CPU Crashes Cold, But Works Great Warm??? Help?

[Billy Saul Hearok]

I'm having a weird problem with my system that started this morning.

The computer wouldn't boot into WinXP. It kept either hanging or
restarting. So, I went in and lowered all the settings in the BIOS.
Still no better.

I popped out one of my memory chips, but it still wouldn't make it
through the boot. So, I put it back in and popped out the other. This
time it wouldn't boot at all(black screen, no post beep). So I removed
that one and went on without it. But I was STILL getting crashes at
boot.

I slowly began to realize that each time I booted, I was getting a
little further in the boot sequence. So, after one crash, I just left
the PC on and let it "warm up". Then, I restarted and WALLA! it worked
fine. In fact, I'm typing this message on it...

--So, WHAT can cause a PC to hang/restart when cold, but work just
fine when warmed up?-- I've heard of the reverse situation, but never
this...

Thanks for any help.

 

[Chris Simpson]

I have how cold does the room get when the PC is off? If the room gets
below 40 F then you need to keepm the PC on all of the time...

 

[Billy Saul Hearok]

40 F? I don't live in an igloo ;-) The room temperature stays around
70 degrees or so...

 

[Jan Alter]

OK,
I'm gonna guess the PS here. Maybe it's got leaking caps and there
isn't enough charge on them to supply enough current to the mb. However,
each time you re-start the system a little more current is retained in them
to supply the system.

If you have another PS try substituting it and send me the $ if I'm right.

 

[Arno Wagner]

I'm having a weird problem with my system that started this morning.

The computer wouldn't boot into WinXP. It kept either hanging or
restarting. So, I went in and lowered all the settings in the BIOS.
Still no better.

I popped out one of my memory chips, but it still wouldn't make it
through the boot. So, I put it back in and popped out the other. This
time it wouldn't boot at all(black screen, no post beep). So I removed
that one and went on without it. But I was STILL getting crashes at
boot.

I slowly began to realize that each time I booted, I was getting a
little further in the boot sequence. So, after one crash, I just left
the PC on and let it "warm up". Then, I restarted and WALLA! it worked
fine. In fact, I'm typing this message on it...

--So, WHAT can cause a PC to hang/restart when cold, but work just
fine when warmed up?-- I've heard of the reverse situation, but never
this...

Could be a electrolyte capacitor gone bad in a sensitive place. These
things can have a large leak-current when the have been off for a
wile. Due to the nature of their construction, the isolation layer
slowly rebuilds when they carry a charge, which decreases the leakage
current. This is counter-intuitive, but happens from time to time in
in practice. Note that this is not the same as the "exploding
capacitor"-problem.

My guess would be that the problem is in the PSU or the reset-
circuit on the mainboard.

Was this PC off for some days?

Arno

 

[Billy Saul Hearok]

Thanks for your advice, Jan. I was thinking along the same lines(PS),
but since I don't have another one to test, I wanted to get a few
"second opinions" before I shell out for one

 

[Billy Saul Hearok]

The PC was off for about 5 hours.

So, do you think the PSU is the likely culprit?

My PSU is some cheap, no-name. When I was fiddling in the bios, I
noticed that the vcore for my Barton is being reported as varying
between 1.585 and 1.60. I bumped it up by 1% so that it would be
closer to the Barton default of 1.65. Vcore is now reported as varying
between 1.60 and 1.618 or so.

 

[kony]

The PC was off for about 5 hours.

So, do you think the PSU is the likely culprit?

My PSU is some cheap, no-name. When I was fiddling in the bios, I
noticed that the vcore for my Barton is being reported as varying
between 1.585 and 1.60. I bumped it up by 1% so that it would be
closer to the Barton default of 1.65. Vcore is now reported as varying
between 1.60 and 1.618 or so.

I was under the impression that the system would have to remain off for
MUCH longer amount of time before the cap degraded, and after power on
attempts enough that it was running, a new oxide layer had then formed
(implied since it then runs), and only leaving the system off for another
LONG time would result in same scenario recurring... not if system were
only off for a few hours or days.

You should never have ran the system with a "cheap, no-name" PSU. It's
likely the power supply is failing or else the motherboard. If you have a
multimeter it wouldn't hurt to take voltage readings at the ATX connector
while system is first cold->powered-on. However failed caps in the power
supply might still be roughly on-spec by a multimeter even after capacitor
degradation making the PSU unfit for use. Considering the "cheap,
no-name" issue, the power supply should be replaced. Question is then if
you replaced it soon enough or if motherboard has been degraded as well...
could require replacement of both for long-term viability but the PSU is
the first focus.

 

[~misfit~]

OK,
I'm gonna guess the PS here. Maybe it's got leaking caps and
there isn't enough charge on them to supply enough current to the mb.
However, each time you re-start the system a little more current is
retained in them to supply the system.

If you have another PS try substituting it and send me the $ if I'm
right.

Maybe in the days of AT PSUs. Modern PSUs (ATX) are powered up all the time
(unless unplugged).[/QUOTE]

 

[Trent©]

I'm having a weird problem with my system that started this morning.

The computer wouldn't boot into WinXP. It kept either hanging or
restarting. So, I went in and lowered all the settings in the BIOS.
Still no better.

I popped out one of my memory chips, but it still wouldn't make it
through the boot. So, I put it back in and popped out the other. This
time it wouldn't boot at all(black screen, no post beep). So I removed
that one and went on without it. But I was STILL getting crashes at
boot.

I slowly began to realize that each time I booted, I was getting a
little further in the boot sequence. So, after one crash, I just left
the PC on and let it "warm up". Then, I restarted and WALLA! it worked
fine. In fact, I'm typing this message on it...

--So, WHAT can cause a PC to hang/restart when cold, but work just
fine when warmed up?-- I've heard of the reverse situation, but never
this...

Thanks for any help.

Yer usin' false logic. You picked the solution...a warmer CPU...and
not yer pickin' the logic to justify the solution.

Did you realize that you've rotated inches further (or closer) from
the moon each time you reboot. Maybe its the intensity of the
gradational pull that is causing the problem.

Every time you boot, xp is doin' things in the background...that you
can't see. More than likely, it was readjusting itself. It could be
offline synchronization, etc...or a myriad of things.

Let it go for a week or so. Give it at least 30 minutes each time to
boot up...just to be on the safe side. If it takes near the same
amount of time to boot each time, download some software that can tell
you what's loading...and how long each process is taking.


Have a nice week...

Trent

What do you call a smart blonde?
A golden retriever.

 

[~misfit~]

I was under the impression that the system would have to remain off
for MUCH longer amount of time before the cap degraded, and after
power on attempts enough that it was running, a new oxide layer had
then formed (implied since it then runs), and only leaving the system
off for another LONG time would result in same scenario recurring...
not if system were only off for a few hours or days.

You should never have ran the system with a "cheap, no-name" PSU.
It's likely the power supply is failing or else the motherboard.

I'm glad you said that or that would have made me the only one thinking it's
the mobo. As it's running a Barton it shouldn't be old enough to be
suffering from the infamous widespread capacitor problem. However I'd still
do a visual inspection of all the capacitors on the board, check for any
signs of bulging (sides or top) and/or leakage.

 

[half_pint]

I'm having a weird problem with my system that started this morning.

The computer wouldn't boot into WinXP. It kept either hanging or
restarting. So, I went in and lowered all the settings in the BIOS.
Still no better.

I popped out one of my memory chips, but it still wouldn't make it
through the boot. So, I put it back in and popped out the other. This
time it wouldn't boot at all(black screen, no post beep). So I removed
that one and went on without it. But I was STILL getting crashes at
boot.

I slowly began to realize that each time I booted, I was getting a
little further in the boot sequence. So, after one crash, I just left
the PC on and let it "warm up". Then, I restarted and WALLA! it worked
fine. In fact, I'm typing this message on it...

--So, WHAT can cause a PC to hang/restart when cold, but work just
fine when warmed up?-- I've heard of the reverse situation, but never
this...

Thanks for any help.

I had a problem very similar to that recently, the good new is it has gone
now,
tha bad news if I don't know what for sure what cured it.
I also thought power supply capacitors etc..
How far does it get and how many restarts are needed?
What do you mean by 'warm up'? doesn't it shutdown?
Could also be dodgy power switch perhaps?
Ever ran Ares file share software? An problems with OE?
Do you have a lot of old newsgroup message? (size of store in megabytes)?

Anyway I tried a few things but basically I am running the same
set-up now, same hardware and software, and the problem
has disappeared,
My problem was not consistant though, sometimes it ran ok,
it was more occasionally for me that there was a problem.

Try a few things, disconnect your harddrives and boot say
from a floppy with memtest on it. Anything to change the
set-up basically to see if it makes a diference.
An open (or closed) curcuit somewhere might also cause a problem
like this.(especially if its a 'hairline' crack.

Anyway my computer is working like a dream at the moment, one
thing which has changed is the weather, its warmer!!!!
Maybe wait untill summer???!!!!

Maybe get your computer up and running and then stiick the
main box in the refridgerator! Its a long shot but it might work....

Anyway you can comfort eat all the food you took out whilst
you wait to see what happens (plus drink the beer!).

half_pint.

 

[Trent©]

You should never have ran the system with a "cheap, no-name" PSU.

I always chuckle to myself when I read this kinda comment.

AMD, Microsoft, Compaq, Dell, ...help me out here. Were these all
'famous' names when they started out?

99% of the folks here who comment about power supplies have never
tested one themselves. All they do is re-comment about something that
they know very little about.

I'm not suggesting that you fall into this category, kony. But it
just irks me that almost all ailments nowadays are blamed on the power
supply.

More often than not, cables (of all kinds) are the bigger problem.

The only PSU problems I've encountered lately is having a
way-underpowered supply for the P4's and comparable AMD's.


Have a nice week...

Trent

What do you call a smart blonde?
A golden retriever.

 

[Billy Saul Hearok]

Hi guys, here's an update. After I posted, I turned the CPU off to
test and could not get it to boot fully into XP again, despite
"warming up".

So, I decided pull the power supply out of my other system to test(no
small task since the CDROM and the CPU heatsink had to be removed to
pull it). I popped it in and (drumroll please) it booted just fine. So
it looks like a PSU problem as many of you suggested. Thanks

BTW: As for the folks who chastised me for using a cheap PSU... Let me
just say that this PSU had been working great for 5 years and had kept
my system perfectly stable through two system upgrades. So, let's not
completely dismiss the "bang for the buck" here ;-) Who knows if a
$100 power supply would have lasted any longer...

MANY thanks to everyone for your help, even if some of it included a
little criticism ;-). I appreciate it.

 

[half_pint]

Nice to know you fixed it, it certaintly had the characteristics of a
capacitor problem, amongst perhaps, other things. Maybe you could
open the PSU up to get a look inside, but be very careful as the voltages
can be dangerous so if in any doubt whatsoever leave it well alone.
Its actually sounds very similar to a problem posted by 'putim'
in an earlier thread 'Intermittant PC lock' however his solution was a new
motherboard and ram.
Maybe he had dodgy capacitors too?

 

[Chris Simpson]

You never know....

40 F? I don't live in an igloo ;-) The room temperature stays around
70 degrees or so...

 

[kony]

I always chuckle to myself when I read this kinda comment.

To a certain extent you're right, but due to the very name of a _generic_
power supply, there's only an assumption of what's inside that metal case,
a gamble with the only way to win being the savings of a small portion of
total system cost. The irony here is that this generic power supply was
the reason for the problem, or at least it's insufficient capacity, which
is too often my gripe with the generics, that you can't trust the labeled
capacity.

AMD, Microsoft, Compaq, Dell, ...help me out here. Were these all
'famous' names when they started out?

The power supply industry is even more mature than the *Wintel* PC
industry though, how much chance do you give a young upstart to compete
with Dell or Microsoft in their core markets?

99% of the folks here who comment about power supplies have never
tested one themselves. All they do is re-comment about something that
they know very little about.

Systems are using an ever-increasing amount of power, yet the average
generic power supply is still built around similar design to that used
several years ago and labeled as 250W. Historically we can see some of
these generics ran a 486 system fine, but the 486 might use 1/5 the power
of a decent modern system while the generic power supply has not increased
in true capacity by 5X... even ignoring the fact that a startling number
of them don't even have basic noise filtration or surge protection.

I'm not suggesting that you fall into this category, kony. But it
just irks me that almost all ailments nowadays are blamed on the power
supply.

Cetainly it can be too quickly suggested as the problem, but it seems to
be more common these days, one of the most common problems outside of a
motherboard misconfiguration, damage or defect.

More often than not, cables (of all kinds) are the bigger problem.

Could be, but I suppose at a certain point I start assuming the user has
at least opened the case and looked around, wiggled a few cables and such,
we have to start somewhere in the whole troubleshooting process.

The only PSU problems I've encountered lately is having a
way-underpowered supply for the P4's and comparable AMD's.

Wouldn't that be most likely to occur with a generic power supply though?
Even a fancy-looking LED neon clear fan sheathed gold anodized aluminum
gee-whiz-bang $75 special may actually be a POS inside, relative to it's
wattage, construction relative to a major manufacturer's unit of same
rating. To me this misstating of capacity goes against industry standards
and should be illegal if it isn't already, which it very well may be, if
not now then in the future.

Certainly a generic power supply can work fine, but generally they're not
worth more than 300W, and a name-brand capable of 300W can be purchased
for $30... the gamble doesn't even have a good payoff if one "wins".

 

[Billy Saul Hearok]

total system cost. The irony here is that this generic power supply was
the reason for the problem, or at least it's insufficient capacity, which
is too often my gripe with the generics, that you can't trust the labeled
capacity.

Just to clarify: The problem with my PS wasn't "insufficient
capacity". This system had been operating just fine with this power
supply for 8 months now.

The problem was/is that the power supply just "crapped out". Of
course, it could be argued that it died because of bad build quality,
but as I mentioned in another post, the power supply worked great for
5 years and two system upgrades. I'm not sure you could expect
anything more from a higher priced unit...

 

[kony]

Just to clarify: The problem with my PS wasn't "insufficient
capacity". This system had been operating just fine with this power
supply for 8 months now.

That's just it, insufficient capacity can and does affect lifespan. Take
any power supply and try running it beyond an amperage it can sustain and
you will reduce it's lifespan... not just a matter of whether it turns on
the first day or for a few months. Likely if you'd kept that PSU in the
original or 2nd system build, it'd still be working today barring some
random mishap like a significant AC line surge.

The problem was/is that the power supply just "crapped out". Of
course, it could be argued that it died because of bad build quality,
but as I mentioned in another post, the power supply worked great for
5 years and two system upgrades. I'm not sure you could expect
anything more from a higher priced unit...

"You're not sure". Ignoring random failures and looking at the odds, I am
sure that a quality name-brand unit of ample capacity will last over 5
years. It's not even necessary to contrast a generic with a $100 unit,
even many $50 units can be expected to support the amperage on their label
and have a serviceable lifespan of roughly a decade. I have stacks of
generics that pile up and get thrown out, they're not even worth
repairing. The stack of name-brand units is non-existant, they get reused
and are still working, barring a few random failures at nowhere near the
rate of the generics.

Now don't get me wrong, there are some generics that aren't terrible.
Then there's the issue of what we call a generic... If a large name-brand
company produces a unit that's relabeled, the user may or may not realize
this, may or may not consider anything less than a very popular brand to
be a generic. Your gamble paid off, but others lose this gamble... system
reliability should not BE a gamble.

The bottom line is that power supply manufacturers can charge whatever the
market will bear, but usually these generics have design, construction
differences. No company is going to throw away money if there isn't some
benefit, purpose behind that. You are free to buy whatever you like but
your random speculation that a "better" power supply wouldn't do any
"better", is in itself quite illogical. If you deny that one of the top
name-brands is better than a generic then open a few up, compare the build
and do some testing.

 

[Billy Saul Hearok]

benefit, purpose behind that. You are free to buy whatever you like but
your random speculation that a "better" power supply wouldn't do any
"better", is in itself quite illogical.

Excuse me Spock, but my "random speculation" was that a HIGHER PRICED
power supply might not do any better. I'm pretty sure that correcting
MISQUOTED logic breaks the Vulcan code

Just kidding
In all seriousness, I am planning to go with a name brand this time.
So any recommendations on the lower priced name brands(Allied,
Antec,AGPB)?