[SOLVED] Overheating problem but not the usual suspects

[sorcerer]

Got an Acer Aspire 5920G running Windows 7 in with "slow running" as the reported problem. When I fired it up it boots to the desktop ok but the fan is constantly running and you can hear it modulating from moderate speed to high speed almost all the time.

I stripped it down to find that the airway was almost blocked solid with dust and fluff and the thermal paste had gone solid, so I cleaned out the airways, put new thermal paste (Arctic MX2) on the CPU and a new thermal pad on the GPU.

I rebuilt it and booted up again, and the machine is now noticably quicker and speedier, almost back to normal I'd say. However, I noticed that there was no antivirus software on it so began to investigate that and found all sorts of rubbish on there, so a run of various things such as Hitman Pro, MBAM, MBAR, TDSS Killer, Rogue Killer and a few others later, and I think I can now say that the machine is clean, running smoothly and back up to full speed now.

However, the temps are still way up there. Before all the work Open Hardware Monitor reported 80C at rest and 85 to 87C when watching a Youtube video (at 480p and full screen). After clearing the airways and new thermal paste/pad there was a couple of degrees fall but no more, and after removing the malware there was about 1 degree fall, so now it's reporting about 76C at rest and about 81C when watching that video.

I know that's way too high but what else can I do?

 

[nlinecomputers]

A quick google search shows that running hot is a common complaint about this unit. Not sure you can do much to address bad engineering design.

 

[frederick]

Thermometer could be out of whack. Take infrared thermometer to it and see what you get

 

[sorcerer]

Thanks guys, I'll have a look with the IR thermometer just out of curiosity but if hot running is a design problem, then fair enough, I've done what I can.

Cheers

 

[joydivision]

The processor can only clock back due to thermal overheating from the signal the thermistor gets so I doubt it is a faulty reading if it is now running better.

Could the heat pipes be damaged internally?

 

[frederick]

The processor can only clock back due to thermal overheating from the signal the thermistor gets so I doubt it is a faulty reading if it is now running better.

Could the heat pipes be damaged internally?

Thermoresistors have a fault tolerance like any other resistor. Bad resistors or poor quality are also common, but not as common as say 20 years ago thanks to better QA testing. its rare, but it still happens in purchased computers where the thermoresistor is at fault or the component that translates it. Testing with an IR thermometer is a good way to see if if its working right or wrong. It may not pinpoint thermoresistor, bios, or anything else, but it does narrow down the things

Heat pipes could be damaged but that would be noticeable. A less pure copper or aluminum mix could be another problem. The pipes are usually hollow and sealed, filled with a liquid/vapor that handles the heat. As long as a soft bend (not a direct 90 degree turn) is taken, you should be fine. If you see anything like cracking, this means the liquid/vapor could have escaped and this will cause heat dissipation to not work correctly. You'd also see discoloration along the outer surface further away from the heat source. Understand bends in the piping are sometimes performed by applying heat (a torch) to the piping when making the bend, so you'll want to check the straights. Discoloration can mean either an improper seal (no more liquid/vapor), it can mean that the mixture is incorrect, or something is running too hot for the liquid/vapor/metal to handle.

 

[Doc+Com]

Thermoresistors have a fault tolerance like any other resistor. Bad resistors or poor quality are also common, but not as common as say 20 years ago thanks to better QA testing. its rare, but it still happens in purchased computers where the thermoresistor is at fault or the component that translates it. Testing with an IR thermometer is a good way to see if if its working right or wrong. It may not pinpoint thermoresistor, bios, or anything else, but it does narrow down the things

Heat pipes could be damaged but that would be noticeable. A less pure copper or aluminum mix could be another problem. The pipes are usually hollow and sealed, filled with a liquid/vapor that handles the heat. As long as a soft bend (not a direct 90 degree turn) is taken, you should be fine. If you see anything like cracking, this means the liquid/vapor could have escaped and this will cause heat dissipation to not work correctly. You'd also see discoloration along the outer surface further away from the heat source. Understand bends in the piping are sometimes performed by applying heat (a torch) to the piping when making the bend, so you'll want to check the straights. Discoloration can mean either an improper seal (no more liquid/vapor), it can mean that the mixture is incorrect, or something is running too hot for the liquid/vapor/metal to handle.

I've had a laptop with a failed heatpipe and the easy way that I diagnosed it was seeing the CPU temp reported with software and comparing it with temp measured at the outlet (side/back of bottom base). The air coming out of the outlet was just room temperature will the CPU was so hot it was about to shut itself down. The heatpipe showed no visible signs of fracture/puncture/staining.

 

[frederick]

I've had a laptop with a failed heatpipe and the easy way that I diagnosed it was seeing the CPU temp reported with software and comparing it with temp measured at the outlet (side/back of bottom base). The air coming out of the outlet was just room temperature will the CPU was so hot it was about to shut itself down. The heatpipe showed no visible signs of fracture/puncture/staining.

Ouch....rare, but it happens. Did you check the solder points by any chance?

I forgot to mention, if they don't properly add the correct amount of liquid/vapor, this can result in the same: over heating. You aren't going to see a bad mix of metal, or if they didn't correctly add the liquid/vapor.

 

[Doc+Com]

Ouch....rare, but it happens. Did you check the solder points by any chance?

I forgot to mention, if they don't properly add the correct amount of liquid/vapor, this can result in the same: over heating. You aren't going to see a bad mix of metal, or if they didn't correctly add the liquid/vapor.

I didn't see much point in doing any further inspection of the heat pipe after seeing the temps.....even if I found the issue with it, I couldn't fix it.

 

[MobileTechie]

I've had a laptop with a failed heatpipe and the easy way that I diagnosed it was seeing the CPU temp reported with software and comparing it with temp measured at the outlet (side/back of bottom base). The air coming out of the outlet was just room temperature will the CPU was so hot it was about to shut itself down. The heatpipe showed no visible signs of fracture/puncture/staining.

Yes I've seen these several times and diagnosed it the same way - hot laptop (often with a hot patch on the back of the case) but relatively cool air being exhausted. Not signs of damage to the pipe.

New HSF assembly and problem solved.