Shutting down under load

cody0851

Member
Dec 2, 2012
35
0
0
Can't remember my old account login info:/

Recently, 400watt psu died. I think. Computer would not turn on. Replaced with this 850w psu, and computer powers fine.

Specs
not overclocked
ASUS P5NSLI
C2D E6300 w stock fan
dvdrw drive
blu ray rw drive
internal card reader
dual tv tuner
MSI 7600gs passively cooled/no fan
firewire pci card
4gb ddr2 800 ram
WD 250GB HD
120mm rear fan blowing out. no fans blowing in.

Idle specs-

idlee.jpg


The shut down problem started when I was watching TV via media center, watching a web video, surfing, and had both monitors running. Something I've done off/on for years.I first thought it was my video card overheating. Rebooted, cranked everything up, and monitored the temp. It hit 73c and shut down..
Then I noticed my cpu temp and looked into it. I downloaded and ran Intelburntest. 10 seconds later, my cpu hit 74c and shut down.
I opened my case, blew everything dust free, made sure all fans were running. Setup a small house fan blowing directly on cpu. Ran IBT again with case open. It hit 71c and shut down.

Ran 3dMark05 with case closed: computer shutdown during cpu test.
Ran 3dMark05 with case open: Completed all tests. Vid card hit 75c, cpu 71c.
My only thought so far, is adding more case fans will not lower my temps.

So is temperature is my problem? Is it possibly my psu is not sending enough power to the cpu? Is something just getting too old?(6 year old build.)
 

AnonymouseUser

Diamond Member
May 14, 2003
9,943
107
106
It's possible the old PSU damaged the motherboard when it died, so if possible, check the PSU in another PC to rule the PSU out.

Also, check your motherboard for bulging capacitors.
 

cody0851

Member
Dec 2, 2012
35
0
0
Don't have a second computer to check:/

I see no bulging or leaking capacitors. They all look very clean and shiny.
 

Smoove910

Golden Member
Aug 2, 2006
1,236
6
81
is there something in your BIOS to shut the computer down once it reaches a certain temp?
 

cody0851

Member
Dec 2, 2012
35
0
0
An hour ago, it shut down and all I was doing was light web browsing.

I shut down and restarted into the BIOS.
CPU INTERNAL THERMAL CONTROL: AUTO
Q-Fan was disabled. I enabled it. Don't know what it means.
My BIOS listed these temps:
CPU 52C
M/B 48C

There is nothing about a temp ceiling.
 

AnonymouseUser

Diamond Member
May 14, 2003
9,943
107
106
Time to remove all but absolutely necessary to run the PC. Remove or disconnect the optical drives, card reader, tv tuner, and firewire card. Boot up, stress test, and see if it shuts down. If it lasts longer than usual, add one item back at a time until it shuts down again.
 

cody0851

Member
Dec 2, 2012
35
0
0
Time to remove all but absolutely necessary to run the PC. Remove or disconnect the optical drives, card reader, tv tuner, and firewire card. Boot up, stress test, and see if it shuts down. If it lasts longer than usual, add one item back at a time until it shuts down again.


Does that mean you're ruling out temperature?
 

Steltek

Diamond Member
Mar 29, 2001
3,034
748
136
An hour ago, it shut down and all I was doing was light web browsing.

I shut down and restarted into the BIOS.
CPU INTERNAL THERMAL CONTROL: AUTO
Q-Fan was disabled. I enabled it. Don't know what it means.
My BIOS listed these temps:
CPU 52C
M/B 48C

There is nothing about a temp ceiling.

Yeah, that motherboard is old enough that it while it does very basic temperature and voltage monitoring I don't think that the BIOS has anything beyond minimal protection functionality. All Q-Fan does is allow the system to automatically adjust fan speeds to try to keep the noise levels down. It being disabled wouldn't cause this problem.

Just to be certain, when you say it dies you do mean that it goes totally dead (i.e. all fans - CPU and case - stop and lights/LEDs turn off, etc) and not that the screen just goes blank?

Does the system shut down if you boot from a Linux live CD and browse the internet?

It might be worth trying to run a couple of MEMTEST86+ passes on each memory module to ensure that nothing is wrong with your memory. Also, re-seat your memory modules and all power connectors.
 

cody0851

Member
Dec 2, 2012
35
0
0
I just looked over my motherboard and made sure everything was plugged in well.
My card reader was dangling off the usb header. I plugged that in, and stress tested. No change.
For giggles, I ran safe mode and stress tested. It lasted 15 seconds before dying.
By dying, I mean the whole computer powers off, immediately.
Just further fyi, my computer, monitor, router, and cable modem are plugged into a year old 1100w battery backup.

Will look into memtest now. I haven't tried a live Linux cd.
 
Last edited:

cody0851

Member
Dec 2, 2012
35
0
0
No blue screen. Just straight dead like a power outage.

Just tried to burn memtest to cd.. My drive has power, but isn't communicating with the m/b. Examined the sata port, and the whole plugs falls off the m/b.

Well, I guess I'll use the blu ray burner. lol

Also, noticed my computers clock is 5 minutes slow. First time it has been wrong in ages.
 
Last edited:

Ketchup

Elite Member
Sep 1, 2002
14,545
236
106
Sounds like the old ps fried some things on the mobo. Had that happen to a customer once.
 

cody0851

Member
Dec 2, 2012
35
0
0
Ran 1 pass of memtest. No errors.

Tomorrow morning, I'm going to put in a different psu, just to try to rule the new one out. If this m/b is indeed shot, I'll be forced to make a decision. Just buy a massive laptop, and go without a desktop - or build a massive desktop and go for a cheap laptop with an ssd. Been contemplating that for at least a year, while trying to squeeze the life out of my current desktop and laptop.
 

Steltek

Diamond Member
Mar 29, 2001
3,034
748
136
Just to make sure your UPS isn't dying, does it shut down the same way if you bypass the UPS and plug straight into the wall? It also wouldn't hurt to pull all of your add-in cards except the video card to totally eliminate the chance that one of the cards is failing and tripping the power supply.

If neither of these things pan out, I really think you'll need to be in the market for a new motherboard. It is still possible to get a relatively cheap Socket 775 board for $40-$60 (ASUS, Gigabyte, and ECS), though almost all of them are micro-ATX (i.e. make sure to get one with enough PCI slots for your add-in cards) and DDR3 (4GB of DDR3 is pretty dirt cheap nowadays).
 
Last edited:

jolancer

Senior member
Sep 6, 2004
469
0
0
I thought it a good idea to always take software temp monitoring with a grain of salt. Also just fyi, i had a bios battery die once, and a no power up was the symptom, i dont know what other symptoms if any a low or dead bios battery may cause.

I'd plug a periferal such as a fan or watever into your old 400w psu, and test that to see if its actually dead by jumping the Green/Black pins on the MB power connetor, and see if it powers up.

if you old psu isn't dead, id try a new battery for bios

if old psu is dead, i'd replace thermal compound on your CPU / Northbridge chip. with AS5 or whatever is used these days, It use to be cheap. easy fix if temp is the issue, as long as the new compound is properly applied. ArcticSilver5s website use to have really good instructions.

I only say northbridge also because i had boards with fanless northbridges to ofcourse and some i notice come with some flacky compound or pad... but would get feel dam hot for being fanless ( i personlly just mod'd a fan on mine set to lower rpm along with new compound)... as long as its in close enough contact(like right on the fins) because the windchill of the fan blades is what primarily drops the temp.

EDIT: oh yeah dont know what your stock HSF looks like, but if its anything like those old P4 HSF's, and you did *Not stick the nozzle to the pray can directly threw the fins onto the hsf.. may help to remove the Fan and wipe the top of the HSF off were the fins are spreading, you'd be saprized how much shit can be hidden under that fan... a thin unseen but even layer of dust even can show an increase in temp just by the area its spread over.
 
Last edited:

cody0851

Member
Dec 2, 2012
35
0
0
I tried the green/black trick with a fan, 2 optical drives, and my hd still attached. The optical drives lights blinked and then nothing. Pretty sure if it wasn't dead, it is now.

I definitely got under the fan blades. I sprayed the canned air while the comp was running just to see if I could drop the temps. lol.. and no, it didn't work. I'll look again. I may pull the fan off just out of curiosity.

This m/b was known to have a hot fanless northbridge when it came out. As old as it is, I'm scared to start wiggling/pulling on things. It might just break in my hands. 2 SATA ports have already fallen off:/

On my list tomorrow:
1. Bypass battery backup
2. Try different psu
3. Unplug cards
 

jolancer

Senior member
Sep 6, 2004
469
0
0
Jumping the PSU when its not attached to the MB is different then jumping the MB's power switch in that, you have to leave the paper clip jumped between the green/black, if you remove it it powers off. is that what you did?

Dont worry about messing with the components, unless you mess everythying up like gorrilla grip everything. only thing htat changes over time is the thermal compound under the heatsinks, so if you mess with them just clean and replace the compound. an make sure your prepared2 befor messing with them.
 

Torn Mind

Lifer
Nov 25, 2012
11,567
2,626
136
EDIT: oh yeah dont know what your stock HSF looks like, but if its anything like those old P4 HSF's, and you did *Not stick the nozzle to the pray can directly threw the fins onto the hsf.. may help to remove the Fan and wipe the top of the HSF off were the fins are spreading, you'd be saprized how much shit can be hidden under that fan... a thin unseen but even layer of dust even can show an increase in temp just by the area its spread over.
Ha. the Socket 478 heatsink I have accumulated such a thick layer of dust that was more like a thick pad. I had to take off the heatsink, remove the fan, and pull off the dust with toilet paper.
 

daveybrat

Elite Member
Super Moderator
Jan 31, 2000
5,725
943
126
An hour ago, it shut down and all I was doing was light web browsing.

I shut down and restarted into the BIOS.
CPU INTERNAL THERMAL CONTROL: AUTO
Q-Fan was disabled. I enabled it. Don't know what it means.
My BIOS listed these temps:
CPU 52C
M/B 48C

There is nothing about a temp ceiling.

What are you all talking about?

Seriously, 52C idling in the bios for this cpu?? Way too hot!

Your E6300 is only a Core2Duo 1.86GHz 65Watt cpu. There is no way it should be putting off that much heat while idle in the bios.

And the shutdowns you describe sound exactly like cpu built-in thermal protection.

Get yourself a new heatsink/fan or take yours off, clean the old grease off and re-grease it properly.
 

Eureka

Diamond Member
Sep 6, 2005
3,822
1
81
It could be CPU overheat. Thermal paste will degrade over time, and it's not a bad idea to pull off the HSF to reapply.

That being said, I've never had a direct shutoff from thermal protection. It'll usually throttle first before anything else, with a very noticeable computer stutter. Won't hurt to reply though.
 

cody0851

Member
Dec 2, 2012
35
0
0
Ok, I am thoroughly perplexed. I plugged my computer straight to wall socket. I ran IBT, and even though CPU temp hit 80c, it didn't die. I figured aha! The backup is the problem.

I plugged it back into the battery backup. Ran IBT, temp hit 80c. It won't die.

My laptop used to shutdown exactly like this when it got too hot. I blew out the vents. Problem solved. So yes, everything points me to temp being the problem. Except why is it running just fine under stress at 80c today, and died at 71c yesterday?

As I type this, without load, CPUID says my cores are 66c and 71c.
 

Ketchup

Elite Member
Sep 1, 2002
14,545
236
106
Ok, I am thoroughly perplexed. I plugged my computer straight to wall socket. I ran IBT, and even though CPU temp hit 80c, it didn't die. I figured aha! The backup is the problem.

I plugged it back into the battery backup. Ran IBT, temp hit 80c. It won't die.

My laptop used to shutdown exactly like this when it got too hot. I blew out the vents. Problem solved. So yes, everything points me to temp being the problem. Except why is it running just fine under stress at 80c today, and died at 71c yesterday?

As I type this, without load, CPUID says my cores are 66c and 71c.

Many things can cause a CPU to let you down like that, Mem, Mobo, PS, heat. I would start by re-applying thermal material to the CPU.

I would also keep it off the battery back up for a little while (day or two). Plugging it back in after an extended period will do a better job of telling you what the culprit it.
 

cody0851

Member
Dec 2, 2012
35
0
0
Ahh. Killed it after I typed my last message. CPUID showed 74c when it shut off.

Maybe everything wasn't properly warmed up yet, as I started all this after it sat off all night. Now that I've killed it, I'm going to start ruling out my batt backup, and psu.
 

cody0851

Member
Dec 2, 2012
35
0
0
Ok.. Ruled out my batt backup, by plugging into wall. It died during IBT.
Plugged back into batt backup. Unplugged both optical drives from psu. Ran IBT and it went through all 10 tests. CPUID showed temp between 76 and 83c during test.
Plugged blu ray burner back in. Ran IBT, and same thing - all 10 tests passed, temp between 76 and 83c.
Left dvd burner unplugged. It's not plugged into m/b anyway.
Turned on my tv. It's plugged into the same outlet as my batt backup. Ran IBT. Computer died during 1st or 2nd test.
Turned off tv. IBT finished all 10 tests. Temp reached 85c consistently.
Turned on TV. Computer died during 7th or 8th test. Temp stayed 83c or lower.


Honestly, I need a break. I really appreciate all the advice from everyone. This thing is really eating up all of our time.