328° C on ati 7950 and CPU freq issue

[Roni]

Hi,
sometimes (randomly) after a game season hwinfo64 show me this :

sometimes is vrm temp, like in screenshot, and some other times is vram temp or even also "GPU VRM Voltage Out" that show me strange values like 3.xx volt.

some other times also my Phenom II 965's freq are wrong, like 4200mhz when my cpu is at 3800mhz.

 

[Martin]

Thanks for the report. Those values are really very off and might be caused by a problem when reading the value from sensor.
To find this problem I need the HWiNFO Debug File from the situation when it happens.
Please see this post about how to create such dump: http://www.hwinfo.com/forum/Thread-IMPORTANT-Read-this-before-submitting-a-report
When you attach the debug file, please also attach a screenshot of the sensors window so I can exactly locate the invalid value.

 

[Roni]

i've a problem with "debug mode", when i run hwinfo in debug my pc freeze at "detecting vga..." and after few sec goes black screen and my pc reboot, and my mobo 3-bips me (seems no vga connected). after a 10 sec swtich power supply off, all returns to normality...
tried a few times, noway to launch hwinfo in debug mode

 

[Martin]

Does it happen during the initial phase (Detecting Video Adapter) or while scanning for sensors ?
Please attach or send me the Debug File from that situation too. I might see there where and why it does such thing and maybe can give you an advice...

 

[Roni]

it happens during "detecting video adapt" (initial phase)...but now (strangely) i was able to run it without system freeze. at start it seems struck at detecting video adapt for a while, as usual, but instead to freeze this time it complete the intial phase

however this time instead of wrong temp reading is vga's voltage, it's displayed 31.xx Volt instead of 1.20 volt (sry i forgot to screenshot this time)

debug file is too large (1.9 mb) to be attached, i've uploaded on mediafire:

http://www.mediafire.com/?2f4z37k10j34ior
hope this will help you.

 

[Martin]

Thanks for the data, I'll analyse it.
If possible, please create other such dumps with other erratic values observed as well.

 

[Roni]

ok i'll try tomorrow, hoping the "detecting" doesnt freeze my pc anymore.

 

[Martin]

I have checked the dump and it seems I'll be able to filter out such invalid values. But I need more samples.
If it freezes during start, then please send me a debug file of that situation too and I'll check it too.

 

[Roni]

this is debug file of gpu vrm temp reading (after about 1.5h) : http://www.mediafire.com/download.php?rbwsa9bvud2d7m8

on afternoon i'll do more test

 

[Martin]

Thanks for the additional information. It seems that the raw reading from the CHL sometimes returns a scrambled value.
So I'll try to filter such things out, but but sure if I can catch all possible situations.
If you have more dumps, just post them...

 

[Roni]

few minutes ago i try to run hwinfo in debug mode and that freezed my pc, like described above. the freeze occured during initial scan on "Gpu Ati #23", this is the dump.
Thanks Martin for your aid, hope this can help

 

[Martin]

Seems to be hanging at checking for the same CHL8228 chip.. Quite strange that sometimes it does and sometimes it doesn't...

 

[Martin]

One more question: are you maybe running some other hardware monitoring tool in parallel with HWiNFO ?

 

[Roni]

no, i only use hwinfo

 

[Martin]

ok.. so I think I can filter out the most invalid values, however not sure if all of them..
If you can catch the problem with Current or Power values in DBG, that would help even more...

 

[Martin]

I have just released v4.05-1735 Beta. Can you please check if it performs better ?

 

[Roni]

I have just released v4.05-1735 Beta. Can you please check if it performs better ?

im still testing for catch reading from others sources, meenwhile i'll check new vers.