PDA

View Full Version : 4 monthes later it needs more voltage?


sendatooli
06-03-01, 11:25 AM
my p!!! 700 cc0 was running 144fsb, which actually puts it at 1016 on my machine. it was running this rock stable at 1.95vcore setting. it would never even post higher than 144, might be a lousy northbridge. on my mobo the 1.95v settting actually puts the vcore at 1.98. now 4 monthes later, it will no longer loop 3d apps for more then about 5 hours. so i set the jumper at 2.0vcore. which runs at 2.03 to 2.05. 3.3 is at 3.58v. my bx was getting a little hotter so i got a bga. it is cool again. actually the mobo read 1 degree cooler with a big greenie, instead of the bga cooler. is that possible?
i think the voltage could go down .5 with the big greenie instead of the bga. but i spent 20 bucks on the bga so i think ill use it for now.
well at the now higher voltage, i can loop all the 3d apps except 3dmark 2001 which eventually crashes on any system i use it on. i think this app is too buggy.

well i was wondering three things.
1) is this chip suffering from electromigration?
2) is this chip dying, and 2.05v is only gonna speed up its funeral date?
3) should i chalk up the bga cooler to a bad investment, and put back the big greenie and go back to 1.98vcore? or will it end up needing 2.05vcore in a few more weeks anyways?


alpha pep66, delta 38cfm,soyo sy-6ba+100,namebrand 222 ram.
i'd rather not kill this chip!

any advice is greatly appreciated,
sendatooli.

sendatooli
06-03-01, 09:52 PM
hmm, no one answered yet. just wanna know, should i turn down voltage(and fsb). i dont wanna kill this poor little guy!

thanks,
sendatooli.

Placid
06-03-01, 11:10 PM
In my opinon I would not run a p3 coppermine at more than 1.95v max for everyday use.
Intel claims the cpu will burn out at 2.3v thou some here run more. Make sure you check the vcore voltage with a 3rd party program to see what its really running at as many motherboards will actually put out more voltage than what you set it to in the bios.
I would highly doubt the greenie does better than active cooling.
Of course if your dumping hot air right on it not much is going to help alot.

outhouse
06-03-01, 11:24 PM
When i started overclocking my p3 1g i tried to burn in by letting the proccesor get to 50C for 5 minutes i ended up loosing stability and my CPU needed more CV to keep it stable i ran it that way for a couple of months until i vidpinned for more CV and now i'm running at 2.05 and have been for a couple of months and i have not seen or had any more problems with stability. All chips are different and its hard to say if your killing your CPU or not but at 1.95 your probably not doing any damage and if you have well the damage is done, many guys around here are running at 2.00+CV theres a good chance it is fine now and has stabilized at its current level up the CV to where its stable and then keep an eye on it if it remains stable you may have room to take it higher or just keep it where its at. if you are that worried about it clock it back down some and keep an eye on it.

DocClock aka MadClocker
06-04-01, 04:22 AM
Could it just be that you need to turn down the fsb for the summer?..A lot of people have had to crank 'em down for the summer...like mine, I was running at 1125mhz and eventualy turned it down to 1050 for the summer (the reason I turned it down so far, is because I will be leaving it alone durring the heat of the day and don't want to burn it up while I'm gone) this comp runs 24/7 and is crunching seti durring inactivity.
Do you use Artic Silver? if you do, it might be time to reaply it...I do this about every three to four months

sendatooli
06-04-01, 04:33 AM
nah, its 65 degrees in this room. i do reapply the artic silver. i also tightend the clamp on the pep66. i guess the chip is just wearing out. it still runs 31-32c at idle and 44-46 max load. system 28-29 idle 35-36 max load. this is about what it always ran. crud, im only at 1016mhz on a cc0 needing 2.05v. this sucks. it would be one thing if i had to raise it to 2.05 to go 1070, but it wont even post over 1016. never did. it runs rock stable now at 2.05, but it was running at 1.95. it must be electromigration and the chip is just wearing out. what do you guys think?


happy overclocking,
sendatooli.

Storm
06-04-01, 05:38 PM
What I'd try is this: remove the heatsink and you should notice that the thermal paste is all dry and cracked up, so not helping heat transfer as well as it used to... clean the chip up and re-apply some thermal paste, sink it back up and see how stable it is again :)

logos
06-04-01, 07:33 PM
i'd ramp it back down for the summer. 700 is still pretty fast!

...lol...

yeah i think a lot of people are having to slow'em down. my temps have only gone up a few degrees and i am having to run her slower anyway. she keeps dropping...and dropping...i also noticed that a few degrees rise in idle temp made the load temp go quite a bit higher.

sendatooli
06-06-01, 02:10 AM
nah, it is freezing in here.
it always ran stable at 1.95 for cpu burn or seti. i never really looped 3dapps before. so it may indeed be that it needed 2.05 all along. strange, this chip. it wont even post at 150 fsb. it now runs rock stable in anything at 2.05vcore at 144 fsb. i would think if it wouldnt even post at 150, it would be unstable at 144.
well, barring the fact that it might fry at 2.05 im not complaining. at 144 fsb my mobo's clock puts it over 1016mhz. soon it will be time for another cpu and mobo anyways.

thanks to all who replied,
sendatooli.