52C full load maze3? what the!@#?

[pHaestus]

You are throwing good money after bad basing cooling upgrades on the Asus cpu measurment from that motherboard. Here are some comments from me regarding exact same mobo and its temperature readings about a year ago:

"Of course it is hard to know how effective that this method of cooling ever was since in-socket probes are notorious for being cooled by secondary methods. You can see in the individual test graphs that not only was the Asus socket probe incorrect in absolute terms, it was not even able to reproduce the proper order for the coolers performance. To further complicate things for the end user, it tends to underreport temperatures when cooling solutions are not performing well, and overreport temperaratures when the cooling is performing well. This is due to the approach that Asus has taken (a probe with a high temperature compression combined with a 10-11C offset to correct for this). The end result is that people who spend lots of money on cooling and tweaking may needlessly mess with their setup trying to figure out why it isn't performing up to par, while people with dangerously inadequate cooling may have a false sense of security. All around it is pretty much worthless."


from here
http://www.voidyourwarranty.net/revi...p1/index8.php3

Just went and looked at my Maze3; there is indeed a "short side" and a "long side" wrt the mounting holes location to the edge. You should probably mount it so that the short side is the one facing the raised part of ZIF socket. I don't have it mounted at this time, so I can't tell whether mounting it the opposite way would catch the raised part of the socket though.

[bigben2k]

The mounting holes are actually 0.10 inches, off of the center of the CPU die (from AMD specs).

You should be able to measure that, as a bit less than 3mm.

[jaydee]

Originally posted by spinky
SK6 with 80mm sunon gives 56C max load.. this does 52C load.. so we can agree that using the block gives better tempreture..but its probably a 4-8C diffrence? (is it right to think this way?)

I am also thinking this is not so bad. 4-8C drop with that setup is probably pretty good over the SK6 with the 80mm. I think you can still do better if you really want to, but I don't know how much good it will do in reality.

See pHaestus last post and read and understand it if you havn't already. I thought it was pretty good and right.
especially this part:

The end result is that people who spend lots of money on cooling and tweaking may needlessly mess with their setup trying to figure out why it isn't performing up to par

And to add to that from my own opinion "people spending lots of money on cooling and tweaking only to find it didn't gain them much if anything"

[airspirit]

Here is my take:

Lap your water block down to under 1200 grit. This is possibly issue #1.

Use just enough AS3 to give your block and CPU a sparkly look, but don't goop it on. I would suspect your joint is bad.

Also, check your thermistor and see if it's pushing against the CPU. If it is free floating, then the air under the die gets really hot since there is no circulation. Naturally, you'll get horrendous readings.

Check your lapping, though, and don't use too much goop!!!!!!!