Help with green stuff on capacitors!

[milkyspit]

This is about my computer motherboard, not a flashlight, but I need technical assistance understanding a possible problem from the electronics gurus around here. Lately my PC has been randomly freezing, rebooting itself, and similar things. I thought it might be a buggy kernel mode driver in Windows 2000, but today the whole box rebooted with no warning, for no apparent reason, and I popped the hood (so to speak) to take a look inside. I thought maybe the CPU fan had failed.

Everything looked pretty normal, and all fans are working, but I noticed something strange: several electrolytic capacitors on the motherboard have varying amounts of hard, crusty green stuff on the tops. I've never seen that before. All the capacitors on this motherboard have crosses scored into the round metal tops, as if they're designed to "blow" gracefully or something. It appears that the ones with the green stuff have tops that are slightly bulged outward, and it's entirely possible that the leakage (if that's what it is) came out the center of the score marks. The non-leaking capacitors don't appear to be bulged outward at all.

My question is, what's going on... can someone tell me what's likely to have happened, as well as how such a thing might happen? Would a power surge do it? Seems strange because this system is running off a power strip with built-in surge protection as well as a UPS. What about overheating? But I've run cases hotter than this one with no such problems, plus I'd imagine the CPU itself or perhaps the graphics circuitry would be showing adverse effects of overheating far before capacitors on the motheboard would pop.

Or maybe this motherboard is just a pathetically designed piece of crap?

For what it's worth, the motherboard is an MSI K7N420 Pro with nForce chipset. It runs an AMD Athlon XP processor.

Do I have a problem here?

 

[WildRice]

piece of crap... sorry /ubbthreads/images/graemlins/icon3.gif

You are correct, the score marks are so the caps dont explode (pop). I have only seen caps pop for two reasons. 1. reversed polarity 2. overheating.

if the caps are leaking, then they are going open. the caps on the motherboard are for noise filtering on the power traces. no caps, more noise, more noise more voltage spikes, more spikes=unstable system.

Either get a new MB or you could replace the caps.

Jeff

 

[LED_ASAP]

Most likely the cap has popped. Just replace the bad one(s) with high quality electrolyte capacitor(s) with equal or higher nominal voltage and capacitance values and low ESR, and things sould be fine. Just be careful to blow/wipe off any of those green crust from your motherboard---they are highly corrosive and conductive. AMD processors are known to generate lots of heat /ubbthreads/images/graemlins/frown.gif

 

[milkyspit]

Jeff, thanks. I did forget to ask whether I could replace the capacitors on the motherboard myself, or would I generally need to buy a totally new one? I know these motherboards are pretty tight as well as multilayered, and blah blah blah... I can handle a soldering iron and order some parts, but is the task of manual replacement likely to go beyond the average hobbyist's soldering skills?

 

[milkyspit]

LED, yes, AMD processors can generate lots of heat, but you would think the mobo would be designed knowing that, no? And it's a whole bunch of crusty capacitors, like maybe a dozen, or roughly half of the larger ones that I see on the board. They're not all near the processor, either.

I've got the CPU way overspec'd on the cooling end. It's using an Alpha PAL8045 heatsink with an 8cm Panaflo fan mounted on top. I've monitored the CPU temps in the past but eventually stopped bothering, because they never rose anywhere even remotely near the danger zone, even when under heavy load.

It just galls me to no end that this has happened. I feel like I've been cheated by the mobo manufacturer. This system hasn't subjected the mobo to anything that would even come close to being considered unreasonable stress.

/ubbthreads/images/graemlins/mad.gif /ubbthreads/images/graemlins/mad.gif /ubbthreads/images/graemlins/mad.gif

What is the green stuff, anyway?

Incidentally, I'm typing this post on the system in question, which is the box I use for most of my CPF'ing, so I guess this thread is sort of about flashlights. /ubbthreads/images/graemlins/wink.gif

 

[WildRice]

the green stuff is oxidized electrolyte. I would recommend using solder wick as not to pull anything from the middle layers. you should be fine. as a side note, use a static strap and a grounded iron.
Jeff

 

[carbonsparky]

Sounds like you have some of those bad capacitors I read about here. Check with you motherboard supplier and see if it is covered under warranty. There was a post back a bit ago, telling of some bad Asian capacitors. These were made using a defective electrolyte.
Here is one of the posts, but it is not the one I was looking for.

http://www.candlepowerforums.com/ubbthreads/showflat.php?Cat=&Board=UBB6&Number=360438&Forum

 

[Eugene]

Yep the old capacitor poop syndrome. I have fixed a lot of broken electronics by replacing a couple capacitors. The other posters are correct, heat and reverse polarity are the main killers. In antique electronics long shelf life can cause failure as well. Warranty or good soldering iron time.

 

[NewBie]

> Here is the story:
>
> Problems!
>
> In September 2002, reports started to surface in the United States among
> brand name computer manufacturers that there were problems with low-ESR
> aluminum capacitors produced in Taiwan. Apparently, a poor-quality,
> water-based electrolyte had been supplied to multiple manufacturers of
> low-ESR aluminum electrolytic capacitors in Taiwan. Our intelligence
> sources
> in Taiwan later confirmed that as many as eleven manufacturers of
> low-ESR
> aluminum capacitors in Taiwan might have been exposed to the poor
> electrolyte. The electrolyte, which is a locally produced knock-off of
> the
> Japanese P-50 type water-based system, was produced without the proper
> additives required to stave off excess hydrogen gassing. Consequently,
> aluminum capacitors produced with the P-50 knock-off tend to fail
> catastrophically, by blowing open the rubber seal and leaking liquid
> electrolyte on the printed circuit board. These failures reportedly
> occur at
> half the rated lifetime of the components.
>
> Fallout?
>
> The fallout in the computer industry is still underway; however, the
> thought
> of potential recalls in desktop computers (laptops are not affected) has
> sent shudders throughout the high-tech supply chain. Even the
> possibility of
> responding to computer motherboard failures under existing product
> warrantees is extremely unpalatable, especially in the current economic
> environment. In addition, it is not known how many other market segments
> may
> have been affected by these poor parts, which can also be used in
> switchmode
> power supplies, modems, and entertainment boards.
>
> As this intelligence on the supply chain for aluminum electrolytic
> capacitors began to circulate in mid-September 2002, it also became
> apparent
> that many OEMs had not been aware of the situation because their CEM
> partners created a buffer between them and the problems in the supply
> chain.
> Concern has been building that CEMs' use of local suppliers in Asia
> might
> end up costing manufacturers more money than they would have saved in
> the
> long term. It is also speculated that the failures of the aluminum
> capacitors might just be the "tip of the iceberg," and that other
> component
> failures from low-cost Asian suppliers might be forthcoming.
>
> As this story continues to develop, it is highly probable that many
> Taiwan-based companies not exposed to the poor electrolyte might be
> inadvertently categorized with those companies known to have been
> affected.
> Regardless, what is now occurring is a shift in value for aluminum
> capacitors away from Taiwan and back to Japan. The value chain here is
> substantial, as approximately 30% of the world's supply of aluminum
> capacitors are manufactured in Taiwan.
>
> Many Western and Japanese suppliers of passive components are now
> feeling
> somewhat vindicated by their constant cautioning to the major CEMs about
> sourcing local vendors in Asia. Taiwanese companies will probably be
> faced
> with the daunting challenge of quality control, as Japanese and Western
> companies do all they can to exploit the present situation to their
> benefit-not just for aluminum capacitors, but for the wider spectrum of
> dielectrics, resistive elements, and magnetic products.
>
> Home Terms of Use Privacy Contact Us
> Copyright C 2000 - 2002 TTI Inc. All rights reserved.

 

[drs2000]

I ran into this with all of our PII access control system's Mobos.. My diagnosis was pretty much what the last comment said. The caps failed, due to either wrong/misquoted specs as far as ESR, or were a bad batch..
Hadn't heard there'd been a problem, tho these are about 2000 manufacture.

You could change 'em out, but you'll have to locate low ESR caps, and be careful when removing.. The board's plated-thru holes can be damaged easily. I usually put a blob of fresh rosin-core solder on each lead, then alternately heat each lead while gently rocking the cap out. Then clean up the holes with a _good_ solder sucker and/or braid, unless you're lucky enuff to have a good desolder station. Pull all the caps out that have bulged. Use an old toothbrush and scrub those empty spots with water until they're clean. Rinse, dry and inspect the areas under and near the caps for any corroded traces or small parts.. Fix any if they're badly damaged.. Don't get the stuff on you. It's a fairly nasty base until you dilute it, IIRC.

I suspect that you're getting some conducting going on that may be what's annoying the system, or the value's just changed too much.

YMMV. Good luck..

Yours, drs the been there..

 

[snakebite]

its the crap cap plague.
i have recapped over 1000 mobo's of various brand.
not to hard to do if you are good at soldering.
check out
badcaps
btw i am kc8adu there.
if you leave it off till you replace the caps it is a near 100% cure.

 

[LEDUser]

If you still have warenty RMA it (Warenty return)

 

[Jack_Crow]

Hi all,
Forgive the intrusion.

I was on the other end of the service business. Not mother boards, but monitors and switch mode p/s units.

I got a tip from one of the better guys that fix monitors for the video game industry.

There is a place in Mo. that markets an "ESR" meter. They advertise in Nuts and Volts.

It comes in a fairly standard pac tec box, has a meter movment on the front, and a wire set for test prods.

It measures the ESR of electrolytic caps and displays this value in ohms. So a brand new cap, lets say 100uF at 33v will show usually less than 1/2 ohm.

Now poke around in a four year old monitor and you will find some that are just fine, and the ones that are part of the power supply or flyback are beat to snot.

With that tool I made a lot of money. Video games, pc monitors, and more than a few switch mode p/s units.

It measures with a four wire kelvin system, using two standard probes. I hold them like chop sticks and work my way through a unit. Marking the failed/weak parts with a sharpie pen. The testing is fast. I can spot check a whole monitors worth of caps in about 5m. It much better to change all the weak ones than to get the five worst and have a come back. Caps are cheap and time is expensive.

The test voltage is low on the order of a couple of mV, and the test frequency is 100Khz.

One thing it won't take is power. Very low internal / source resistance inside the unit. Un plug and let the caps drain. It can measure different caps even if they are part of the same buss. About the only time you have to pop a part is if the ckit places two caps right next to each other on heavy pads.

Caps have a shelf life, and a service life. They are electro chemical devices and don't last as long as resistors or semiconductors.

This is a truly amazing tool. Also works on low value resistors, pcb traces and such. Coils will give wrong readings to to XofL issues.

Good luck guys
Jack Crow in Iraq

 

[snakebite]

the best esr meter. **** smith esr meter

 

[UVLaser]

Ya my old old computer had that problem too /ubbthreads/images/graemlins/icon8.gif But not on the mother board. It was the battery that kept the time when you shut down and all the other info that the computer needed to turn on. It still turns on but you have to reset the time, what port the mouse and keyboard uses, what size the floppy drive is, etc.

 

[milkyspit]

Is it possible that an LCD monitor could be experiencing the capacitor problem as well? I've got a ViewSonic VG175 monitor that I bought sometime in 2001. It still works, but for the past several months there have been horizontal wavy lines in the image. At first I thought the backlight was bad, but lately I'm wondering if it might not be yet another manifestation of the dreaded capacitor defect? The wavy lines begin farther apart and more intense, then as the monitor warms up they get a little less pronounced and move closer together. They never completely go away. Other than the lines themselves, the image is sharp, focused, and well backlit.

Thoughts, folks? /ubbthreads/images/graemlins/confused.gif