Heat problem - Tualatin

I have a Celeron 1.2A that's a bit odd I think. With idle temps above around 39-40 degrees it becomes very unstable. And this is a fact. This happens at default speed as well as in overclocked speeds, and regardless of vcore. I get it stable at 130fsb with 1.575v, idle 35degrees. At 133fsb the only thing that can make it stable is low temp. It is almost as more vcore only makes it more unstable. Got it stable at 135fsb one day with my window open on a cold day, idle 27deg at 1.7v. Ran 3dmark2001 until the low room-temp finally got to me and the window had to be closed. =)
Can also tell you that the biggest problem isn't that the computer hangs. That only happens at fsb 135+. Most common is just that f.e. "iexplorer.exe has generated an error...." and so on.

Now to the question. Is it anyone out there more than me that has this problem. That the tually is very sensetive to "high" temps?? (Personally I don't regard 39deg as high but...)

I found something similar with mine. With the vid pin trick to bring Vcore up to 1.675 I can reach 128Mhz stable, I can post and half boot @ 130Mhz but nothing past that (133 wouldn't even post). Then I tried a test using a spare peltier and found that when the chip temp got down to around 18C I could then post and fully boot into windows at 133Mhz. This would seem to imply that the chip is very temp sensative.

Why do you people not tell what you are cooling with. You have a very good cel-t and yes with the cel-t 1.2 sometimes too much vcore will make it unstable. Over time my cel-t though is able to stay stable at vcores above 1.75v. At first I could only run 1.625v @ 1.6GHz and anything above would result in it being unstable. This would happen at any speed. Say if it was at 1.2GHz and 1.8v it would still crash! Get it! THis was not dependant on the temp of the chip but more on vcore for some odd reason!

What is your HS!

It's a Taisol CGK760092. It's not a bad heatsink at all. http://www.blargoc.co.uk/reviews/taisol760/index.shtml
My Tualatin has not a problem with booting anyways. It boots at fsbs over 140 but it never reaches w2k then. Well, I know it's hard for these 1.2A to reach 133 fsb but it seems that if I can get mine to boot at 1680 with 1.7v it would be able to run stable at 1600 atleast. Nope. 130fsb is the limit. At 131fsb I get program-errors in windows.

I had the same problem of instability. 130FSb is fine, runs all games and benchmarks, but go to 133FSB, and it's unstable, even with a Vcore increase.
The whole thing of my CPU Temp at high 30's made me think it was something other then a heat problem, but I stuck a strong 60mm Fan on my HS and had 2 x 120mm's blowing in the direction of the HS and it was perfectly stable at 133, anything above 135 would crash. I think the CPU likes temps in the very low 30's.
I am going to add a second 60mm to the HS and see if it is stable enough to get above 133. (I stuck a P3 into the board and got an FSB max of 160, so it's not the board (VH6T))

This was really nice to know!!
Yesterday, just for fun I decided to go at default speed, default voltage and se how warm the cpu ran and so on. To my surprise the cpu isn't stable!! I get program-errors with default settings. The temp is at 32 degrees idle so I KNOW that's not the problem. I ran it at 2 degrees higher at 1560MHz with 1.575v and no program-errors, perfectly stable.

Should I return it and ask for a proper one???

(I should mention that the first time I tried to run this cpu was with the BX-tualatin-mod on a BH6 but my BH6 got fried so I had to buy a tualatin-board. The cpu could have been damaged there, right?)

You guy's concidered water cooling, a deffo saviour.

Ooops. :|
Seems I have made a fool of myself. Amateur-error actually. It seems my so called pc133-memory is the villain in the drama. It is my memory that is causing the program-errors in windows. For some reason it can't take fsb over 130 without being unstable. I am in luck for having a Via-chipset so I can set the memory speed to (fsb, fsb+pci, fsb-pci). Without that I could never have known. I am currently running a perfectly stable [email protected] at 1.6v. Shall do some more testing but it seems my cpu isn't halfbad at all... ;o).
My memory though is. Anyone know if memory-cooling works well?

Yes! TT passive cooling is excellent and cheap.

Brosken,


Same happened with my mem and a celly 1.2 2 months ago. Try relaxing settings and/or removing one stick (if you have more than one) and/or swapping memory slots. Then, to test memory stability, 3dmark2001 is a good test ... besides the 'traditional' ones.

Regards
FTC