ASrock motherboard temperature problems

[Kamen Mladenov]

I'm using an ASrock Z68 Pro3, but i've recently had some blue screens when i run a demanding game or my computer is overtaxed, so i decided to clean it up, update some drivers and run some diagnostics. What i noticed was some weird fluctuations in my "Motherboard" temperature, i viewed it with Everest and this showed up

http://imageshack.us/a/img526/3120/87cs.jpg
http://imageshack.us/a/img10/5978/103cx.jpg
http://imageshack.us/a/img854/8279/67c.jpg
http://imageshack.us/a/img594/1391/98549671.jpg

These pictures were taken when my PC was idle and I've seen it go from 3 all the way to 130 degrees Celsius. Most of the time this "Motherboard" sensor and it's temperature doesn't even show up, it just appears every few seconds and has a different value...Anyone else have this problem or know the cause?

Thanks in advance for your troubles
- Kam

 

[unksol]

The sensor or the program is faulty

 

[RetiredChief]

A question, Which are at Idle and which are under load.
Which CPU and is it OCed (I know its a Intel something).

Your first screen shot which shows a large diff between core 1 and Core 4 could simply be that you are running a application that is single threaded and using just core 1.

Look at idle, should not be more than about 5 C diff between cores. Temps should be about 5C to 9 C above ambient temps, ie if room is 22C (72F) then cores sould be around 26C to 31C . If higher then probaly a poor HSF, ie using the LOW performance Intel stock HSF (I never even take it out of the Box), or if the 212+ HSF, improper Paste application.

Next Run a program that stresses the cores equally. I run Prime 95. Again temps per core max diviation about 10 C (Good HSF and good paste application diff should be 2-5C difference). MONITOR temps when runing Prime 95 do not leave unattended for the first Half hour. Normally they will reach close to max within the first 15->30 Min. PS also monitor Your +12V - Min is 11.4 (I use a Min of 11.6V) and the diviation beween Idle and loaded w/prime 95 is 0.3 V ( ie if at idle it is 12.0 volts then it should not drop below 11.7V.

You can check GPU temps with furmark. Again inaddition to Monitoring GPU temp, monitor Your +12 V - SAME constraints as above.

PS does not look like a sensor, or monitoring program problem. More likely a HSF installation problem or incorrect spread of the Thermal Paste.

Prime95 link: http://www.mersenne.org/freesoft/
Furmark link: http://www.ozone3d.net/benchmarks/fur/

 

[Kamen Mladenov]

Sorry, my full post wasn't showing for some reason, as i stated all of them were at Idle and i thought i had a sensor problem or something, but when i read about improper thermal paste application it made sense, i wasn't the one to put the machine together, it was ordered and also overclocked at the retailer's office, i might have to get the custom cooling off and have a look at it, since the warranty expired about a year ago

Also, it seems i might be having bigger problems than a sensor glitch if the temps should be at most 10 degrees higher than ambient (26-28 ambient) and at most with 5 degrees difference in cores...
I'm using an intel i5-2500k originally at 3.3GHz, but clocked to 4.5GHz with scythe cooling (http://www.scythe-usa.com/product/cpu/121/scnj3100_index.html), either there are some major ventilation issues in my coolermaster tower or the cooling is insufficient..
When i'm watching the CPU core temps, cores 1,2,3 variate from 42 to 55, almost the same values, core 4 seems to differ from them and goes from 37 to 44.
I'll find that program and run it for a while to see what results i get, thanks for the info and links, Chief

 

[RetiredChief]

When sitting Idle (or should I say what should be idle), you have to include background tasks that may kick in. Reason I put more stock in using prime95.
Currently only have my i5-2500K @ 4.2 and W/Prime 95 my temps were just below 60C. I did check @ 4.6 and it was in the neiberhood of 65C (Memery serves me right-lol). My HSF is a Zalman 9900 Max. Might double check it when I get home tonight.

PS the reason I backed off from 4.6 -> 4.2 is that that diff only repersents a 9.5% gain - NOT worth it as it generally takes about a 15% increase to even notice it outside of a benchmark.

 

[Kamen Mladenov]

P.S. Turns outs it wasn't completely idle, i noticed some crappy Intel process than was using up 25% of the cpu all the time, might aswell have been only one of the cores, I disabled it and uninstalled the corresponding program (intel rapid storage technology), since i don't need it with a single HDD and apparently after checking cpu usage logs it's been doing that for a long time, i just didn't notice it. After I removed it thouh, the temps dropped to this
http://imageshack.us/f/7/cpuil.jpg/
"Motherboard" temp still fluctuates wildly, cpu core temps dropped a bit, core 4 still different

 

[Kamen Mladenov]

Well i'm not a very knowledgeable person when it comes to overclocking so i left it to the people that sold it to me, had i know there was no difference, i would've made a different choice, i'll do some test later and post what's the result...i hope i don't get any more blue screens after cleaning it up, but i'm still wondering about this "Motherboard" temp

 

[Kamen Mladenov]

All of a sudden a sensor for the temperature of "Aux" appeared in Everest. After testing for 10-15 minutes with Prime 95 my max temp was about 62-65 on all cores except for the 4th one, which as usual was about 7-8 degrees below the other ones.
Here's the mysterious Aux reading:

http://imageshack.us/f/585/auxsens.jpg/

By the way i'm not very clear on where to check for those +12V values, a +12V appeared under "Voltage Values" in the sensors tab in Everest AFTER i ran the first cpu test, along with the Aux sensor, but i'm not sure if that's what i'm looking for.