Overheating Problem - DFI + Athlon x2 5000+

[trala-la]

Hello every one,

First of all, I'd like to tell you that I'm happy to find this forum. I hope that someone can help me with my problem.

I just build a system using a rather old DFI motherboard and a AMD CPU. I'm having problem configuring the board, so that the system unexpectedly shuts down because of overheating. This, even though I'm not stressing the CPU and the MB is not overclocked.

I have the following hardware;

MB: DFI LanParty UT NF590 SLI-M2R/G
CPU:AMD Athlon 64 x2 5000+ (Windsor) CPU.
RAM: 2x 2GB DDR2 800​

The BIOS is setted up, so that the MB will be turned off if the CPU temp reaches 70°C.

In order to reproduce the shutting off, I tried stressing the system using PRIME95. During the stress test, I kept watching HWMonitor until the system crashed. Just before the system dies, I get following readings:
PLEASE SEE ATTACHMENT


I've read in the forum, that the MB temperatures COULD mean the following:

THRM: CPU temperature
TMPIN0: ???
TMPIN1: Northbridge temperature
TMPIN2: Case temperature​

THRM and TMPIN0 raise almost identical. So I assume that they measure the 2 CPU cores.
As you see in the image, a few seconds before crashing I get this readings:

THRM: 67°C
TMPIN0: 68°C
CORE#1: 49°C
CORE#2: 54°C​

Now my questions...

Which is the real CPU temp? THRM/TMPIN0 or CORE#1 and CORE#2?
Why is the MB shutting down the system if the Core temps are way under 70°?
What would you suggest to prevent crashing?

Thank you in advance!

Trala-la

 

[||Console||]

The core temps are the ones you care about .
Have you checked the ram with mem test ( do you have the right setting set for your ram in the bios timmigns and voltage )
do you have a second PSU you can test with ?

 

[Mr.Scott]

The BIOS is setted up, so that the MB will be turned off if the CPU temp reaches 70°C.

This would be measured at the socket, not the core. Either increase the shutoff point or disable it. The core temp is what you need to watch. Windsor cores are hotboxes. It won't take much CPU overvoltage to get the core temps into an unstable range. Try to keep them under 55c, preferably under 50c.

 

[caddi daddi]

ok, the therm or tmpin0 are the vrm section of the motherboard.
this feeds the power to the cpu, this is what is shutting down your system.
I do not disable this feature unless I am after mandrake benching.

you need to increase the airflow to the area around the vrm that is next to the cpu on the motherboard to cool the thing off some.
the motherboard is doing what it is designed to do.

 

[trala-la]

Hello Everyone,

Thank you for the suggestions. I've raised the shut down temperature to 85°c and ran a Prime95 test to see the temperature development. After 2 Hours I get the following readings:
The THRM and TMPIN0 are stable at 71-72°C
Both core temperatures are stable at about 60°C.

I think that the temperatur problem is solved....

Now I get a new problem. I had the PC running for some time, and I got a blue screen because of a "memory management" problem. I checked the configuration with CPU-Z and in the BIOS and couldn't find anything strange (The RAM is not overclocked). The only thing what seens rather odd, is that my ram is running in 5-5-5-18 mode eventrough CPU-Z says its 5-5-5-15. This is because I configured the Ram according to the numbers printed in the sticker on the memory chip. I've tried 5-5-5-15 and got the same error. So I don't think that this is the cause. The memory is 2x 2GB Exceleram EX-4800p2-SX DDR2

I ran a MemTest86+ for over 3 hours (3 passes) but no errors were found. I read the minidump using BlueScreenView (see attached file).

I've also attached the results of CPU-Z just for info.

Does anyone knows what could be causing the bluescreen?

Thank you in advance!

 

[Mr.Scott]

Problem is still temp. PWM area is roughly 20c too warm. Core temps are roughly 20c too warm for the vcore being used.

 

[trala-la]

Hi, Mr. Scott.

I got THRM 72°C and Core Temp 60°C only after several hours running Prime95. When I just surf in internet or use Office, the THRM stays constant at around 50-53° and Core Temp at around 32-38°C. Are these values too high?

If so, where could the problem be?
The system ist not OC, CoreV and RAM-V are within spec. I have a AMD-cpu fan, a case fan on the back of the tower and a fan in the powersupply (Kiss Quite 420W).

Thanks again for the help

Trala-la

 

[RGone]

Back in the day of that older board we used to run Memtest86+ for at least 4 hours in test #5 and then test #8. Those two tests seemed to much better determine if the ram were okay on that older board with the memory controller on the motherboard.

I agree with the rest in that the temps are higher than I would like to see and believe in overall stability. Cases today are in a way much better than most cases available when that board was new and in use. I had to get some good fans in the case. A real fan to push air into the bottom front of the case and a very good fan to pull air from the case at the upper rear as that was how most cases of that day were designed.

 

[TeknoBug]

Funny timing, I JUST pulled out an old PC that has an Athlon X2 5000+ as well and it behaves very similar to this, using a Phenom II heatsink it idles ~40-45C and reaches 75C under load (not overclocked). Using a Gigabyte MA785-US2H board and 6GB DDR2 800. Wonder if I can tone down the voltage a little while keeping the same clock, I think the volt is 1.32v.