PC powers off when CPU is not being taxed...CPU/Voltage issue? not virus...

[dkolacz]

I am a PC tech, but not to the mobo component level, so I do not know if this problem is easily fixable.

I have an older laptop from 2008 that has been running fine up until a couple months ago. I am very fond of this model and would love to salvage it if possible.

In a nutshell, the laptop will just power off (no bluescreen, no errors) if the CPU is not getting taxed.
When windows comes up, I have to immediately start a 720P video or run a game in the background and then it will stay powered on. I can then browse the web, or do other things. If I close down all programs, the laptop will power off within seconds/minutes.

My troubleshooting steps so far:

It is not overheating as HW Monitor reports normal and I have re-pasted everything.
Swapped out RAM, and hard drive.
It happens when just the battery is supplying power.
It happens when just the power cord supplying power.
It happens when both the battery and power cord are used.

This is not an OS problem as when I swapped the hard drive, I tried reinstalling Windows 7 and it powered off during the initial setup phases in the partitioning section. I have tried Live Boot Linux distros and it shuts off in them as well.

Now, some strangeness...If I boot into BIOS, MemTest, DOS, or the Hirens/Falcon4 CD and boot into mini Windows XP, it will stay up forever. I am guessing this is because these environments don't support different power states of the CPU? I don't know if this is correct, but I am at a loss.

I'm pretty sure this post would be better off in another forum area, but did not know which to select. If the mods would like to move it, please do!

Anyone have any ideas?

 

[mindless1]

See if there are any power state settings in the BIOS that you could toggle. If not, see if there is such a software utility for your CPU.

Ultimately it seems likely that if you can't or don't want to do these things then it needs a new mainboard, but frankly no matter how much you like it, it has achieved it's expected lifespan and then some. Might be time to bite the bullet and get a replacement laptop rather than mainboard, and eventually fan, battery, hard drive... though I suppose by now you might have already replaced some of those items.

 

[Paperdoc]

Are you sure it's not overheating?

My first reaction to your story was this: OK, the machine is old enough that the main cooling fan is wearing out its bearings so that the fan runs a little slower than it should. At high processing loads it can still keep up. But drop the processing load and the CPU cools down, so the fan is slowed down, too. That is, the mobo reduces the fan voltage to a low level. Maybe that's low enough that the fan, which is struggling, actually stalls and then cannot re-start easily. If that happened, two possibilities ensue:

1. Certainly before long the CPU would overheat, and that would happen over less than a minute, I think, so you MIGHT not notice it. Then the machine would shut down for that reason.
2. On SOME mobos (I don't know about your laptop) there is an extra CPU protection system. It monitors the CPU cooling fan (maybe that's your only fan in the laptop) and, if it detects that the fan has stopped, it does not wait for the CPU temperature sensor to alert an overheating situation. It just beeps a brief warning and shuts down everything to prevent the possibility of overheating with no cooling fan.

Oh yes, I thought of another cause of reduced cooling fan performance in laptops, mainly because I saw this exact thing happen in my grandson's laptop. His machine managed to build up a collection of dust and lint that got trapped in the cooling air ductwork near the fan. You could not even see it until the machine was disassembled enough to open up that dict and see all of it. The wad of dust was rubbing the fan rotor and slowing it down badly.

 

[THRiLL KiLL]

Are you sure it's not overheating?

My first reaction to your story was this: OK, the machine is old enough that the main cooling fan is wearing out its bearings so that the fan runs a little slower than it should. At high processing loads it can still keep up. But drop the processing load and the CPU cools down, so the fan is slowed down, too. That is, the mobo reduces the fan voltage to a low level. Maybe that's low enough that the fan, which is struggling, actually stalls and then cannot re-start easily. If that happened, two possibilities ensue:

1. Certainly before long the CPU would overheat, and that would happen over less than a minute, I think, so you MIGHT not notice it. Then the machine would shut down for that reason.
2. On SOME mobos (I don't know about your laptop) there is an extra CPU protection system. It monitors the CPU cooling fan (maybe that's your only fan in the laptop) and, if it detects that the fan has stopped, it does not wait for the CPU temperature sensor to alert an overheating situation. It just beeps a brief warning and shuts down everything to prevent the possibility of overheating with no cooling fan.

Oh yes, I thought of another cause of reduced cooling fan performance in laptops, mainly because I saw this exact thing happen in my grandson's laptop. His machine managed to build up a collection of dust and lint that got trapped in the cooling air ductwork near the fan. You could not even see it until the machine was disassembled enough to open up that dict and see all of it. The wad of dust was rubbing the fan rotor and slowing it down badly.

i would call dust as the issue as well... when the cpu is being taxed the fan is moving more, and can blow through it..

i would get a can of air and blow it out, if that doesn't fix it, give us the model and we can probably find a guide for you to disassemble your laptop

 

[herm0016]

I have been reading about switching power supplies for a bit today....so i may have a clouded vision...


Could it be a power issue with the power mosfets on the main board? a laptop has voltage regulation built in because the switching supply only provides 19v on most and some other voltage 12? 24? on other laptops. At a higher current they supply the correct voltage, but at a low current the voltage raises or lowers to something the cpu cant handle. I am not sure of a failure mode of those components, and depends on design. they regulate the voltage supplied by the battery as well.

knock any resistors off around the cpu?

 

[mindless1]

What is probably happening is that at low power/idle states the CPU is throttling down to a lower clock speed which by itself shouldn't be a problem but also there's a voltage decrease that could be a problem if there is capacitor rot or cracked PCB solder joints. To test this, find some software that just disables whatever power saving the particular CPU supports.

BTW, it is good to tell us make and model of laptop as well as the specific CPU, etc. That can be more important than the other info in case the problem is common to that model as it gets older and more people are reporting it.

 

[Jfletcher86]

Obviously this can be a frustratingly wide range of issues and very hard to pin point for the simple fact that the issue seems to have no valid consistency to it. For that reason it leads me to believe that there may be an issue with the actual d/c jack or the internal power socket failing on the motherboard. I know that seeing the battery is being charged seems like it would rule this out, but if the voltage being supplied to the board is off by any amount (less of an issue under full load) then it will cause the board to shut off.

 

[mindless1]

^ I have to respectfully disagree, that your post was not consistent with the problem. A power jack failure would be more of a problem at load than idle, and no, voltage being off won't shut the board down as the input ac/dc adapter has a higher voltage than the battery pack in order to charge it efficiently but the board and components will run fine down to the cutoff voltage for the battery pack which is only a threshold to keep that pack from damaging itself from low discharge.

 

[Flapdrol1337]

Guess that laptop is on it's way out.

Maybe try to set the windows powerplan to high performance, that'll keep the cpu in a higher power state.

 

[Jfletcher86]

At that point I was referring to the actual power socket on the motherboard. That's why I said it may not seem like it because the battery is being charged. I worded it very poorly now that I look at it. I am talking from dc to motherboard socket, not dc to power brick.

 

[sm625]

Now, some strangeness...If I boot into BIOS, MemTest, DOS, or the Hirens/Falcon4 CD and boot into mini Windows XP, it will stay up forever.

You should expound upon this since it seems to go against the dust theory. What is this mini XP? Is it a fully functional XP? Can you run a CPU usage and temperature monitoring tool?