View Full Version : Curious: how hot can/should a K6-2 Get
phenolite
01-06-02, 03:57 PM
Going way back w/this chip boys (and girls)...friend of mine has a 500Mhz K6-2 chip currently idling at about 43C (mbm5)...just wondering how much farther i can push it? (how hot, i mean) thanx guys
JetMech
01-06-02, 04:12 PM
There was a thread in which a member did some extensive work on pooling (all) processors into a guide which gave maximum speeds as part of the info. I thought it would end up a sticky but haven't seen it since. Wish I had printed it. Because of it's voltage demands I would cool it like a Thunderbird and be conservative on the high side meaning I don't think 5c more will hurt it. I Know from experience that overtemping that chip will take out the board before the chip.
Hey, I dug this out of AMD's developer pdf files.
With the older processors, they call the die temps 'case temps'. That threw me for a while, until I read the entire pdf. Because it has the aluminum lid, it's called case temp....go figure.
nil_esh
01-06-02, 04:32 PM
I was running my K6-2 450 sometimes at 60C when I first got it and didn't know a thing about cooling. I had to clock it down to 400 for Windows to work without crashing. Now I get like 34C load max with the case temp at 28C and ambient about 27C. It'll idle at the same as the case temp, but I don't idle it much (Folding). These are onboard thermister temps though.
I would say under like 45C load is fine.. I even ran it at 49C load running SETI for a while and it was stable. I didn't kill anything with all the heat I put it through. I've OCed it up to 550mhz with no problems too.
But the cooler the better. A cheap heatsink like the aluminum Coolmaster that I'm using is enough to get good temps on these processors at stock speed anyways. These CPU's are very easy to cool, so that 43C idle can easilly be dropped to the same as the case temp.
nil_esh
01-06-02, 04:37 PM
Originally posted by Diggrr
Hey, I dug this out of AMD's developer pdf files.
With the older processors, they call the die temps 'case temps'. That threw me for a while, until I read the entire pdf. Because it has the aluminum lid, it's called case temp....go figure.
Weird..
To clear up any potential confusion when I said case temps in my last post I was speaking of computer case, not the CPU case. :beer:
Yep. Just because AMD used to use the term, I'm not going to change what I call case temp.;)
Their engineers term for what we call case temp was "internal ambient temperature" They also measured die temp differently than is possible on the t-bird, they would use thermal epoxy to glue the probe dead center on the aluminum cover, with a corresponding hole in the hsf.
I just wanted to make sure no one objected to Tcase reference in the pic. That is what they meant to be die temperature. I didn't make the pic, I just cut and pasted into PaintShopPro straight from their pdf, so I felt an explaination/translation was in order.
One more thing, max stability temp is far different from max 'dead chip' temp. I personally would keep it under 50C at all times.
Have a great one.
They don't get hot as my dad's K6-2 500 is usually 32 degrees when using his programmer's application.
The only thing you should worry about is probably cleaning off all the dust off the fans since they're old processors and I think that there is alot of dust on them to reduce tons of CFM and create more heat.
The next thing is applying ASII as a thermal compound since the compound maybe dry and that should cut some serious degrees.
Yodums
vBulletin® v3.8.7, Copyright ©2000-2012, vBulletin Solutions, Inc.