WTF?? Alpha FC-PAL6035 better than Thermoengine??

[Celemine1Gig]

I received my Thermoengine (black version) today and installed my delta fan on it. I was waiting to see a difference between the Thermoengine and my Alpha FC-PAL6035 but my Alpha produced better temps than the Thermoengine!!?? Now I asked myself if I mounted the Thermoengine correctly? The Alpha's Load temp at 1008 MHz was 36°C and with the Thermoengine I got up to 40 °C.

I noticed that the Thermoengine was sitting slightly off center on the core of my Celeron 600 FCPGA. Do you think I should get a copper shim to level the Heatsink on the CPU.

Awaiting your suggestions!

 

[OnDborder]

My 6035 did better than the MC-370 from Swiftech.

 

[Lex]

My ALPHA BOTH PEP66 and FC-PAL6035 w/ Delta 38 cool my CPU down the same level as the ThermoEngine. Sometime even better. And also my Thermalright SK6 w/ Delta38 can't even beat my old PEP66 w/ the same fan up. May be those guy at ALPHA know that their cooler are still sufficient enough to cool current CPU so they havent' released a new cooler yet.

LONG LIVE ALPHA!!!

 

[Hoot]

Celemine1Gig (Jul 14, 2001 07:15 p.m.):

-snip-

I noticed that the Thermoengine was sitting slightly off center on the core of my Celeron 600 FCPGA. Do you think I should get a copper shim to level the Heatsink on the CPU.

Awaiting your suggestions!

When I was experimenting with a hollow core ThermoEngine, I was also surprised to see that it does not sit centered on the AMD cores either. For all the Engineering effort that went into it, someone sure dropped the ball on that dimension. I don't think you need to worry about it though. I cobbled up a different retention system and eliminated the clip. This allowed it to sit centered on the core and the resultant temps did not change. As far as a shim goes. If you un-mount and remount it a lot, it would probably be a good thing to have. It may even offset the in-socket temperature reading skew caused by the exposed CPU surface inherent in the ThermoEngine design.

Hoot

 

[Nevin]

Celemine1Gig,

Your system measures actual internal core temperature via the Intel CPU's internal diode. This measurement is much more accurate than measurements taken external to the core (heat source) and is not fooled by a heatsink that blows air down on the motherboard around the socket. Once AMD users switch to the new Palomino series CPUs that incorporate internal diodes and acquire motherboards that can read and report the actual internal core temperatures of their AMD chips, they will see results similar to yours.

BTW, while the internal diode in both Intel and AMD Palomino CPUs will accurately report temperature Deltas, the absolute reading can be off by several degrees. i.e. 30C is actually 32C and 36C is actually 38C. If you want to calibrate the diode in your chip to insure that is reports accurate temperatures, here is a simple procedure:

http://www.arcticsilver.com/diode_calibration.htm

Nevin

 

[Celemine1Gig]

Thank you!

 

[CalCoolage]

Nevin (Jul 16, 2001 08:05 a.m.):
Celemine1Gig,

Your system measures actual internal core temperature via the Intel CPU's internal diode. This measurement is much more accurate than measurements taken external to the core (heat source) and is not fooled by a heatsink that blows air down on the motherboard around the socket. Once AMD users switch to the new Palomino series CPUs that incorporate internal diodes and acquire motherboards that can read and report the actual internal core temperatures of their AMD chips, they will see results similar to yours.

Nevin

It seems to me, the diode only reflects the temperature near its on-chip location. Provided the location was well chosen, so it is near the maximum temp, this could be a great thing. Maximum temp is what we really need to know. Naturally it won't have the averaging effect that measurement by a macroscopic probe just under the HSF has, nor a probe in the socket. It does not seem to me that the diode temperature is more accurate, although everyone else seems to think so. External probes measure something different.

We may find out that all those magnificent and expensive heat sinks do very little to subdue the maximum temperature and those inferior sinks the manufacturer provides are close to optimal.

 

[SCOdoubleT]

I also noticed that my ThermoEngine is slightly off, but it didn't seem to affect anything my idle CPU temps are around 29-32, nowhere near 40. Try reseating it with some AS II.

 

[Celemine1Gig]

When I talked about the Thermoengine reaching 40°C, I talked about full LOAD and not Idle temps. Celerons and PIIIs don't get that hot! They're literally COOL!

 

[SCOdoubleT]

If you've only reached 40 under load, you haven't really put it under load. Try playing UT for 3 hours and see what happens.

 

[Celemine1Gig]

Oh my god you still didn't get it!! This is an Intel CPU not one of these burning hot AMD's. Intel Celeron and Pentium III FCPGA CPUs don't get warmer than 45°C with adequate cooling. My room temperature is normally between 20 and 22°C and I have processor idle temps of about 25-28°C with CPUidle and LOAD temps of max. 42°C if room temp is higher than 24°C.

You got it?? Intel isn't the same as AMD!! I don't mean to be rude, but it's strange that you are an overclocker and still didn't get to know of the temop difference between AMD and Intel.