Upgrading poses possible temp problems

I just recently bought an AMD Athlon XP 3200+ processor to put in my computer. After setting it up and running it, I discovered that my motherboard wouldn't support the processor (so it was running at 1/2 FSB) and so I bought a Foxconn 600A01-6LRS motherboard and it now works. Oh, I forgot to mention that I bought a Volcano 11+ fan and heatsink to cool it. OK. My GPU is a nVIDIA GeForce FX 5200 (128 MB). I don't know if this is worth mentioning, but when I put the thermal stuff on the CPU, two possible problems arose: 1) Being inexperienced with the alignment and installation of the fan and heatsink, I accidentally got the stuff all over the little ?transistor? thingies on the top, but I gently wiped it off and 2) The heatsink sits so close to the CPU that I notice that it squeezes all the thermal stuff to the sides when I put it on (I see the results of that when I take it off). Here, I explain the problem. Overall system performance is great, but during CPU and/or GPU intensive applications, either the CPU warning alarm goes off (which is set to 60 C) or Windows XP SP2 crashes and says that the graphics driver has stopped working, and shortly after “blue screens” and resets. The funny thing with the CPU though is that I can’t figure out if any of the temperature readings I’m getting are correct. In the BIOS, I get a normal temperature of about 40-45 C. I have downloaded a program called Speed Fan to monitor the temperature, and I can’t tell which temperature it is because for some strange reason, it didn’t label them (the many temperature readings). Here is the process by which the program obtains sensor info:
---I/O properly initialized
Linked ISA BUS at $0290
Linked VIA VT8237 SMBUS at $5000
Found nVidia GeForce FX 5200
Linked nVidiaI2C0 SMBUS at $3D403E3F
Linked nVidiaI2C1 SMBUS at $3D403637
Linked nVidiaI2C2 SMBUS at $3D405051
Scanning ISA BUS at $0290...
SuperIO Chip=VIA VT1211 (PLEASE REPORT)
Linked ISA (flat) BUS at $6000
Scanning VIA SMBus at $5000...
Scanning I2CNVidia SMBus at $3D403E3F...
Scanning I2CNVidia SMBus at $3D405051...
SMART Enabled for drive 0
Found Maxtor 6Y080P0 (82.0GB)
SMART Enabled for drive 1
Found Maxtor 33073H4 (30.0GB)
End of detection---

Here are the temperatures it gives me, along with the labels:
HD0: 40 C
Temp2: 42 C
Temp3: 70 C
Temp4: 69 C
Temp5: 0 C
Temp6: 0 C
Temp7: 0 C
CPU diode: 16 C
Temp5-7 and the CPU diode never change, but I’m worried about Temp3 and Temp4, I just can’t figure out what they are. When the CPU Temp Alarm goes off, indicating that it’s passed 60 C, I quickly reset the PC. I go into the BIOS and see, with the alarm still going off, that the CPU temp is ±51 C. That isn’t a bad temp. Just as the temp fluctuates between 50 and 51 does the alarm also begin to subside. If that were all, it could be narrowed down to either a problem with the BIOS temp readings, a flaw in the alarm system that actually starts the alarm (if the BIOS CPU temp is correct), or a malfunctioning CPU temp diode sensor, but that’s not all! At about the time when the alarm goes off, in Speed Fan, I get Temp3 at about 71 or 72 and Temp4 at about 70 or 71. These don’t match with anything I see in the BIOS when I restart (about 15-20 seconds to cool down to ±51 C if they have anything to do with the CPU temp). So I’m very confused. Could these (Temp3/Temp4) be pertaining to my GPU, and if they are, where is my CPU temp in there, because (except for the hard drive sensor) none of the other numbers ever really change very much (± 1 C). When I run “Toast” (puts CPU at 100%, you can find it on this website), the Temp3 quickly rises to from 70.0 C to 71.0 C in 30 seconds (I hope that’s not the CPU!). It is the only sensor that really moves. If I leave it on too long (a minute or two) the alarm sounds and the temp gets to about 72 C. I quickly reset and go into the BIOS (about 15-20 seconds) and see the temp at 51 C. Well? ANOTHER SITUATION occurred while executing a graphically intensive program and after about 15 minutes, I got an error saying that the graphics driver had been disabled (everything was now in 256 colors and 640x480) and that followed with a blue screen and automatic restart. Amazingly, the temperature sensors as indicated by Speed Fan were all normal AND the alarm didn't go off and then after restarting (after the blue screen) I saw the CPU BIOS temp at about 49 C. If this is GPU related, it might be worth mentioning that my GPU does not have a fan on it, only the little thing that looks like a plastic heatsink because that’s all that came with it, as it is non-ultra (I think the “ultra” ones come with a fan). Well, thanks for reading this book, and I would appreciate ANY help that ANYONE could provide. Thank you.

 

Bump

Hello drtmiller, check out this tutorial at the arctic silver website, it shows the proper way to apply thermal compound, you dont want it to thick, the main idea is to make full contact between the heatsink and die,
the link,
http://www.arcticsilver.com/arctic_silver_instructions.htm
Sounds to me like you should cleanup and reseat that heatsink.

Im running mbm 5 on my pc,
amd mobile 2500+ overclocked at 12.5 x 200 (2500 mhz)
Idle temps
Case temp 32c
CPU Diode 57c I have alarm set at 70c
CPU socket 41c

Also, if your not getting proper cooling, that Diode temp will really begin to fluctuate, it sounds like this could be the case with your cpu.

P.S. also note: does the mating surface of your heatsink have slight roughness or grooves, it should be smooth and shiny.

 

I'm pretty sure I've got the thermal compound on right, now. I took the 3200+ out and replaced it with my old 2400+ and ran "Toast" and it's been running for awhile without any major problem, but the "Speed Fan" program is still registering the temp at about 70 C, which is too hot for me. I believe that I've come to the conclusion that "temp3" is the CPU and "temp4" is the system temp. Unless the program just isn't reading them right, 70 C for the CPU is too hot! My fan is at 5200 RPM. Now what should I do, assuming that the thermal stuff is on right?

 

Did your motherboard come with a program called "superstep"?
And if so, did the program seem to show more realistic temps?

::EDIT:: are driver/bios updates available for this MoBo?

 

Yes, but as soon as I opened it, the CPU alarmed. That's strange, isn't it? The temp it gave me was about 36. A little too optemistic I'm afraid. I have yet to run "Toast" to see if it's even a real temp but I will in the morning. Oh, I forgot to mention that when I opened SpeedStep, the system was not under ANY kind of stress whatsoever and had actually been idle for a while.

 

It does in fact give me a realistic temperature--- 32 C when I turned it on this morning. Running toast for 20 minutes yields a temperature of about 42 C, but right now, our house is 67 F. I can't find out why the alarm goes off though when I open it. I don't have any alarms enabled in SpeedStep. Only the ones in the BIOS. What could the possible problem be now? And why on earth is the other program, Speed Fan, giving me those awful temps? And if the system's only running at around 40 C with a heavy load, why do some graphic intensive programs quit, followed by a message saying that the graphics driver has been disabled and a blue screen of death and the finale... automatic restart? I have the latest nVIDIA driver---all of my drivers are up to date!