VOGONS


socket 4 cooler

Topic actions

Reply 20 of 25, by 5u3

User metadata
Rank Oldbie
Rank
Oldbie
retro games 100 wrote:

I'm going to try keeping these CPUs cool using a fast case fan, without a heatsink, just to check whether the CPUs work OK. Stop me if you think I'm going to start a fire. 😉

This doesn't seem to be a good idea - you wouldn't start a fire, but running without a heatsink puts a great strain on the CPU package, as some areas will get much hotter than others.
Better try the other way around: No fan, but heatsink installed. You can leave out the thermal paste for testing (make sure the surface of both CPU and heatsink is flat and clean).

Reply 21 of 25, by retro games 100

User metadata
Rank l33t
Rank
l33t
5u3 wrote:
retro games 100 wrote:

I'm going to try keeping these CPUs cool using a fast case fan, without a heatsink, just to check whether the CPUs work OK. Stop me if you think I'm going to start a fire. 😉

This doesn't seem to be a good idea - you wouldn't start a fire, but running without a heatsink puts a great strain on the CPU package, as some areas will get much hotter than others.
Better try the other way around: No fan, but heatsink installed. You can leave out the thermal paste for testing (make sure the surface of both CPU and heatsink is flat and clean).

Thanks very much for the advice. I won't "cut corners". Instead, I'll dig out a heatsink. But I won't bother with any paste! 😀

Reply 22 of 25, by 5u3

User metadata
Rank Oldbie
Rank
Oldbie

If you want to maximise the effect of your heatsink, but don't want to apply/clean off thermal gunk, just use a few drops of tap water instead (apply with a wet Q-tip, for example). It will work better than any thermal goop, but only for a short time (until the water evaporates).

I got the idea from this very entertaining article about thermal grease.

Reply 23 of 25, by retro games 100

User metadata
Rank l33t
Rank
l33t
5u3 wrote:
retro games 100 wrote:

I'm going to try keeping these CPUs cool using a fast case fan, without a heatsink, just to check whether the CPUs work OK. Stop me if you think I'm going to start a fire. 😉

This doesn't seem to be a good idea - you wouldn't start a fire, but running without a heatsink puts a great strain on the CPU package, as some areas will get much hotter than others.
Better try the other way around: No fan, but heatsink installed. You can leave out the thermal paste for testing (make sure the surface of both CPU and heatsink is flat and clean).

I decided to foolishly ignore your sage wisdom, and tested a socket 7 board with an Intel Pentium 1 MMX 166mhz chip, without using any heatsink. Instead, I positioned a very fast 12cm case fan really close to the CPU. I had the machine running for about an hour, installing Windows 98 from scratch. The installation succeeded. I powered down the machine, and immediately pressed my fingers against the CPU to see how warm it was. Without a trace of exaggeration, I would say that the feeling of warmth was barely noticeable. Extremely surprising.

About a week ago, I ran Windows 98 on a 486 using an Intel 486 SX 33mhz chip. That had no heatsink either, but I didn't use any fan. After about 30 minutes, I tested to see how warm the CPU was, and it was very hot. Again, rather surprising.

Finally, testing a slot 1 Pentium 2 233mhz CPU with and without a big case fan propped up next to it, I noticed substantial temperature differences. Without the fan, the CPU quickly climbed over 50C. With the fan, it hovered about the 30C mark. (Out of curiosity, I'd like to test any P2 CPU without its heatsink, but I don't think they can be removed!)