Why is my new build going Chernobyl on me?

Toggle sidebar Toggle sidebar
Savagearms

Savagearms

Honorable
Dec 12, 2013
35
0
10,560
Hello all,

I've been lurking on Tom's for years now, but I've recently hit a snag in my new build; bad enough to warrant asking here for help.

After a careful assembly, I booted up the computer for the first time, and huzzah! everything seemed to be working. It started windows seemingly without issue. Once start-up had finished, I immediately opened CPU-Z to check on things, and noticed my CPU clock speeds varying wildly.

I shut down and waited a few minutes, then went in the BIOS, where I discovered my CPU temperature was an astonishing 93 degrees Celsius.

Again, I shut down and waited for it to cool before going back into the BIOS. I was able to watch the temperature steadily rise from 30c to 60c and beyond before once again, I shut down so as not to fry the CPU

The heatsink was then removed, and the thermal paste (Arctic MX-4) was re-applied.

I turned it on only once more to mess with the power settings in the BIOS, such as turning off C6 State, APM, and AMD Overdrive, but this did not affect the temperatures.

The specs of my rig are as follows:

-----

AMD FX-6300

GA-78LMT-USB3 Rev. 5 Motherboard (a Micro ATX board)

CoolerMaster Hyper 212+ CPU cooler

Antec TruePower Classic 650w 80+ Gold

8Gb of Crucial 1333mhz DDR3 Ram (Two 4gb sticks)

Sapphire 7870 Ghz Edition

In Win Dragon Slayer case

-----

The case is extremely well ventilated, with a total of 7 fans on the case, and 1 on the CPU cooler itself.

The only thing I can think of, is that either the Motherboard or the CPU are faulty. I suppose the PSU could be at fault, but I kind of doubt it. These AMD AM3+ mATX boards seem to be a bit on the low quality side of things, and mostly with 760g chipsets (If that even matters). The board and CPU can both be easily RMA'ed, which I'm giving serious consideration at the moment.

All of the parts are brand new from Newegg.

Any advice would be much appreciated.
 
Solution
Unolocogringo
Sort by date Sort by votes
W

Warrgarbl

Honorable
Jun 4, 2013
78
0
10,660
Such a high temperature normally means either incorrect cooler installation or overvolting. Have you tried resetting your BIOS to default? Check the processor voltages and set them to auto or to the specified voltage for your chip. I assume that you had a look whether your cooler has good contact with the CPU? Maybe the fan is not working?

If all of this does nothing then I would assume that you could have a defective CPU. From my experience (almost exclusively Intel) CPUs are VERY rarely defective, though.
 
Upvote 0 Downvote
siddharthmukul007

siddharthmukul007

Honorable
Sep 16, 2012
2,356
0
12,460
cpu temps does not really depend on the motherboard quality....depends on the cpu cooler, air flow inside the cabinet and TIM.

you can try clearing the cmos and also updating you motherboard bios (just to be sure that you are not auto overvolting)

once through, you can than go ahead and rma your processor
 
Upvote 0 Downvote
Unolocogringo

Unolocogringo

Illustrious
Dec 31, 2007
7,764
657
38,390
Upvote 0 Downvote
Solution
Savagearms

Savagearms

Honorable
Dec 12, 2013
35
0
10,560


It's a little tricky to check all the BIOS settings, as I only have about 15 seconds, if that, before the CPU goes over 60c, forcing me to shut it down to cool.

However, all the voltages and clocks seem to be at their minimum setting, and the Vcore was at something like 1.161v, which from what I've read, seems to be exceptional. However even after setting the BIOS to default, the temperature still rockets up.

The cooler seems to be fully seated, and the fan is working flawlessly.



I was just thinking that since this is a bit of a budget board, that maybe something in it may be defective.

The CMOS battery is a bit of a pain to get at (Under the GPU), but I'll give that, and re-flashing the BIOS a try later today.



I actually followed that exact guide the first time I applied it. I tried the dot method for the second application, but no change. =\

Thanks for the replies, guys. After I try fiddling with the BIOS a bit more later today, I'll also try plopping in a Phenom II X4 765 I have in my other compy, and see how that fares. I'll be back with an update of how I make out.
 
Upvote 0 Downvote
W

Warrgarbl

Honorable
Jun 4, 2013
78
0
10,660
This may be a bit of a stretch, but maybe there is some error with the solder connecting the die with the cap? Afaik the cap is soldered to the die. Maybe that came off somehow? It would explain the high and quickly rising temperatures. I work with notebooks a lot and those chips tend to get hot VERY quickly because notebook cpus have their die exposed and are not connected to any kind of metal which would help with heat dissipation.

Please try to run your PC and allow it to get hot. Then touch the cap of your CPU and see if it has heated up accordingly. Do so without a heatsink or paste on it. If the cap isn't getting hot I would RMA the CPU.
 
Upvote 0 Downvote
Savagearms

Savagearms

Honorable
Dec 12, 2013
35
0
10,560


I don't think that's the case. When I attempted to mess with the BIOS a second time, I had taken off the case panel beforehand to feel the copper heatpipes coming off the cooler (At the bottom, nearest the CPU), and it was quite hot indeed, nearly burning hot. I assume they would be cool to the touch if there was a problem with the solder.

Also, I forgot to mention in the OP that I also re-seated the CPU when I reapplied thermal paste a second time, thinking that possibly one of the pins may have been arcing. So at least I can be sure that isn't the cause.
 
Upvote 0 Downvote
Savagearms

Savagearms

Honorable
Dec 12, 2013
35
0
10,560
Right! bit of an update.

I decided to switch CPU's out before messing with flashing the BIOS, so after releasing my Pheom II X4 965 from the clutches of my other computer, I slapped it in the new build, applied paste (Which spread out pretty perfectly, looking at it after taking it off), and booted her up.

In the BIOS, I watched as the temps rose up and over 60c, albeit slightly slower.

So, this confirms that the FX-6300 is, in fact, perfectly fine. The problem now must be either with the MB or the PSU.

I then attempted to re-flash the BIOS, however for some reason, Q-Flash (The flashing software on the board), simply will not see the replacement BIOS on the USB flash drive. I reformatted it in FAT-32 as the manual for Q-flash stated, but no matter what port I plug it into, or what files I put on the drive, it simply says that it's empty (Though it does see the flash drive being there. Strangely it calls it Floppy Drive A).

It also seems to just freeze up if I attempt to save the BIOS on the MB to the Flash Drive.

Very strange indeed...

 
Upvote 0 Downvote
W

Warrgarbl

Honorable
Jun 4, 2013
78
0
10,660
I am pretty sure that the PSU can be ruled out because if it would supply an overvoltage it would blow the VRMs or damage the system in some other way. Therefore the board has to be defective, the BIOS update failing is a bad sign. It is quite possible that the VRMs don't regulate voltages properly and give the CPU more than the BIOS displays, or the sensor might be defective and the displayed voltage is lower than the actual voltage.

Do the VRMs get excessively hot?
 
Upvote 0 Downvote
Unolocogringo

Unolocogringo

Illustrious
Dec 31, 2007
7,764
657
38,390
VRM's get very hot and need airflow over them to keep them cool.
Hot VRMs cause a domino effect. The hotter they get the less efficient they are. The less efficient they are the more power they lose to heat. The more power they lose to heat the more power they use to supply the cpu with correct voltage. So they get even hotter, and the never ending cycle continues.

Most boards are designed for a down flow heat sink creating airflow over the VRMs too cool them.
The 212 does not cool the VRMs.
On mine I went with a push-pull setup with 2 fans.
One fan is all of the way to the top and the other is slid down below the fins to blow air over the VRMs.
similar to this image.
http://ecx.images-amazon.com/images/I/81-3S3oBvRL._SX385_.jpg


 
Upvote 0 Downvote
siddharthmukul007

siddharthmukul007

Honorable
Sep 16, 2012
2,356
0
12,460


why are you getting into vrm overheating????

op has processor overheating problem
 
Upvote 0 Downvote
Unolocogringo

Unolocogringo

Illustrious
Dec 31, 2007
7,764
657
38,390
VRM overheating can cause processor overheating.
It is right next to the processor on the motherboard. If the VRM's overheat they heat up the motherboard PCB, which heats up the processor socket. Which in turn heats up the processor.

In bios the temperature that you see is the sensor in the motherboard socket, not the diode(s) in the processor.
These sensors can be way off on their readings. At 93c his processor would of already crashed from overheating if this was the true cpu temp.

The only way to be sure of true cpu temperatures is to boot into windows and check the cpu temperature with software such as CPUID Hardware Monitor, OCCT, Speedfan, or the Gigabyte control center.
These take readings from the motherboard socket and the diodes on the cpu and compare.
 
Upvote 0 Downvote
Savagearms

Savagearms

Honorable
Dec 12, 2013
35
0
10,560


I'd be very hesitant to use that method..It would require the computer to be on for possibly 5 minutes or more, and if the temps are correct, that amount of time could fry the CPU altogether.



That's actually one thing I forgot to check, I may do that tomorrow.



Sounds like a good idea. Having a second fan on the 212 would mean having to take off one of the side panel fans, but it probably isn't helping much there anyway.



Yeah...I suppose I really should have done a temp reading from Windows, unfortunately I've taken the HDD's out and put them back in my old compy. It would be quite a hassle to put the OS drive back in for testing...I may end up doing that, but honestly I think I'm just going to RMA the board, and probably the CPU as well, seeing as it may have been subjected to life-shortening temps.

Also, I must say once again that I really do appreciate all your replies, they've been quite helpful.

I'll come back with an update if I do end up doing some more tests.
 
Upvote 0 Downvote
siddharthmukul007

siddharthmukul007

Honorable
Sep 16, 2012
2,356
0
12,460
the only thing you can do now is turn on your pc, get into the bios. keep that on for 15-20 mins. (see what temps do you get to)


turn it off if you see your cpu reaching 90 ish C, if it stops rising at one point you can start up with booting the os. perform a bios s update from within windows see what results do you get
 
Upvote 0 Downvote
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom