Chapter 3. Thermal tests
Tue, 10/02/2007 - 23:09 — Lubomir Kuzman
I'll begin with BTF95 - Fanless. This cooler I tested in the passive mode bz static air conditions. The values in tables, that were used for graphs were chintzy green coloured.
Measured values of the CPU seems to be fairness-sized. But in every "passive" PC has to be used "maybe only by my mind" a minimal 2 of the fans. But for our copper made guy is this pretty enough. You must bear up to me, that real passive PCs using only a chimney efect without fans are mostly only a dreams. So I measured it un-uselessly? No, here they are :) .
Cooler of type BTF80 had so good results. It assured me that at the room temperature can cool power of 215W of IHS of the CPU. I wittingly typed of IHS If somebody don't know why, please let know that temperatures of my CPU are temperatures of IHS in real. That means that temperature of the chip of CPU will be always higher. This depends of used heat transfer technology between the chip and IHS. If here is used classical thermal paste , the thermal gradient at the specific thickness and surface is quite big. Or for the heat transfer is used tin ( back side of the chip that is aluminium coated is tinned to IHS ) and the result thermal gradient is thanks to it's around 10 times better thermaol conductivity so smaller. So the revisions of the CPUs that uses tin are better in thermal transfer from chip to IHS, like the revisions that uses the thermal paste.
What can I say to thermal paste OEM delivered for this coolers. Durring my tests with the others, these pastes resulted as the very good. I didn't make this test with for example AC-Silver 5 or mercury :), but with Titan paste and many a things, mainly cheap. I tested also a silicone pads for the power transistors, gap-pads and many other things. Please do not this at your PC, it is not a tester. :) That's a geat pitty because I didn't write any result of those measurements, and also that OEM thermal paste is delivered only in 0.5g package.
Measured values of the CPU seems to be fairness-sized. But in every "passive" PC has to be used "maybe only by my mind" a minimal 2 of the fans. But for our copper made guy is this pretty enough. You must bear up to me, that real passive PCs using only a chimney efect without fans are mostly only a dreams. So I measured it un-uselessly? No, here they are :) .
Cooler of type BTF80 had so good results. It assured me that at the room temperature can cool power of 215W of IHS of the CPU. I wittingly typed of IHS If somebody don't know why, please let know that temperatures of my CPU are temperatures of IHS in real. That means that temperature of the chip of CPU will be always higher. This depends of used heat transfer technology between the chip and IHS. If here is used classical thermal paste , the thermal gradient at the specific thickness and surface is quite big. Or for the heat transfer is used tin ( back side of the chip that is aluminium coated is tinned to IHS ) and the result thermal gradient is thanks to it's around 10 times better thermaol conductivity so smaller. So the revisions of the CPUs that uses tin are better in thermal transfer from chip to IHS, like the revisions that uses the thermal paste.
What can I say to thermal paste OEM delivered for this coolers. Durring my tests with the others, these pastes resulted as the very good. I didn't make this test with for example AC-Silver 5 or mercury :), but with Titan paste and many a things, mainly cheap. I tested also a silicone pads for the power transistors, gap-pads and many other things. Please do not this at your PC, it is not a tester. :) That's a geat pitty because I didn't write any result of those measurements, and also that OEM thermal paste is delivered only in 0.5g package.
