Increase in rpm not in proportion with decrease of temp

ACook · Post by **ACook** » Sun Oct 14, 2007 12:25 pm

hoi

is my assumption that when I see only a few degrees drop in temperature on my zalman if I go from 60% to 100%, that there is bad contact being made with the cpu, and I need to reseat it?

just upgraded quickly to a x2 4200+ from 3000+ (see below for system)

Felger Carbon · Post by **Felger Carbon** » Sun Oct 14, 2007 12:36 pm

ACook wrote:is my assumption that when I see only a few degrees drop in temperature on my zalman if I go from 60% to 100%, that there is bad contact being made with the cpu, and I need to reseat it?

The change in temperature with airflow is low in HSFs with wide cooling fin spacing, and high with narrow cooling fin spacing. The Zalman has relatively wide cooling fin spacing, so no, you don't need to reseat it.

Overclockers tend to gravitate towards narrow fin spacing, SPCR types toward wide fin spacing. This generalization gets modified by other factors and specific situational needs...

ACook · Post by **ACook** » Sun Oct 14, 2007 1:52 pm

I did a rather hurried replacement. I will reseat it in a week or so. cause even at 100% at full load the temps go around 60C without undervolting.

eventually will have to take the whole board out and put a backplate and front retention bracket thingy on for the ninja I got as well. need to find the screws first.

andyb · Post by **andyb** » Mon Oct 15, 2007 4:36 am

60c for a 4200+ is too hot for any half decent cooler, bad contact as you have mentioned is possible, it could just be that the sensor is way off, what was it like with the 3000+

Andy

Solid Snake · Post by **Solid Snake** » Mon Oct 15, 2007 4:44 am

Do you really think it will be proportional? Think about it, if you run your fan at 100%, do you expect the CPU to hit absolute zero? Diminishing returns applies here.

Bluefront · Post by **Bluefront** » Mon Oct 15, 2007 4:57 am

If you want to see what "diminishing returns" means with respect to airflow, take a look at some of the temperature numbers in this thread. An increase in airflow from 22CFM to 70CFM, only lowered the CPU temp about 2C. There is a certain "sweet spot" in airflow ...... go much higher, and your cooling benefits level out.

ACook · Post by **ACook** » Thu Oct 18, 2007 10:35 am

Of course I didn't expect it to continue linearly down to 0. just a bit more than the 2C difference between 60 and 100%.

Anyway, I did just reseat the zalman, placed a grain like as5 says and pressed it down with a bit of twisting, and screwed it on well, better than last time as well.

temps are indeed lower, at 1.25V stay under 60C at 60% cpufan and 2xpi95, and it's running at 1.15V now, doing few hours pi95 for stability before I lower it to 1.10 even. and gradually move down.

all this at the stock 11x200 again, so don't have to lower the multiplier anymore to feel safe, which is nice.

Increase in rpm not in proportion with decrease of temp

Increase in rpm not in proportion with decrease of temp

Re: Increase in rpm not in proportion with decrease of temp