn something out, or get someone killed. Send it back to the manufacturer if you're not up to the task, or don't have much experience in such matters. There is a touch to this sort of repair that comes from practice, I think, and the only way to develop it is to get your hands dirty. Or burnt. Or something. Previous to this, I had only a little experience with this sort of rework, so don't be too shy.
Besides, Natalie Portman demands a guy with trained fingers.
This is one of the reasons why I am an advocate for heterogeneous systems for server farms, clusters, labs, and such. If you have a server farm, you have some sort of tolerance for loss of systems. That redundancy works if systems fail at random, but this shows that that doesn't always happen. Let's say you have 200 machines with the same motherboard with the leaky caps. They all start leaking at about the same time, so they are all starting to wear out at the same rate. Now, let's say that the power company puts too much voltage on your line for a few seconds, the heat rises in the room, what have you, and all the motherboards blow at once. Or just 80% over an hour, it doesn't really matter. In any case, you're screwed.
Now, if you had differing systems, only half the systems are affected by the design flaw. The key here is to have systems with nothing in common. Power supplys, motherboards, cases, even cables must come from different companies. For example, half my server group is x86, the other half is SPARC. One runs Solaris, one runs NT. No matter what happens, with any design flaw, only half the servers will be affected at one time.
-twb