E4500 super abnormal temps...

[asbshyam]

I have an old box with an Intel C2D E4500 processor. It was running fine for the past three years. Recently the computer shut down frequently with random BSODs which was due to dirty ram.

After that the computer would just shutdown randomly without warning. I checked the Temps, and it was around 71C at the BIOS and a fast 100C on full load. After 100C the temps shown were random like -25C. I just pulled the plug.

The cpu is on stock HSF. I removed, cleaned, reapplied the thermal paste and now get idling temps around 55C. Still when on 100% load the temps would rocket to around 97-98C. And the system still shuts down randomly (even under light load).

I have reseated the HSF properly and there are no issues with it. I have tried HWMonitor, Speedfan, CoreTemp varying versions and readings are still the same. The stock fan spins to a max of 3000 RPM, so I've ruled out the fan. What else could be the issue???

sys config: C2D E4500, Mercury PI945GCM, Nvidia 7100GS, 1GB + 1GB Hynix DDR2 667Mhz, 400W OEM PSU, 160GB Seagate HDD..

 

[asbshyam]

I just noticed that the HSF is a low height one, generally issued with 45nm processors, but this is a 65nm. Will it cause such high temp raises?? I did not build the sys, so I have no idea if there was a mixup!!

 

[a4rocks]

Can you remember if there is anyway the CPU coolers could have swapped? I am guessing the HSF isn't seating properly and hence the high temperatures. Just remove your Motherboard and check for an existing gap with your naked eye. If there is one, try alternating the positions and check where it fits correctly.

 

[king smp]

I just noticed that the HSF is a low height one, generally issued with 45nm processors, but this is a 65nm. Will it cause such high temp raises?? I did not build the sys, so I have no idea if there was a mixup!!

I have a E4500 also in a Dell Optiplex 745 and it is OCd to 2.93
Idle on coretemp - 30c
load on coretemp - 60c
ambient room temp - 20c
distance to tj max coretemp - 58c

I tell you that so you can compare

first use coretemp and in options advanced select option to show readings of distance to TJmax

since the cpu was running fine for past three years and if ambient room temps didnt drastically go up (over 80F)
then the stock HSF was fine

1) What was the problem with the ram and what did you do to fix?

2) How did you know that was the problem?

3) Is your current problem the same as before you fixed the ram?

Please answer questions with the number format
Thank You

 

[asbshyam]

Nope, the system was never opened once built. I reseated the cooler myself and it sits perfect. Seating is not the issue.

 

[asbshyam]

I have a E4500 also in a Dell Optiplex 745 and it is OCd to 2.93
Idle on coretemp - 30c
load on coretemp - 60c
ambient room temp - 20c
distance to tj max coretemp - 58c

I tell you that so you can compare

first use coretemp and in options advanced select option to show readings of distance to TJmax

since the cpu was running fine for past three years and if ambient room temps didnt drastically go up (over 80F)
then the stock HSF was fine

1) What was the problem with the ram and what did you do to fix?

2) How did you know that was the problem?

3) Is your current problem the same as before you fixed the ram?

Please answer questions with the number format
Thank You

1) The Ram had some sticky dust on the contacts. I just cleaned it.

2) When the system BSODed randomly, the OS was reinstalled. The system BSODed during and after the install. I unplugged the RAM and cleaned them. And then there were no BSODs.

3) The current problem does not involve BSODs. It shuts down randomly without any warning.

I also have others C2D systems idling at 40c and maxing around 60, which is not that high considering the ambient temp is around 32c.

Thanks

 

[king smp]

three possibilities

1) ram was in process of failing, you cleaned them and it worked ok but now
ram has gotten worse and causing shut downs
if possible to use/borrow ram to try out that would be good
if not then try just one stick

2) power supply could be going, if possible to use/borrow another power supply that would be good

3) possilble bad motherboard - look at the motherboard - pay special attention to capacitors (if you dont know what they look like I will post a link to image)
see if they are bulging or leaking yellow fluid

Also the tower is still running hot
try taking off side of case if possible and using a house fan to blow into tower
and see if problem still occurs while monitoring temps
I like latest Coretemp for the E4500
and using distance to tjmax
really whatever temp you have is not important
the distance to tjmax is the best judge of temperature conditions

 

[asbshyam]

three possibilities

1) ram was in process of failing, you cleaned them and it worked ok but now
ram has gotten worse and causing shut downs
if possible to use/borrow ram to try out that would be good
if not then try just one stick

2) power supply could be going, if possible to use/borrow another power supply that would be good

3) possilble bad motherboard - look at the motherboard - pay special attention to capacitors (if you dont know what they look like I will post a link to image)
see if they are bulging or leaking yellow fluid

Also the tower is still running hot
try taking off side of case if possible and using a house fan to blow into tower
and see if problem still occurs while monitoring temps
I like latest Coretemp for the E4500
and using distance to tjmax
really whatever temp you have is not important
the distance to tjmax is the best judge of temperature conditions

Thanks for the help

1) I will swap the RAM sticks and try it out. AFAIK, in most cases bad RAMs leave some kind of trail to follow, BSODs, corrupt screen and stuck systems, no display with box on, beep codes. But this system always boots up normally, never shuts down in the BIOS, when in OS under varying load shuts down instantly and takes some time to let us start it. It does not restart automatically. what do ya think???

2) I haven't checked the PSU yet. I will check that out.

3) I will also look for bad caps in the motherboard.

The case is already open on the side with empty front bays removed.

 

[king smp]

number one is making think PSU

ram is weird
if had a short could cause strange problems

also this might sound silly but replace the mobo battery with brand new one
wouldnt hurt
bad mobo batteries are known to cause strange problems also
just good to rule it out

also in bios reset to defaults at same time changing battery
again just to eliminate possible factors
wont hurt to do

 

[asbshyam]

number one is making think PSU

ram is weird
if had a short could cause strange problems

also this might sound silly but replace the mobo battery with brand new one
wouldnt hurt
bad mobo batteries are known to cause strange problems also
just good to rule it out

also in bios reset to defaults at same time changing battery
again just to eliminate possible factors
wont hurt to do

Will try those stuff and let you know wat happens... Thx

 

[king smp]

I will check back later today
Good Luck

 

[asbshyam]

The hsf surface is not completely level, only half of it makes contact with the cpu. Its not worth lapping. I seated the hsf in a different direction and now the cpu idles around 50c and max temp is around 75c. Thats good for now. The tower is running at temps acceptable by me.

But the system still shuts down. The caps on the board are clean. Nothing is bloated or leaking.

There are two stick of 1gb ram. I will be swapping them to find the bad one if there is one.

After the RAMs i will be swapping the psu with a different one too.

The CMOS battery reads a voltage of 3.08 in my multimeter. Its think its about right??? I will be swapping it too along with the psu after testing the RAMs.

Also I noticed that the Graphics card (7100GS) is rendering artifacts all over the screen while playing videos, its temperature is fine though. Could this card cause the system to shutdown????

 

[king smp]

absolutly the card could be causing this
should never have artifacts especially if not over clocking it
if you have onboard graphics the first thing you do is
remove the card and use the onboard graphics
if no onboard then replace with appropriate
new gpu
need to know if agp or pci express
in either case they make a HD4650 for both form factors
under $50 USD
I own one
Definitley not the best card out there
but does not need external power and draws low wattage
also much mcuh better than current card you have now
dont get the 1gb DDR2
get the 512 DDR3 if possible
has many newer features regarding video playback
and can handle newer games at lower settinga and resolutions
if that matters
even if the card doesnt turn out to be the problem
you are really overdue for new card
and the HD 4650 is cheap
I like the Sapphire model
due to cost and quality combined
Not sure where you live
if you live in US and want me to post link
to good deal then let me know
also just for the Nvidia fans out there
a GT240 would be a good option also
though not made in AGP

cmos battery voltage is perfect

i would not worry about rams or psu
at this point
though a inexpensive PSU upgrade would be a good idea
if the PSU in unit is old and/or no name

i am not a psu expert
but antec,corsair and seasonic are the best choices generally
I know that for a less expensive one for a lower powered system
that the Antec 380watt model is nice
if not using a wattage hungry card

 

[asbshyam]

The machine needs no gpu, igp would do. I will remove it and try it with the IGP.

And the power supply 12v reads 12.20v idle and 11.80 on load. So I think its okay for now.

 

[king smp]

psu readings good

cool I like somebody that has a purpose for their
tower and uses what is needed to accomplish the job

Let me know how it works

Even if it running perfect
please let me know
because I do care

 

[asbshyam]

Best answer selected by asbshyam.

 

[asbshyam]

You were right Thanks for the help the gpu was the culprit, the sys is booting fine on the intel IGP.... but there is another issue and an new thread http://www.tomshardware.com/forum/291735-30-access-problem

 

[king smp]

Ok I will check it out
thanks for the Best Answer