Unexpected reboot problem - Possible sollution -

[Johnny Solar]

Hi

Sorry for the major crossposting, but I feel that this information has a
better
chance of reaching the people in need, when they use the various search
engines to find an answer to a quite common, but so far, very mysterious
problem.

The problem im talking about is computers that for no apparent reason reboot
(no not because of viruses or worms).
In windows, typically you would only see an entry in the system log like
this;

Previous shutdown was unexpected


There are no other log entries, no bluescreens, no memorydumps, just this
entry.
When tracking the problem, people tend to find nothing wrong with their
hardware, memory, powersupply or with their OS.
There have been discussions about this subject, touching things like memory
problems, powersupply problems and so on.

A very possible explanation:

It may be that your motherboard has the types of capacitors/condensers that
fails after a while, we've had two different brands of motherboards with the
same problem on them, one VIA chipset mb and one MSI motherboard;

Typically the computer started to reboot without warning at random times
after about 2.5 years of troublefree service.
It was nothing in the windows 2000 log except "Last shutdown at xx:xx:xx was
unexpected"
After testing the computers to the best of my knowledge i finally contacted
IBM support who then changed the motherboard on my computer, no questions
asked.

The sollution to the mystery:

You can see it easily by opening up your computer and check for the
components that look like small barrels around your processor (can be placed
elsewhere too though), these "barrels" are called capacitor or condensers.
These capacitors/condensers are supposed to be flat ontop (usually they have
a cross marking too), if they bulge upwards ontop (like a cone, tuch it with
your finger and feel it), it means they have failed, and your processor is
no longer getting a stable voltage/current, causing it to reboot at random
times, just like if you had overclocked it too much. The aim of the
condensators as it was explained to me by the serviceman from IBM, was to
keep the voltage at a stabile level towards the cpu, as it can fluctuate
during uptime.

If this is the case, contact your dealer and get a refund or similar,
becuase it's not a user error, but a vendor error.
The capacitor/condenser type was faulty produced and they are common across
a wide range of motherboards, so it can happen to IBM, Compaq, Dell and the
whole lot out there.

This problem was most common in motherboards produced around 1999-2002 i
think, but offcourse this may happen on some various brands today for all i
know.


Anyway, I just wanted to contribute to a sollution to a very annoying
problem for those it conserns, it's almost impossible to know/see it if you
aren't an expert, but now you know.


Best regards
Johnny Solar

 

[Ben Myers]

Johnny,

Yes, this is the well-known capacitor problem that plagues mostly the cheap
brands of motherboards manufactured out there somewhere on the Pacific Rim. In
some instances, capacitors actually exploded with a loud popping sound.

.... Ben Myers

 

[Johnny Solar]

Well,

I don't know about 'cheap'; Our MSI mainboard that failed was a dual
processor thingy which was indeed very expensive, and the VIA chipset mobo
which had the same problem, was installed in an IBM small to medium server
which wasnt very cheap either.

The capacitors were probably very cheap though

JS

Johnny,

Yes, this is the well-known capacitor problem that plagues mostly the cheap
brands of motherboards manufactured out there somewhere on the Pacific Rim. In
some instances, capacitors actually exploded with a loud popping sound.

... Ben Myers

 

[Rob Stow]

Well,

I don't know about 'cheap'; Our MSI mainboard that failed was a dual
processor thingy which was indeed very expensive, and the VIA chipset mobo
which had the same problem, was installed in an IBM small to medium server
which wasnt very cheap either.

The capacitors were probably very cheap though

Cheap <> Inexpensive. There are quality aspects to be considered also.

It only takes one "cheap" capacitor to make the whole
motherboard "cheap" - in much the same way that a chain
is only as strong as its weakest link.

JS

Johnny,

Yes, this is the well-known capacitor problem that plagues mostly the
cheap

brands of motherboards manufactured out there somewhere on the Pacific

Rim. In

 

[Ben Myers]

Johnny,

MSI motherboards have been especially prone to capacitor failure. No matter how
many processors an MSI board supports, I have to place the brand in the cheap
category.

Capacitor failure of a motherboard in an IBM server is a real surprise. The
board was an IBM-branded board? Unfortunately, IBM has also become as
susceptible as other name brands to the use of cheap motherboards and other
components in many of its systems.

Even though VIA chips themselves have a decent reputation, I have also come to
associate VIA with cheap, because of the motherboard manufacturers to whom they
sell. Same problem as with nVidia, who sell their chips to anybody, even
companies with pathetically low quality and manufacturing standards. As a
consequence, there have been quite high numbers of failures of video boards with
nVidia graphics chips, mostly because of overheating, inadequately small heat
sinks, failure of cheap on-chip cooling fans... Ben Myers

Well,

I don't know about 'cheap'; Our MSI mainboard that failed was a dual
processor thingy which was indeed very expensive, and the VIA chipset mobo
which had the same problem, was installed in an IBM small to medium server
which wasnt very cheap either.

The capacitors were probably very cheap though

JS

 

[Tony Hill]

MSI motherboards have been especially prone to capacitor failure. No matter how
many processors an MSI board supports, I have to place the brand in the cheap
category.

Capacitor failure of a motherboard in an IBM server is a real surprise. The
board was an IBM-branded board? Unfortunately, IBM has also become as
susceptible as other name brands to the use of cheap motherboards and other
components in many of its systems.

Err, should it really be a surprise given that MSI actually
manufactures a lot of the motherboards for IBM servers? Actually
they make the entire server for a many of them, though I would guess
that IBM does a lot of the design work.

At the very least the IBM e325 is identical to the MSI K1-1000, and
IBM's e335 is basically the same as well except with Xeon processors
instead of Opterons.

 

[Bill Wolff]

Sounds like we are talking about electrolytic capacitors! Well I
have news for you! These electrolytic capacitors has been this way
long before computers ever came about.

As they are one of the most expensive capacitors one can buy
(although they can store huge amounts is what makes their cost
reasonable). And age really effects them greatly. As I seem to
recall from my engineering experience that they are generally good
for about 5 to 10 years and they evidently dry out and become
worthless. If you fire one up when they are bone dry (or near), they
can indeed explode!

I don't know anything about cheap vs. high quality electrolytic
capacitors, but I do know there isn't one on the market that can
last forever. As they all die in a matter of time. It is just their
nature and current technology can't change this. And currently, I
don't know what we can replace them with that would be cost
effective. Although some electrolytic capacitors can indeed last far
longer than 10 years. Although the following may explain this.

Now I do know that sometimes that engineers will push the voltage to
the limit. As electrolytic capacitors have a voltage limit which is
declared on the electrolytic capacitor. Pass this limit and the
capacitor will have a short life! Maybe this is what is happening.
As the higher voltage they are capable of, increases the cost of
them tremendously (and thus why they are costly).

Well this is just my 2 cents on the matter.



Bill






Johnny,

Yes, this is the well-known capacitor problem that plagues mostly the cheap
brands of motherboards manufactured out there somewhere on the Pacific Rim. In
some instances, capacitors actually exploded with a loud popping sound.

.... Ben Myers

 

[CJT]

Sounds like we are talking about electrolytic capacitors! Well I
have news for you! These electrolytic capacitors has been this way
long before computers ever came about.

As they are one of the most expensive capacitors one can buy
(although they can store huge amounts is what makes their cost
reasonable). And age really effects them greatly. As I seem to
recall from my engineering experience that they are generally good
for about 5 to 10 years and they evidently dry out and become
worthless. If you fire one up when they are bone dry (or near), they
can indeed explode!

I don't know anything about cheap vs. high quality electrolytic
capacitors, but I do know there isn't one on the market that can
last forever. As they all die in a matter of time. It is just their
nature and current technology can't change this. And currently, I
don't know what we can replace them with that would be cost
effective. Although some electrolytic capacitors can indeed last far
longer than 10 years. Although the following may explain this.

Now I do know that sometimes that engineers will push the voltage to
the limit. As electrolytic capacitors have a voltage limit which is
declared on the electrolytic capacitor. Pass this limit and the
capacitor will have a short life! Maybe this is what is happening.
As the higher voltage they are capable of, increases the cost of
them tremendously (and thus why they are costly).

Well this is just my 2 cents on the matter.



Bill






Johnny,

Yes, this is the well-known capacitor problem that plagues mostly the cheap
brands of motherboards manufactured out there somewhere on the Pacific Rim. In
some instances, capacitors actually exploded with a loud popping sound.

... Ben Myers

Read this and see whether it affects your thinking:

http://www.spectrum.ieee.org/WEBONLY/resource/feb03/ncap.html

 

[Ben Myers]

There's not much design work with computers any more. Design work seems to be
confined to an arcane corner of industrial design, selecting the color of paint
to put on the metal and how to sculpt the plastic front bezel to give the
computer its unique and VERY SPECIAL name-brand look-and-feel.

Pity to hear that IBM uses MSI boards in its servers. Maybe IBM at least has
had the sense to specify use of high quality capacitors? ... Ben Myers

 

[Johnny Solar]

Pity to hear that IBM uses MSI boards in its servers. Maybe IBM at least has
had the sense to specify use of high quality capacitors? ... Ben Myers

Just to clarify on my previous posts, the IBM server had a VIA chipset
motherboard.
The MSI motherboard was placed in another computer, but this was not an IBM
branded computer.


Sorry if I was a bit unclear on that point.


JS

 

[Bill Wolff]

Yes it does. Thanks Charlie.

Bill

Sounds like we are talking about electrolytic capacitors! Well I
have news for you! These electrolytic capacitors has been this way
long before computers ever came about...

Read this and see whether it affects your thinking:

http://www.spectrum.ieee.org/WEBONLY/resource/feb03/ncap.html

 

[chrisv]

Now I do know that sometimes that engineers will push the voltage to
the limit. As electrolytic capacitors have a voltage limit which is
declared on the electrolytic capacitor. Pass this limit and the
capacitor will have a short life! Maybe this is what is happening.

Well, a designer would have to be quite the bumbling fool to expose
caps to higher-than-rated voltage. Essentially, that never happens.
A far more common life-shortening problem is that of a cap which,
because it is under-sized, is exposed to excessive charge/discharge
(ripple) currents, which heats it up and hastens the drying of the
electrolyte.

 

[George Macdonald]

Sounds like we are talking about electrolytic capacitors! Well I
have news for you! These electrolytic capacitors has been this way
long before computers ever came about.

As they are one of the most expensive capacitors one can buy
(although they can store huge amounts is what makes their cost
reasonable). And age really effects them greatly. As I seem to
recall from my engineering experience that they are generally good
for about 5 to 10 years and they evidently dry out and become
worthless. If you fire one up when they are bone dry (or near), they
can indeed explode!

I don't know anything about cheap vs. high quality electrolytic
capacitors, but I do know there isn't one on the market that can
last forever. As they all die in a matter of time. It is just their
nature and current technology can't change this. And currently, I
don't know what we can replace them with that would be cost
effective. Although some electrolytic capacitors can indeed last far
longer than 10 years. Although the following may explain this.

Now I do know that sometimes that engineers will push the voltage to
the limit. As electrolytic capacitors have a voltage limit which is
declared on the electrolytic capacitor. Pass this limit and the
capacitor will have a short life! Maybe this is what is happening.
As the higher voltage they are capable of, increases the cost of
them tremendously (and thus why they are costly).

Well this is just my 2 cents on the matter.

This is a specific problem related to industrial espionage, theft of trade
secrets and bypassing patents etc. Here's the IEEE's take on it:
http://www.spectrum.ieee.org/WEBONLY/resource/feb03/ncap.html

Rgds, George Macdonald

"Just because they're paranoid doesn't mean you're not psychotic" - Who, me??

 

[George Macdonald]

Johnny,

Yes, this is the well-known capacitor problem that plagues mostly the cheap
brands of motherboards manufactured out there somewhere on the Pacific Rim. In
some instances, capacitors actually exploded with a loud popping sound.

Not at all. It affects several quality mbrd mfrs. including Asus & MSI
etc. over a specific time period, who *do* live on the Pacific Rim. Those
are *not* "cheap brands". It sounds like you think all Pacific Rim mbrds
are cheap; so where would you expect to find mbrd mfrs which are not cheap?
Where do you think quality laptops come from?

Rgds, George Macdonald

"Just because they're paranoid doesn't mean you're not psychotic" - Who, me??

 

[RusH]

This problem was most common in motherboards produced around
1999-2002 i think, but offcourse this may happen on some various
brands today for all i know.

you say "capacitor problem", but i instantly hear Abit problem
Its been since BH6, seems to stoped at KT7A.

Pozdrawiam.

 

[CJT]

you say "capacitor problem", but i instantly hear Abit problem

That's a pretty limited world view.

 

[chandugite]

MSI Motherboard reboot without any reason

i have the MSI motherboard which reboot without any reason can
anybody tell me why ?????