Slashdot Mirror
PCs Plagued by Bad Capacitors
Hawaiian Rules writes "CNET has a story
detailing a new threat to Dell PCs, Apple iMacs and other computers with Intel boards. This has been documented on BadCaps.net for some time, but the article also discusses what to do if you suspect you've got a case of the bad caps."
I've had a couple AMD boards go bad because of leaky capacitors.
Bad Caps have been a problem since 2002 at least. For awhile, I was making some bucks repairing Apple Airports, with all their bad caps.
I drank what? -- Socrates
Never buy brand new high-tech toys before they've actually passed major consumer testing.
It's the same for everything technological! Only through trial and error, consumer brute force sort of do they get the best product after 1-2 years for most products such as Dell's, i'd cite motor companies too but bah.
Coding projects blog - Code Slim
I had this happen to an old Asus board I had a couple year ago. It was covered on /. before.
Slashdot - Taiwanese Capacitors Leaking, Exploding
Watch out for all the 'Geeks popping a cap in your mother' jokes.
-Eric
Call Capman
Geeks Popping a Cap in... Oh, wait...
War isn't about who's right. It's about who's left.
If a car maker can get away with a cheaper, flimsier [insert part here], save a few cents on each car, and sell millions of cars, they can make a mo'load more profit than if they'd gone with the slightly better quality part on every car. Same thing here only with mobos and capacitors -- nothing new.
This sig rocks the casbah.
You can read the whole history of dying iMacs on Macintouch.
Strange women lying in ponds distributing swords is no basis for a system of government.
Nothing like a good dose of the cap.
Well, my doctor said... huh? oh, ok... never mind
My Dad bought a microcontroller-based water pump for construction work that had bad caps. You would think that companies would try to shave a nickel off a less vital part.
I had this happen on an Asus motherboard I got in 2001, noticed it when I was swapping out RAM in 2003. Board still works to this day, but you can see a line going from one of the regulator caps down to the PCI slots. I wrote down what kind of cap it was in case I was bored and wanted to replace it.. but honestly, after almost 5 years with this T-Bird board, it's not a big worry of mine. Still running, still over clocked, still a heck of a Linux system.
-=JML=-
"... a new threat to Dell PCs, Apple iMacs and other computers with Intel boards"
Not really a problem for me, at least. During the last 5 years, all PCs I bought were AMD based.
The Internet really has become quite a zoo. Once the chairman of IBM thought "there is a world market for maybe five computers". Now there's a server farm just for bitching about bad capacitors. We really live in an age of miracles.
--
make install -not war
It was easy to spot obviously bad capacitors once I knew what they looked like. The ones I notice look like little cylinders on metal legs, with a rounded instead of flat metal top.
My least favourite kind of capacitor though, is one that works properly, but has been put in the worst place possible so that putting the heat sink on that is supposed to match the CPU, is impossible. And you can't exactly bend those suckers over out of the way, so you have to buy another heat sink that conforms to the annoying motherboard layout.
Saskboy's blog is good. 9 out of 10 dentists agree.
This is no surprise. IBM fixed there issues with this TWO years ago. Just shows where some companies do better work of taking care of their customers. We had to go thorugh our data centers and proactively replace a bunch of server power supplies because of bad capacitors. All free, before they failed.
Ok, I realise most people (in general, not the /. population) probably wouldn't know which end of a soldering iron to hold, but it's not that hard to fix the issue..
Read the values of the leaky caps, get replacements, or near enough in value replacements. This will probably cost about $5.
Desolder the old caps, use a stainless steel pin to clean the solder out of the hole (since solder won't take easily to stainless), pop the new cap in (with the correct polarity), and solder it.
I had an asus board go like this a couple of years ago, it took me about 1/2 hour to fix the issue, but most of that was getting the board out of the case, and reinstalling it.
I called up asus, and had a runaround, before I identified the caps as the issue, and decided to fix it myself.
I doubt it's going to cost $300 million dollars to fix this. I'm typing this on a GX270, and it's had the motherboard replaced in it already, but I don't know if caps were the reason for that.
It's my work machine, first the hard drive died, so I called Dell and got it replaced, then the mobo died, and I just called Dell and got it fixed, I didn't investigate the issue myself, like I would have done if I owned the equipment, or if it was out of warranty.
Anyway, while it might cost them a bit in labour, the hardware's not going to be all that much, replace the first few boards with working ones, then refurb the retrieved boards, and use those to replace the dodgy board, rinse, repeat.
I have a Mac Mini that just died, I thought bad power supply first, but upon picking up the box I smelt leaky capacitors. After searching the web for a bit, I found the CNET webpage. so Bingo I though... darn.
The box was in sleep mode for the last couple of days. I only noticed it died when I couldn't see the illumination for the power LED.
I got hit with 4 bad Abit boards in 2001 back when it was first reported several years ago. I had assumed since then Taiwanese manufacturers would have stopped using the knock off capacitors or would have stolen/bought the correct formula if only to stop threat of lawsuits.
I guess the lawsuits never came. Maybe it would be one time sleazy lawyers could do some good? It is pretty sad that electronics from ten years ago are better quality than todays and they know exactly why and yet it isn't fixed...
ME NO UNDERSTAND!
warning: not suitable for younger viewers
Same thing here only with mobos and capacitors -- nothing new.
I don't know about that. I've been into this hobby for a long time and it seems as though quality in all forms of electronics has taken a major dive since the late 90s. I certainly don't remember any of my XT and AT boards ever having bad caps, but I've had my fair share with ATX.
Not new at all - quite a few Rev. A iMac G5s had this problem. I bought this 17" in November, and the bad caps finally failed in March. Apple sent me a new midplane and I swapped it out myself, but from what I hear they're now requiring people to take their machines to an authorized Apple service provider to get the work done.
I took pictures of the midplane/motherboard replacement process, clearly showing the bad/bulging caps on the original system board.
I was working at a local computer shop and we got a bad batch of Amptron mobos that had substandard capacitors on them. Ususally, when they failed, it was within one minute of powering up. So we'd put them on the bench, cover the caps with the manual and power them up. If the caps didn't blow after one minute we'd put them in the systems.
LK
"Hi. This is my friend, Jack Shit, and you don't know him." - Lord Kano
if you suspect you've got a case of the bad caps
yeah, this one time in college, there was this girl... it was my first time, not hers though... i didn't know...
oh! caps! never mind...
--iggy_mon - www.ananonymouskiller.com - Die Trying -
They use cheap capacitors in their baords!
Cue the 1.21 jiggawatts / flux capacitor jokes.
I inherited a bunch of 3 GHz P4 Optiplex machines back in '03 after they were decommissioned from a student computer lab. The university buys cheaper machines as they only keep them around in the labs for a year or so normally.
Well, I roped them together into a really nice Beowulf cluster for running my simulations and for the past 2 years I've had nodes die left and right. I'm sure the machines are out of warranty now, but I really hope Dell fixes these machines. I seem to remember Gateway doing this back in 2002. Now that the official word is out, maybe the computer department will take my word for it. What does a silly physicist know about computers and motherboards anyway?
Mike.
Mmmm......sacrelicious.
you all deserve a kick in the joules.
The problem with the Airport base station capacitor failures is described on this web page:
http://www.vonwentzel.net/ABS/Repair/
There are also instructions buying and replacing the failed parts, with good images. I followed these instructions a couple years ago very successfully.
The article says that the caps have "... a letter "X" stamped on the top." They are not stamped with the letter "X" - they are stamped to allow the caps to deform and vent the boiling liquid contents in a predictable manner when it fails. That is why the top of a failing cap bulges and not the sides.
..
Not that it always works - plenty of caps still just "pop" violently and spew their content across the electronics anyway.
So don't look for a stamped "X", chances are all your caps have them
No joke there -
Ford put plastic water pump impellers in Duratech engines found in Contours, Escapes, and others... a few cents off the cost, and they're MTBF is 40k miles - perfect. Just long enough to get out of warranty, but not long enough to not make money on the repair.
Now look, Dell's paying for it out the arse this quarter because they had to go fix all their Optiplex. Good - they should have paid attention and bought the three cent caps, not the two cent deals.
The caps were made by Nichicon. Nichicon has been in business for 50 years and has had, up to now, the reputation of building *the best* low esr high quality electrolytic caps on the market. I've specified Nichicon caps only in designs because they work better than anything else.
That's why this is such a surprise.
I know it's bad form to bitch about moderation, but I can't see any way that the parent is insightful. Nichicon has produced good caps for years. Manufacturers pay a premium for Nichicon caps. Something or someone fucked up a Nichicon. Has nothing to do with trial and error.
"Eve of Destruction", it's not just for old hippies anymore...
I like how at BadCaps they have pictures of Mobos with BadCaps.
Like how the hell are you supposed to tell which one is the bad cap?
If they had a probe or something next to the cap showing something kind of blinking red light or something...
Arash Partow's Philosophy: Be a person who knows what they don't know, and not a person who doesn't know.
Anyone else misread this as "what to do if you suspect you've got a case of the bad clap"?
Sheesh, I've got to start reading the headline first (although on /. that doesn't always clarify things either).
A large number of the PC's in operation there are the GX270s, my own desktop machine was one. The problem for me though was that before I was in the most expendable position in the bank, an intern for the IT guys so whenever a machine blew it was my pc that was taken. Finally I my workstation was changed, but to a GX280, luckily though the hard drive in my 270 has been saved so if this one goes bad I'll be able to grab a spare 270 hopefully and not miss a beat.
But that's not what happened. The capacitor company in question, Nichicon is, or rather was, the best in the business. Manufacturers pay a premium for Nichicon caps because they were the best available. The motherboards in question were made by Intel and Intel uses quality parts.
The problem is that Nichicon screwed up somehow, not that Intel got burned for buying the cheapest parts.
"Eve of Destruction", it's not just for old hippies anymore...
i bought one this month that some bad Capacitors in it spent like $80.00 USD on it at Microcenter, i put it in and 2 hours later Wham !. no more power supply.
but at least the guy gave me a new one without too much trouble since i had just been there not that long ago.
CH
My Abit KT7 board which had been running fine for over 4 years blew 3 caps this week.
Anyone got a replacement Socket A board that supports PC133 RAM and not DDR?
capbusters.
CTRL-Z!!!!CTRL-Z!!!!!ohgod, i swear i didn't mean it!
/. is what happens when geeks talk. get used to it.
*This* is how we'll all be forced to use trusted computing/soundcards with DRM.
They just dont make them like they used to.
If we both capitulate all this introductive assonance to the altar of CowboyNeal, then a sword of fire will come forth from his mouth and slew all those fools in charge of moderating these aweful posts.
without prejudice
In Soviet Russia, bad caps blow YOU!
Zap!!
According to the article "Various postings on message boards claim the trouble was caused by capacitors that were overfilled with a liquid electrolyte that helps the component protect the processor from excess power; convert energy from 5 volts to around 1.5 volts; and deal with current surges."
Capacitors Maintain VOLTAGE, Inductors try to maintain current. Time to return to school and
re-learn Components 101.
The definitive study, from The Computer Aided Life Cycle Engineering (CALCE) Electronic Products and Systems Center , is "Identification of Missing or Insufficient Electrolyte Constituents in Failed Aluminum Electrolytic Capacitors". CALCE actually took capacitors apart and analyzed the electrolyte.
To see if the excessive hydrogen was being produced by impurities in the capacitor foil, wavelength dispersive x-ray spectrographic (WDS) analyses of foils from a capacitor from the lot of Taiwanese capacitors known to bulge and foils from a capacitor from a lot of non-bulging Japanese capacitors were performed.
A small amount of magnesium was detected in both the Taiwanese and Japanese foils, and copper was detected in the Taiwanese foils alone (see Table 1). Ignoring the topical constituents of oxygen and carbon, the purity of the cathodic aluminum foil from the Japanese capacitor worked out to be approximately 99.1 wt%, which was within the limit set by Dapo. The purity of the cathodic aluminum foil from the Taiwanese capacitor was approximately 97.5%,which was below the minimum value stated by Dapo. The insufficient purity of the Taiwanese aluminum foil could cause gaseous hydrogen production that would not be impeded by a depolarizer, but the galvanic couples were not thought to be sufficient to account for the rapid production of hydrogen gas that was necessary to cause the relatively rapid bulging of the capacitor cans. There were other anomalies in the ion chromatographic analyses,chiefly variations in the amounts of ammonium and phosphate ions present. Ammonium ions in water form ammonium hydroxide, which is strongly basic. This raised concerns about the pH of the electrolyte in the bulging capacitors,as a review of the chemical properties of aluminum oxide - the dielectric - showed that it is slightly soluble in basic solutions (but not in acidic)[8 ]. Measuring the pH of electrolytes from capacitors from the Taiwanese lot known to bulge and from a Japanese lot that had not exhibited bulging showed that the electrolytes of the bulging lot were weakly basic (7 < pH < 8),while those of the non-bulging lot were acidic (pH 4).
And that's the cause - internal corrosion because the electrolyte has a highly acidic Ph.
Well there was this one time at band camp... :)
Had this infamous problem and they replaced it for free with the upgraded one.
It's not just motherboards that are affected - I've lost a couple of nVidia GeForce4Ti's that way.
It's now gotten to the point where I have to specify motherboards and graphics cards without GSC capacitors. Every single Gigabyte GA-7VRXP we've had has had bad caps develop over two years - three happened in the space of two weeks, just after the warranty expired of course!
"Nine times out of ten, starting a fire is not the best way to solve the problem." - my wife
Allow me to get my asbestos suit on, because I know this is going to piss some people off, but whatever.
Anybody who's been a hardware tech and built machines for any significant length of times, has known about issues like these for years. And the funny thing is, a lot of people wind up blaming Windows on problems that were really being caused by faulty hardware. In the early days of win 95, I noticed a lot of times that Windows was really unstable due to low end and sometimes bad hardware. Granted, windows is hyper sensitive to crap hardware and could of been made a little more fault tolerant, but a lot of times h/w failures gets blamed on the OS...
maybe this article will chill people out a bit
I lost my concept of community when my community lost all concept of me.
wow, i got an email yesterday from the campus IT group asking if we've been having problems with optiplex gx270s. and today in a meeting with our dell reps, they explained the problem in further detail. we've had only two of the machines suffer from it (so far) out of 15. from what they told me, it affected a relatively small number of motherboards, but their tracking isn't granular enough to be able to tell preemptively by checking the serial #. he didn't have time today, but offered to show me the exact capacitor in question later, so i can look out for the bulging/rusting look. i'll try to take a pic and post it up soon. basically it was some thermal issue and would cause the machine to randomly shut down. the gx270s (2/15) and gx280(7/150ish) motherboards have been flaky thus far, and it's pretty annoying. the optiplex line is supposed to be more reliable. i understand that they have little control of this, and it doesn't help that the gx280s were pre-orders (they were the first out with pci-express slots). but we still have about 50 gx240s from years ago that still perform like champs, the only parts we have had to change is the battery on the mobo on some. the dimension line used get the latest tech, while the optiplex was "tried and true". i guess being behind the times is a hard sell, even at the cost of a some reliability.
We bought six identical iMac G5s, plus two of us bought identical systems for our homes. Out of the batch of 8 machines, we have:
n program/).
1. Replaced the motherboard in two of these machines.
2. Replaced burned power supplies in one of these.
3. A third machine burned both the motherboard and the power supply. It has taken Apple over a week to ship the parts to be replaced.
Al repairs so far have been under warranty. Half the service transactions have been done thru the genius desk, half thru Apple Care. Both methods are painfully slow.
Also, on the iMac G5 Apple will extend coverage specifically for the capacitor issue, so even if your warranty coverage expires they will fix your machine at their expense (http://www.apple.com/support/imac/repairextensio
Pedro
----
The Insomniac Coder
I had a mobo from Gigabyte with some bad caps a while back. The board had lasted a couple years, and at one point I ended up putting it in a machine to put in my lab on campus. I knew there was something wrong with the caps, because the tops of them had sort of bulged open, and a bit of the electrolyte had leaked out and crusted on them. But since the board seemed to work fine even in that condition, I didn't worry about it.
A few weeks later, I was unable to connect to the machine from home, so I went down to the lab to see what was up. One of my coworkers said that when he came in that morning, the machine was already off and there was a nasty burning smell in the air, so I popped open the case to see that one of the caps and several of the surrounding components had caught fire, leaving a nasty black residue. Apparently, the fire had gone all the way to the other side of the board. Some of the wire traces had lifted off of the back of the board, and there was a black smoky streak straight up from that spot to the top of the board.
Fortunately, the processor, memory, and other devices still work just fine.
CowboyNeal actually posting a story on /.? I thought he was just a myth.
I dream of a better world... one in which chickens can cross roads without their motives being questioned.
As someone who has a literal 'pile' of dead motherboards from the last round, all I can say is not again. Worst thing is that they seem to usually outlast the warranty. Or atleast the symptoms dont show up till after the warranty expires.
Better get ready for another round of upgrades !
DSLIP Web Design and Content Management Australia.
That's right, business as usual. It's not just motherboards, it's nearly every type of componant. I've seen motherboards, power supplies, and monitors blow capacitors, too.
You have to remember, (A) by the time the caps have blown, the products are usually well out of warranty, and (B) the percentage of customers who will ever know that you used cheap capacitors is next to nil. That doesn't give you much of an incentive not to use the cheap ones, does it?
steve
Oh, you're not stuck, you're just unable to let go of the onion rings.
I thought a Dragon Plus would have been one of their best, and maybe it was, but after months of swapping out stuff trying to figure out what was crashing my system, I finally pulled the motherboard out, and it looked like I had dried bloodstains on it. This was just over a year after I bought it.
What needs to be remembered is that often a system with bad caps can damage other components, from memory to the CPU to hard drives, even cards attached to the PCI bus. This was devastating when it happened to me.
... the most important thing to remember when buying Dell computers:
Don't
"Nine times out of ten, starting a fire is not the best way to solve the problem." - my wife
Capacitors are a fact of like when it comes to electronics.
they go bad in TVs, Computers, even your clock radio.
thing is, we demand bigger and better and the power requirements are getting higher and higher.
Processors are getting faster and faster and that demands absolutely clean power.
I had 2 motherboards fail 2 days ago (how ironic)
Recapped with low ESR, hitemp caps and im good to go for at least another couple years.
Some of these caps are cheap, dont get me wrong.
Expecially on motherboards. but motherboards are cheap, too!
300 million seems like a big number. I wonder how many of these things they sold.
As far as Apple covering it up, what else is new.
My RevA iMac 1.8 G5 took a shit too. In the middle of a WOW session a cap blew, stinking up the entire room and killing power to the machine.
The best part was that after the mobo was replaced, it turned out the hard-drive had been damaged in the incident, but not enough to flat-out fail. It just kept silently corrupting the filesystem worse and worse, causing progressively worse crashes and truly freaky bugs. So THAT had to be replaced.
God bless Applecare, but fuck does the whole experience suck. Apple hardware isn't any better than anyone else's... it's just that they actually SUPPORT you when shit happens.
a bad rubber.
Unless you call that a cap. In that case, more power to ya.
If all you have is a hammer, everything looks like a nail.
I had an MSI board die in the middle of playing the hampster dance. I havn't watched it since.
calling them bad is a bit judgemental. dumb perhaps. http://www.gopapparel.com/ProductCart/pc/viewPrd.a sp?idcategory=6&idproduct=7
and of course... i happen to have one of the Intel boards listed on that sight... and of course... i experiance random shuttdowns on occasion... and weird heating issues... *Sigh* dam bloody hell dam... this always happens to me. =/
So that's what those cylinder things are. Capacitors! I always thought those things were just decorations, to make boards look more "techy" -- the PC equivalent of greebles, if you will. Maybe that also explains why my motherboard failed when I chucked a toy Death Star at it.
Slashdot requires you to wait longer between hitting 'reply' and submitting a comment.
I just got an upgrade two months ago. I may need someone to hold the flashlight while I check the caps. Here is the system diagram - http://guidant.com/condition/pop20.html .
A friend just gave me their IBM NetVista which had bsod'd, doing a little research I found out about this and supposedly (according to multiple newsgroup posts) the whole NetVista line had this problem. I checked it out and there are six capacitors around the cpu and two by the power supply connector and they are all bulging and leaking, some more some less. I talked to IBM and while they are replacing some of the motherboards free of charge this machines "Serial number is not on the list" of machines that qualify but as the tech was quick to point out they'd be happy to sell me a replacement motherboard.
thank you.
people always think I'm dumb for going cheap (second hand, bottom-tier, whatever) on cars and computers and electronics, but any brand that's produced in numbers high enough to get the product into best buys and comp usas is made like crap. they might have more features or something, but they're just as likely to break on you.
unless you're willing to pay the price for industrial grade products or for (in the case of some things, like audio) hard-core hobbiest stuff, you're just getting a plastic box of electronics produced in the third world and glued together by an assembly line. "quality control" (has there ever been a better piece of PHB-speak?) helps a little bit, but not much)
price hasn't equaled quality since your grandpa's day when everything was built out of painted steel and machined parts.
In Capitalist America, bank robs you!
Where I work at, we had a test lab which had hundreds of machines (Dell) with bad caps on the mobo.
This kind of problem really blows, especially in a test lab, because you end up seeing all sorts of strange software failures. Random reboots in the middle of tests. Crash dumps that don't make any sense (ex: how do you dereference a null pointer immediately after verifying that the pointer is non-null?). Impossible code paths executing.
Lots of one in a million errors that happen once or twice and you never see again. This will drive a programmer absolutely nuts.
We now have queries which flag machines with test failure rates much higher than the lab average... You'd be surprised at how often machines start to "malfunction."
I declare this horse officially dead. It's getting disgaussting.
Wow that was bad. But Ampere was harder.
Yea, those infamous pumps. When it dies, replace it with a metal one.
Anyway on this topic, I used to have a manager who was an "automotive engineer" in the 80s. His job? Re-engineer parts to "Save a few cents".
He worked that for less than a year, quit and went the computer route. And it sure did pay off for him.
Grump
Is it true that more people vote for the winner of American Idol, than vote for the president? -Ali G.
i had one in a standard el-cheapo power supply. impressive bang, cloud of white smoke, box full of cap shreds.
You don't want to know how many iMac G5's I've had to repair due to bad capacitors.
Okay, I'll give you a hint: I live and work in the same town as the single largest installed base of Macs in the USA.
We started seeing a number of strange security failures during boot on older IBM 300PL machines:
The RF-ID Tag has changed.
System-security - Keyboard is locked.
After which the system was dead - it would never boot further.
Turns out it had nothing at all to do with the RFID device inside (an inventory control tag). Instead, after trying everything under heaven and earth to revive the systems we finally noticed that the three 560uf caps in the processor VR had bulged.
We pulled the three old caps and replaced them with a pair of 1000uf units and viola! the machines were restored to their former glory.
Perhaps off-topic, but does the number of capacitors correlate to the design, i.e. is a soundcard with lots of capacitors of less design than a comparable with less capacitors?
We hae 75 or so HP (Compaq design, hence Hewlett-Paqard) d530 USDT desktops (the ultra slim desktop). These have had two problems which have required the replacement of every single motherboard under warranty and about 30% of the hard disks.
Firstly, the design for cooling is terrible - the hard drive bay has a fan by it, but it is blocked by the structure of the bay, so the hard disk essentially is uncooled. The Maxtor low profile drives seemed to be very susceptible to heat - we had a failure rate on hard disks of around 20% of our installed base. (The d530 USDT generally runs very hot).
After 9 months of having the machines in use, then the capacitors began to fail - either just bulging (and then one day, the machine wouldn't boot) or actually bursting (not explosively, just seeping electrolyte). Our HP agent ended up replacing every single motherboard under warranty after the failure rate reached 30% in less than a year of use.
I have to wonder if it was lowest-bidder parts. So much for getting reliability when buying a name-brand.
Oolite: Elite-like game. For Mac, Linux and Windows
The funny thing is that I have a Sony SLV-R5UC S-VHS VCR that had its power supply capacitors fail way back in the mid 1990s. Guess what? These were Nichicons.
Well, I think AMD do create reference designs at least.
But yes, I think the OP was refering to the Intel x86 architecture, not the Intel's current physical products range.
Yes.. you'd get charged
There is another point.
In the market sector where bling is the thing the motherboards are a fashion item, along with the processor and graphics card. Basically, this season's motherboard is out of fashion next season. Hard core gamers (at which these are targeted) will have swapped out the motherboard in a year or so, long before most of the motherboards with faulty capacitors will fail.
Anyway, why should the motherboard manufacturers care? The products will be out of warrenty anyway.
Agrajag: "Oh no, not again!"
The Nichicon website states ISO2000, which means the batch numbers should NAIL the matter - we just need them to post which batches have issues.
Note - It may not directly be the caps fault alone - poorer cooling, and el cheapo PSU's that pump out spikes, may be a contributing factor - Just open a PSU and see all the caps inside - they wear out too!
AFAIK - The bigger the cap the better - Minatures have a shorter life.
Slightly offtopic, but I've had this problem with a compaq presario 6000 awhile
back. The problem was the caps were mounted too close to the northbridge heatsink
(which runs unusually hot) and they began leaking electrolytic fluid onto the
motherboard creating a mass of other problems. Thankfully the damage was not too
severe, all it took was some old caps from the junk box, some wires and it was
fixed. They are now resting comfortably in between the PCI slots. I've seen other
compaq presario 6000's do this too so if you own one, pop it open and check the
caps just below the northbridge heatsink, if the tops are puffed out, remove them,
replace them, and relocate them and the board should survive.
Perfectly good electrolytic capacitors can fail because the power supply design engineer tried to save money by using capacitors that were barely adequate for the circuit they were used in.
Mea navis aericumbens anguillis abundat
When this story first came out the last line was "All the faulty capacitors have the letter 'X' engraved in the top."
I posted a comment to the effect that *all* modern electrolytic capacitors have a cross cut into the top - it gives them a point where they can safely rupture rather than just exploding if something goes wrong.
Now that line seems to have disappeared from the story, and so has my comment...
We have alreay replaced almost 200 Dell GX-260/270 system boards out of the 450+ we have at work that are being recalled and repaired under warranty by Dell. I guess you get what you pay for. I would rather spend a few dollars more and get better quality control.
I have a 6579-PDU which may have this problem. At least it's a mATX so I can get another MoBo from Tiger.
I also have a PC300PL which has started to show this problem. But since the 300's are those strange ILX form factors or whatever - with the daughter boards, it's not worth it, just chuck it out and scavenge what you can.
I have several computers that I have built from scratch using motherboards
made by asus and tyan. Anybody know about capacitor problems on these?
- You're really sure you have bad capacitors.
- You have successfully removed capacitors from a board before.
- You have the right tools:
- A fine-point soldering iron, 47 to 150 watts. NOT your typical 20-watt pencil iron.
- A solder-sucker.
- Known good capacitors:
- Not from Rat-Shack.
- Not salvaged from a dead car stereo
- Same uF.
- Save Volts
- Rated for HIGH RIPPLE CURRENT.
- Rated for 85 or 105 degrees C.
- A grounding strap for your body and soldering iron.
- Willing to take the 25% risk of killing the mobo anyway.
The reason for all these cautions is that mobo power supply capacitors are highly stressed-- those square black FETs are hitting the caps with 30-amp pulses about 200,000 times a second! Your basic Radio-Shack 49 cent capacitor can't handle this kind of stress.(Best bet is to order them from Digi-Key, they list the full specs.)
You also need a big honkin' soldering iron as each of those capacitor leads are soldered to many layers of copper foil, which make excellent heat sinks. It takes 50 to 100 watts of heat to heat up all those layers in an expeditious fashion.
I would first practice this art on an old scrapped motherboard. A true geek always has a few of these around. Practice your unsoldering technique until you can get a capacitor off (no jokes pls) in 20 seconds with no damage to the board.
Don't ask me how I learned all the things not to do.
Anybody want to buy a few "as-is" mobos?
Working in desktop support, I have to change a dell motherboard (usually an optiplex) probably twice a week.
Your sig(k) has been stolen. There is a puff of smoke!
Nichicon appears to be the only company manufacturing those short (~5mm high) axial capacitors, and our repair facility had to order them directly from Japan, as I wanted the 105 degree C rated caps, in contrast to the 55C rated caps that were installed in the radios. In South Florida, you could easily get in-dash temperatures over 200F: 90F air temperature outside the car, sunny day cooking the inside of the car.
The 200X cap problem seems to be due to the fact that all the caps were coming from one Taiwanese manufacturer. The dangers of one group cornering the market on anything.
I drank what? -- Socrates
We use Dell GX280's here at Rhino Video Games for our store computers and they're dropping like flies. We've already had to replace 5 motherboards just at our headquarters and now they're starting to go out company wide at the stores. Five might not sound like too much but when you're only a 100 store company it's quite a few. Needless to say when a store's main computer goes out that's not a good thing for business. Creates a lot more work and frustration for us in IT as well as for our customers.
I will forever be a student.
I had a Dell GX270, started shutting itself down randomly. Pulled it & started playing with it, the diagnostics came up with nothing. It was only while talking to the Dell tech (who was actually pretty helpful) that I gave a good look at the board & noticed the bad caps - I had seen them on another non-Dell board a couple of years ago. They next-day aired me a replacement board & we were back in business. I can see how this will end up costing them a lot if it's widespread - tech time, next day air shipping, return shipping, and a lot of people who will need an actual tech to come out to swap out the board for them. I bought a number of these machines at the same time, I'll be opening them all up to see that they're ok.
This week I've had the suso.org server fail nearly every day right after the backups start running (read, heavy disk access). One of my customers is a electrical engineer and pointed me at the bad capacitor problem on Wednesday, then today its on Slashdot. Usually my servers are up for at least a year at a time.
;-)
I think I'll be replacing my motherboard ASAP.
The problem is, how do you know if a motherboard company has fixed the problem by moving to better capacitors. I hate to find out in another 4 months that the capacitors have gone bad. I can't exactly replace the capacitors on a server since the risk of screwing up the operation is high. Maybe I can convince Dosman and Zach to come over from the Packet Sniffers to help me out, since they live in Bloomington as well.
But the demands on electronics are way higher now. The power density, temperatures and frequencies are much higher now (remember those 486s that ran only slightly warm with no heatsink?). You get more current being constantly pulled through the capacitors at a higher frequency (although that applies more to the switchmode side of things, rather than the linear regulators around the CPU), with hot air being blown around.
It's pretty normal in all fields for high performance to equal tighter tolerances and less room for error. It's more than just "the good ol days when things were made properly", although that is a factor too. There's a shitty cycle of consumers shopping for the lowest price and companies compromising products in order to compete. I've noticed that if you try to buy outside of that cycle you pay way more, instead of just the cost of better components and a slightly more complex design. This must be because a slightly more expensive product doesn't acheive a high enough sales volume. But these better products aren't enough better to account for the big price jump. And so the vicious cycle continues. Booooo.
My company bought 120 of these Dell PCs. Of the 120, we've had 115 of these stop working after a reboot due to a leaking capacitor. We are waiting for the last 5 to "pop" too. Fortunately, Dell's been pretty good at getting us replacement motherboards. I wish they would just recalled them instead of us having to wait for them to break. It seems pretty pathetic that 100% of them break, yet they don't do anything about it. Keith
Support bacteria. They're the only culture some people have.
Buy a new one and replace it. Granted they should not be made to low quality and not everyone can do that, but it's not that terribly big a deal, is it? I mean, I have a TV where the tuner is going, and it's out of warranty - if I want to keep the TV, it has to go to the shop for repair.
My team got a bunch of the d530s (they are CRAP!) last year.
We have 5 people using these machines. We have had 8 replacements in the 18 months we've been using them (yes, each and every one of us has had at least one machine go on the blink).
I can't believe our company standardized on these machines, but I'll be glad when we get something new. FWIW I have a ThinkPad T42 now that I'm happy with, but I use the HP/Compaq machine for coding.
You can accomplish anything you set your mind to. The impossible just takes a little longer.
New!
For fucks sake, the computer industry has been plagued by shit capacitors for over SEVEN YEARS!!!
No one lears to use good quality caps because by the time they usually die people just upgrade.
Shit like this would never last as long in the consumer electronics industry (well, SONY can get away with shit drives in PS2 for as long).
Well I've wrestled with reality for thirty five years doctor, and I'm happy to say I finally won out over it.
... the article also discusses what to do if you suspect you've got a case of the bad caps.
From the fine article:
If there is a brownish substance oozing from the [caps], check your warranty and contact your computer company.
Too bad opening the unit voids the warranty:-)
I had one of those IBM All-in-one PC's It had a 17" LCD with Mobo,HD.. Built into the main case.. After about 8 of the capa-asiters leaked I called IBM, I was about 3 months outta warranty, and even though they knew it was Manufacturer defect, they wanted to charged me $350.00 to ship me a new Mobo.. For a P2-400 !!?? I now, buy all my pc's from DELL, at least they don't try to stick it in ya. I don't mind Companies having problems with their equipment, thats part of life. Owning up and fixing the problem is what I consider superior.
Ad eundum quo nemo ante iit!
In all probability, this problem had it's roots in the 70's and 80's when the Asian capacitor makers drove the North Americian manufacturers out of the market. The Asians started selling parts with +/-10% tolerances vs. the North American parts which were +150%/-10% tolerance. What wasn't advertized was the North Americian parts were expected to meet the -10% tolerance at the end of the part's life. The Asian parts were allowed a -50% tolerance at end of life. Obviously, the Asian parts were cheaper and with "tighter" tolerances. Not only that the Asian parts were smaller.
Well fast forward 25 years the parts that worked perfectly well in non demanding consumer gear found their way into computer gear. Bad move... The high frequencies found in computer gear causes excessive ripple current in the capacitors, resulting in excess internal heating. A mere 10 degree C temperature rise in the core of a capacitor will cut the part life in half. Most engineers aren't aware of (or don't pay attention to) the aging characteristics of these capacitors. They only look for the least expensive parts.
Given my experience, with capacitors, I wouldn't expect the parts identified in the article (and yes I have designed with those parts) to last more than three years in the application. It isn't an issue of bad parts, it is an issue of bad engineering using the wrong part for the application.
Comment removed based on user account deletion
Is this only the 4 banger Duratec's or does it affect the V-6 too? I have a 99 Taurus with 173K miles on the 24 valve engine and I haven't had many problems so far, but it did get a little hot during the peak of the summer months, just wondering if I should be looking at replacing the water pump.
There are 4 boxes to use in the defense of liberty: soap, ballot, jury, ammo. Use in that order. Starting now.
I worked at HP as a contractor (before the big merger).
I caught some heat from a vice president or something for posting a complaint about
one of HPs software products that we were forced to use. There were public newsgroups
and internal newsgroups. I was encouraged to stick to the internal newsgroup.
It turns out the product in question used XML configuration files which referenced the DTD
out on the internet. It would not work with a proxy server. I tripped across this
trying to run the tutorial. It seemed to me like a problem that anyone in a
large corporation at the time would run across starting to use this product.
HP had bought the small company that produced this product and ended up discontinuing
the product. They later gave their external customers the migration path of using
the market leader -- the same product we couldn't use internally the year before.
I wonder how their customers felt fighting similar problems to implement on this product,
only to have it discontinued later.
The article said that the caps are only designed to last 7 years. Doesn't that kinda suck? Even with good capacitors, your computer is designed to die on you. Now, maybe you think that 7 years is a long time for a computer, but in many places, computers stay in use longer than that.
A) Solder-vacs are much easier to use than A solder-sucker.
B) I prefer the term solder-pult over solder-sucker.
C) When in a pinch, I find that heating the solder joint than quickly banging it against a table, oddly works very well.
These are not those "capacitors invented in the thirties." They type of capacitors used in this application are of the ultra-low ESR variety. They have a frequency response and other characteristics that were formerly attributable only to dipped tantalum caps even as late as the 80's. They are typically made from an organic material and they are more sensitive to stress than those old chemical electrolytics.
These capacitors are products of very modern technology, and so it can be expected some plants that were used to only producing the older technologies might fail.
I experienced this exact problem with my OptiPlex unit at work -- random system shutdowns & video failure. A tech said we'd had 65 Dell machines suffer similar failure in one department alone.
That's not exactly true; that's a guaranteed 7 years at a given temperature extreme and max-rated-spec usage pattern. Essentially, in a situation where the cap is being abused at the very limits of its ratings 24/7, it is rated to last seven years. The specific criteria of the test are far in excess of what the cap will see in real usage. Good engineers spec caps that are overkill for any given use.
Monty
However Edsyn probably prefers that you use their trademark, Soldapult.
I see even classic Slashdot is now pretty much unusable on dial up anymore.
"Not enough information"
Both manufacturers have sold mobos with bad caps. That said, I've owned a few of both and none of those ever developed the problem.
Monty
I have a few RAIDs built of consumer drives... If you lost 20% of your hard drives under warranty, you were doing better than I'd expect. All my boxes are overcooled and on-drive temp readings are in the low thirties Celcius... and I still lose 20-30% of the hard drives within two months of usage (depending on batch; the worst recent case was a 9-drive RAID5 built from Samsung Spinpoint drives, that one had four of nine die first month, then about half of the RMAs failed. Of the drives that did not fail, no errors after three years).
Monty
I think the Caps were fine in my last EPOX board, they wern't buldged at all.. but they were hot. One day when I came home, I was told by my brother that the computer had been on for 20 minutes but wouldn't boot up. I went to take the lid off, and could feel things were very hot. The coils were too hot to touch, and the caps were so hot i could pull a few of them right out of the board! Of Course that board never worked again, I got a new one, and a better replacement for my $20 power pupply.
I seem to recall a rumor that it was a stolen electrolyte recipe that was flawed (perhaps deliberately) and when used, produced the exploding capacitors. Anyone else remember this?
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
seems the problem was deeper than the stolen formula. I kind a sensed that because the problem lasted a lot longer than what one would expect from the one time fault of a single company. Come to think about it more now, it is suspicious that the problem was blamed on a one, unknown company with the story that sounds like taken from a cheap detective novel.
Your motherboard manufacturer could stick with the same capacitor manufacturer safely (most likely) unless the manufacturer was a REALLY shady company.
What happened was that a few years ago, it was discovered that one of the electrolyte manufacturers (basically a supplier to the capacitor manufacturers) was using a substandard mixture. By the time this was discovered, almost all capacitor manufacturers had bought at least some of the substandard electrolyte. A few years later, it's becoming obvious which capacitors used electrolyte from that bad batch (or batches.)
I forget most of the details, but even many of the traditionally reputable capacitor manufacturers were affected to some degree.
retrorocket.o not found, launch anyway?
"I'm not sure why the fluid was corrosive"
Electrolytic capacitors, by definition, have an electrolyte in them.
Electrolytes are corrosive.
retrorocket.o not found, launch anyway?
A lot of knowledge never makes up for bad judgement. It's broke, what you do won't make things worse. This is a case of little to lose and something to gain.
The board is dead or flaky because it has cheap caps. Do you think putting new cheap caps will be worse? The worst you can do is screw up the traces with a cheap soldering iron. Then your dead board remains dead and you move on.
Back in 2002, I fixed a board this way. The cheapest caps from a reputable dealer cost me less than $10 and the board still works. I had little to lose and some time. It was worth the time and money. It cost much less than buying a new motherboard. It has run continuously and still serves as an email spam filter and back up computer.
Friends don't help friends install M$ junk.
When in a pinch..... sure, the hot solder (now in liquid form) drops neatly off the board. Used to to do this myself until the day when the drop of hot solder fell between my sock and my shoe. Now I use the right tool. Everybody gets to learn, I guess : good luck to you, but don't wear lace-up shoes.
How many beans make five, anyhow ?
I am the proud owner of an old Toshiba Sat Pro ;) which was beginning to get bad cracking around the catch, I found it quite easy to take out the screen, drill out the crack tips(to stop them propagating) with a small drillbit then laid up fibreglass over the outside of the case to strengthen the area. I then sanded it back so that it is only 1-2mm thick with smooth edges. It has worked well and I have got at least another 3 years out of the repair and still going strong. I did not have a service manual for the Toshiba so was going by guesswork with the screen removal but I hear Thinkpads have quite good service manuals available so you may be able to dismantle it a little easier than I found the Toshiba. The material for the Toshiba was the old grey brittle plastic they used in the mid nineties so it was quite suitable for the fibreglass, if you attempt it on the Thinkpad you may want to do a quick test on the floppy drive or a surface you wont see to test it adheres and doesn't damage the finish near the area you are working on. It may look ugly but it works.
I am the proud owner of an old Toshiba Sat Pro ;) which was beginning to get bad cracking around the catch, I found it quite easy to take out the screen, drill out the crack tips(to stop them propagating) with a small drillbit then laid up fibreglass over the outside of the case to strengthen the area.
Interesting. I have been contemplating gluing some flat plastic to the top (when closed) of my girlfriends TP, to get some more life out of it for her.
I wonder if I should make a mold of my Sony's base so that I may be able to copy it in fibreglass some day when I need to? Have you ever worked with carbon fibre? Any ideas if a hobbiest could work with it on the cheap?
Maybe I should cut my losses and sell my Sony when the new x86 Powerbooks come out. Assuming they will be equal to my criteria for buying this Sony, 1920x1200 LCD, 2GB DDR2, quick Pentium M (~2.13GHz).
War crimes, torture, lies, illegal spying... Would someone give Bush a blowjob, already, so he can be impeached?
Anyone else noticed that many video extension cables and headsets etc. tend to stop working properly in a few months? I suspect that the copper inside is too brittle.
6 cylinder 2.5/3.0L DOHC were the Duratech engines I was speaking of. Not sure about the 4 - never worked on one.
:)
:(
My 99 'tour lasted 136k miles (95% highway) before I traded it in - luckily, I did not have to replace the water pump. But it was a very common failure... as always, your milage may vary
When the pump fails - it starts to freewheel in the housing, causing lower flow of coolant, but not total failure. You'll notice a slower to heat up cabin, warmer engine temp (remember your temp guage will not show a 30-40F swing in temp, they are buffered to prevent false positive dealer visits... so when it starts to run warm on guage - you're cookin' your motor.)
After a while of semi-freewheeling, you'll start to see bits of plastic in the overflow tank... and then eventually - no more pump.
Get a metal impeller replacement and toss it in the trunk, if you don't plan on replacing it now - you may not have a choice one Sunday night when the parts stores are closed or when the closest pump is three states away.