Sign in with
Sign up | Sign in
Your question
Solved

My CPU temp always reads 98C in BIOS

Last response: in CPUs
Share
September 30, 2009 10:33:53 PM

Hey everyone,

I did a build yesterday using an aBIT IP35-E board and an Intel Q9400. Origionally the CPU fan was not spinning when plugged into the CPU fan header. I plugged the CPU fan into the system fan header and it spun just fine so it was not the fan. I went into the BIOS to see the CPU temp and it said 98 degrees celcius!!! I could boil an egg on that! Anyways I did a little research and found out my mobo was on BIOS versio 15 from aBIT and that version 18 added more 45nm CPU support. I upgraded the BIOS and cleared the CMOS memory just fine. Tried the CPU fan header again and now it's working (one problem fixed). The CPU temp though is still reading 98 degrees celcius. When I feel the CPU heatsink it does not feel overly hot. Of course I cannot really feel the CPU to tell.

All in all I cannot think what could be wrong here. This mobo ran my QX6600 just fine when I was using it while my ASUS Striker was getting fixed. It was a rock solid board for me then. I am going to try reseating the heatsink with some Arctic Silver 5 tonight when I get home to see if that changes anything. So I know how to get the heatsink seated properly, is it OK to assume that maybe the temp diode in the CPU or the sensor in the mobo is shot? The CPU is brand new. I have run a CPU without a fan before just for testing in the BIOS and it never got above 40 degrees in the two minutes it was running.

What does everyone here think I should do? aBIT seems to have stopped support in the US so there is no chance of RMAing the board probably.
September 30, 2009 10:47:25 PM

I reseated the heatsink using Arctic Silver 5 and made darn sure it was secured all the way. It is still reading 98 degrees Celsius.
m
0
l
a b à CPUs
September 30, 2009 10:56:54 PM

Download CoreTemp or RealTemp. Check what your core temperatures are.

First thing I'd do though, go to the Abit website, download the latest BIOS.
m
0
l
Related resources

Best solution

a b à CPUs
October 1, 2009 1:04:25 AM

Red822 said:
... is it OK to assume that maybe the temp diode in the CPU or the sensor in the mobo is shot?

Regardless of the offset, if the Analog Thermal Diode within the substrate layers of the processor package is functioning properly, then you'll see it scale up and down with the Core temperatures by using SpeedFan and Prime95 Small FFT's.

SpeedFan detects and labels thermal sensors according to various motherboard, chipset and super I/O chip configurations, so the label for CPU temperature (Tcase) might be CPU, Temp 1, Temp 2, or Temp 3. Repeatedly start and stop Prime95 Small FFT's at 15 second intervals, while observing whichever SpeedFan temperature that scales with an Idle to Load Delta similar to the Cores. This will identify your CPU temperature sensor, which is the "Analog Thermal Diode".

If it's working OK, then the problem is miscoded BIOS, which happens all too often, since BIOS programmers must encode each new release with every possible processor variant for a given socket type. However, the good news is that the offset can be calibrated in SpeedFan by using the "configure" button. (See Section 11 in my Temperature Guide). When properly calibrated, CPU temperature should be 5c lower than mean (average) Core temperatures during Prime95 Small FFT's. It'll still be wrong in BIOS, but at least it'll be correct in SpeedFan.

If it's not working OK, then the Analog Thermal Diode is bad, which means your Q9400 is fauly. You can RMA it, or just ignore CPU temperature and use Real Temp to monitor Core temperatures only, which should never exceed 76c. Also, there is no "sensor in the motherboard". There is no thermocouple in the center of the CPU socket touching the bottom of the CPU, and hasn't been for many years, since the erra of the single core processors. Here's how it works in current technology:

From my Core i7 and Core 2 Temperature Guide - http://www.tomshardware.com/forum/221745-29-sticky-core...

Section 5: Findings

(A) Tcase is acquired on the CPU substrate from the CPU case Thermal Diode as an analog level, which is converted to a digital value by the super I/O (Input/Output) chip on the motherboard. The digital value is BIOS Calibrated and displayed by temperature software. Motherboard BIOS Calibration affects the accuracy of Tcase, or CPU temperature.

Does this help?

Comp :sol: 
Share
October 1, 2009 3:30:10 AM

Ok, so right now since the heatsink is not feeling overtly hot, I am going to go ahead and install an OS and run those tests you told me to. I'll let you know how it turns out.

Thank you
m
0
l
a b à CPUs
October 1, 2009 3:51:46 AM

Very well. I'll check on your progress.
m
0
l
October 1, 2009 10:14:50 PM

So now Windows does not want to load on properly. I am trying to put Windows Server 2008 on it, but I am having the same issue with Windows XP as well. I only have one 2GB stick of memory in the mobo with no peripherals plugged in. As soon as it is done starting Windows (The blue screen that asks you to press enter to continue installation in XP) the computer will just shut completely off and not restart. I have read that this could be caused by the HDD? It soulds like it is working properly and the HDD activity light is blinking when it is loading the kernel from the CD. Any thoughts?
m
0
l
a b à CPUs
October 1, 2009 10:35:11 PM

Either the CPU or motherboard is dying/dead/damaged.

MB could have been damaged in storage.
CPU could have been damaged in water incident.
I doubt it's HDD related, as the 98C issue points to a MB/CPU failure already.

I wouldn't know how to decide which it is without spare parts to swap in/out.

If you can find any other 775 processor, try that in there.
m
0
l
October 1, 2009 11:08:14 PM

Ok, so I put in my QX6700 in there and it is working. This begs the question is this mobo just incompatable with 45nm CPUs? Or is the CPU overall junk and I should RMA it? The mobo is giving a believable temp now for the CPU, although this thing is running at 73 degrees Celsius. It is a stock Intel cooler so I don't think I am very surprised there. The Q9400 like I said is either bad or incompatible (though the aBIT IP35-E board says it is ready for 45 nm CPUs).

Thoughts?

I took the processor out of my other desktop to do this, should I try the Q9400 on that motherboard? It's an ASUS Striker Extreme nVidia 680i board.

Never mind, that idea is right out because the Striker Extreme on the ASUS website does not say it is compatible with the Q9400.
m
0
l
October 1, 2009 11:21:53 PM

So the aBIT website lists the Q9400 as test for this motherboard. I guess that means they are technically incompatible. So now the question is do I get a new motherboard and chance that the processor is still good? Or just get another processor that is listed as compatible with this motherboard. I am thinking the first options would be the best as if the processor is bad, I can just RMA it and get it replaced from Newegg.

For a good good, solid motherboard that is compatible with this processor, what would you recommend? It has to be a rock solid board as this PC will be used mostly as a server.
m
0
l
a b à CPUs
October 1, 2009 11:47:00 PM

Just a reminder RMA is for manufacturing faults, not you spilling water on your computer. Computers aren't waterproof. I'd say the CPU got it's ass kicked when your water cooling system did it's spat. More people who abuse this, the higher prices end up, if you're gonna do it at least lie about it!
Shouldn't encourage others to do this.

But for the meantime put it to one side, and try it when your other motherboard comes back.

Are you sure Abit don't have a BIOS update to add 45nm support? Seems really unusual.
m
0
l
October 2, 2009 12:04:16 AM

First thing, I don't believe I ever mentioned anything about my system being water cooled. It has and will always be air cooled. It is using the stock Intel cooling fan. Second, I have already upgraded the motherboard's BIOS to the most current version. I am not sure though if it added support for this CPU. Thirdly, since it would be an honest RMA if I sent the CPU back for it seeming to not work, I am not going to know whether it is CPU incompatibility that is the root cause, or a generally bad CPU. In the end though, for me to get this server up and going in the time I need it to, I think I may just have to pony up for another mobo and test this CPU on it when it arrives. I don't think mobos are as stringent as CPUs in return policies to Newegg. Either that or I could just ebay the old mobo.

Next time please make sure you have your facts straight before laying down on me or others like that. I really do appreciate your help though and I thank you very much for it.

I am looking to get an ASUS P5N-D for this as it is fairly high rated and I have not had issues with ASUS before, unless you have another board in mind for me to get for this rig for around $120.
m
0
l
a b à CPUs
October 2, 2009 12:09:48 AM

Crap, sorry dude, totally got you totally mixed up with someone else.

Best of luck.

Asus are known to be a good brand, haven't heard anything bad about that board. I've had a slightly bad run with Asus but that doesn't seem to be public opinion, next time I'm going Gigabyte but I wouldn't say 'dont buy asus', I just have a grudge.

Wish I could help more, but only real tests to be sure is to swap in known good bits. Doing that with a bunch of Athlon XP machines I got given at the moment.
m
0
l
October 2, 2009 12:35:57 AM

I decided to get the GIGABYTE GA-EP45-UD3P. We'll see how it goes.
m
0
l
October 9, 2009 12:27:49 AM

It was the motherboard incompatibility. New mobo with the CPU in it is running beautifully.
m
0
l
October 9, 2009 1:56:02 AM

never purchase abit mainboard as they have long history of bad bios record. my oppinion.

msi is always be my top choice
m
0
l
a b à CPUs
October 9, 2009 1:58:18 AM

Awesome.
Glad you're going well.

Oddly, I bought the exact same motherboard yesterday.
m
0
l
November 19, 2009 10:59:34 PM

Hi

I just stumbled upon this thread as I was searching what was going wrong with my "new" IP35P & E5200 (45nm) and it is exactly the same thing.

The symptoms:

- BIOS says the CPU is running exactly 98 celsius, it does not fluctuate
- Windows XP installation stopped every time to the same point as you described (the pc just shut down), until I disabled the "shut down pc when cpu is xx celsius" from bios and after which it worked fine.
- my CD-drive was working really slow

Even though I have Win XP up and running now, my PC crashed just 5 mins ago after I opened a third tab in firefox.. haven't tried gaming yet.
m
0
l
November 23, 2009 11:41:51 PM

It took me using a different motherboard to see that it was not the processor causing the issue. The aBIT website does not list the CPU I purchased as compatible with the IP35E motherboard.

I solved this by using the different motherboard. The only other thing that might help is a BIOS update, but since aBIT does not make motherboards anymore, the chance of a newer BIOS update coming is probably slim to none.

Sorry I could not be of more help.
m
0
l
!