AMD FX(tm) -4200 Quad-Core Processor Temperature Problem

BallisticMouse

Honorable
Mar 27, 2015
29
0
10,530
My desktop has been shutting itself off lately so I checked the fan speeds and temperatures in the BIOS.

My CPU, the one listed in the title, is pre-set to shut off at 70c/158f and it seems to be at a resting temperature of 69c/156f and that does not seem right to me at all. The CPU's fan is showing 3096RPM in the BIOS. Is the set shut off temperature too low? What's making it hot? I don't believe it's as simple as dust because I cleaned it out pretty thoroughly yesterday.

Any help to figure this out is much appreciated!
 
If your system is idling at 69°C you have a very serious problem. If the CPU cooler fan IS running, and I'd check visually to be sure, then it's most likely because either there is no thermal paste or the heatsink is not secured to the CPU lid. Check to make sure all the corners/sides of the CPU cooler are fully fastened down and none have come loose.
 
I checked visually before posting this and all fans are running. I'll check after work today if it's fastened down correctly or not, do you think thermal paste is something I could pick up in an electronics/hobby store or is it a order online only kind of deal?
 
Any computer shop, Best Buy or service center should carry some form of it. If you don't have one in your area, then you may need to order it online. I'd probably recommend just going with something like Arctic Silver 5 or another name brand. It doesn't have to be top of the line. There's not that much thermal difference from top to bottom between brands and formulas anyhow. The type of paste is certainly not the issue.

When you "cleaned" out the computer, did you remove the CPU fan heatsink and then reinstall it? Did you blow out the CPU cooler heatsink fins and the inside of the power supply? If not, I'd also pick up a can of compressed air from a hardware center or Wal-Mart and so so. Clean using short blasts. Continuous release of pressure from the can will cause freezing and moisture buildup on components, which of course is undesirable.
 
Like what is shown at the following tutorial. It's for the 212 EVO, but all modern aftermarket heatsink applications are similar.

http://www.tomshardware.com/faq/id-2520482/solving-temperature-issues-hyper-212-evo.html



And don't take this the wrong way, you'd be surprised what we see here, but keep in mind the heatsink needs to be removed, as does the CPU, then use isopropyl alchohol to remove ALL the old paste from the bottom of the cooler heatsink base and the top of the CPU lid using a lint free cloth or coffee filter so no fine particles are left behind. Blow the surface off when done. Then reinstall the CPU and apply the past to the top of the CPU lid. Don't overdo it. Less is probably better than too much, although the RIGHT amount is what you really want. If you can see thermal paste squeezing out on any of the four sides after installation, you used too much. A very small pea or large uncooked wild rice grain sized amount is just right.

 
I'll get back to this thread tomorrow after work when I pick up some Thermal Paste. I'll respond with what the idling temperature is after the thermal paste is applied. I want to make sure you, and future thread finders, know if this worked or not.
 
I cleaned off the old thermal paste residue, applied a proper amount of Arctic Silver 5, waited about 30 minutes (I've read you don't have to wait at all for it to settle), and turned on the PC going straight into the BIOS to monitor the CPUs temperature.

The first number I witnessed it at was 42c but it slowly climbed its way to 60c before I turned it off to post this.
 
Have you manually overclocked your CPU? Try resetting the bios to optimal defaults and see if the problem is still there. If you're not on the most recent bios version, upgrade to the latest bios firmware. It really sounds to me like the heatsink is either not fully seated or the fan is simply not working right. Make sure after you reset the bios to assign a specific cooling profile to the CPU fan in the bios. I'd recommend the "standard" setting to start with.

What are your full hardware specs. Motherboard, case, number of case fans, power supply, cpu, ram, etc.
 
I have never overclocked the CPU.
It doesn't have an optimal defaults but it has a setup defaults, should I try that?
I'll check out the BIOS version in a bit.
The heatsink seems to be fully seated and, unless the BIOS is reading it wrong, the fan is still spinning around high 2000's to low 3000's.
I set the CPU fan and shut off heat to standard, it was set to manual so I could turn up the minimum fan power on it.

I'll try and get the hardware specs soon but it's a 2-3 year old pre-built iBuyPower that I can't seem to find in a 2min search.
 


The only thing in that entire monologue that made sense is the fact that when you're in the bios the system does in fact provide full spec voltage, with no decrease in voltage or multiplier, therefore no "idle" type environment. This will cause a slightly higher "sitting" temp, maybe 5-10 degrees above what you'd see at idle on the desktop. It would certainly not cause a 65 degree temp just due to being in the bios. 65 degrees is what you should see after running Prime95 in full stress for fifteen minutes, not sitting idle or navigating the bios.

The motherboard does not "create" more resistance. EVER. The resistance of the motherboard circuitry is precisely designed for the chipset, CPU and hardware that will be used with it. If the motherboard were causing higher temps it would be due to faulty VRM's, dust choked heatsinks or a short circuit that was causing a direct short thereby overvolting, well, probably everything, or at least some things. On a system that POSTS fine, but overheats, the motherboard is the last thing I'd suspect and it wouldn't be due to "resistance". It would be due to using the wrong voltage or incompatible hardware. If there was too much resistance in a circuit, it would likely burn out or simply cause a hardware failure. Or not POST at all.

Changing the CPU voltage to manual, but not changing the values, is the same as doing nothing at all. I have no idea where you got that idea, but it's definitely incorrect. Even manually set voltages will fluctuate somewhat due to other settings and features baked into the bios control functions. This theory is simply full of fail. Now, manually setting a LOWER voltage might decrease temps, but it's also likely to create instability, depending on various other factors. If the system is set to default settings and not overclocked, it should not overheat unless the CPU or motherboard are purely faulty or the cooling system is not working.


I'd highly recommend taking the CPU heatsink back off, removing the CPU and using a magnifying glass or other device to inspect both the motherboard and cpu for bent pins or sockets. Bent pins can short the CPU and cause a voltage/heat issue as well as a lack of cooling. I'd also make absolutely certain that nothing, including the clasp that locks down the CPU, is interfering with the heatsink being fully seated on the CPU lid. CPU temps that rapidly shoot up are usually due to no thermal paste or incorrect/obstructed heatsink mounting. It could also be a faulty BIOS thermal table, which a bios update is likely to correct. Having an unsupported CPU that requires more voltage than the system was designed to handle could also create issues. Low end motherboards with high end CPUs often present such issues.


Knowing your full hardware specs as I asked for before would certainly be a good place to start. Another possible issue could be using a cooling fan that was not designed for use with a CPU of a specific TDP. If the fan you're using is the one that came with the processor, or a decent aftermarket cooler, it should not be to blame. If it was one off a different CPU, it might not be up to the task, but without knowing your hardware, it's impossible to say.
 
In HWMonitor I see AMD FX resting between 50-55c (Edit - After a couple minutes it was resting at 70c's area.) and sometimes jumping to 60-65 for one moment and I also see ASUSTek sitting at 64c.

My Bios needs to be updated from version 1303 to something around 1600 but I don't have a flash drive and am having hard troubles finding another way to do it.

The fan i'm using for the CPU is the one that came with it.

Computer is an ibuypower i-series 301 trying to find out the specs asap.
 
A friend recommended I try SpeedFan to test if the motherboard's sensor was wrong.

With SpeedFan in control of my fans it has my CPU down to 36c at 1% cpu usage with it running.

From what he said that means the motherboard's sensor is messed up and he doesn't know a way to fix it. New motherboard is my only option besides setting up SpeedFan as a task to start upon startup?
 


You are aware of the forum rules? You are aware people clicked on your thread to assist you? That attitude is totally unnecessary and unappreciated.

 
Just as a suggestion, to you both, in the future, I'd use HWinfo rather than HWmonitor or the Open hardware monitor. Both those utilities are known for problems with reporting certain sensors or just plain inaccurate reporting of sensor data. Core Temp and HWinfo are the more reliable utilities.