08-10-2016, 04:45 PM
(This post was last modified: 08-10-2016, 04:45 PM by Beta_Ravener.)
Hi,
I have the board for little longer and didn't have temperature problems so far doing only basic stuff (the highest I got was during some compilation), but your post inspired me to give my board a heat-sink today.
I looked around and found a little heat-sink that was precisely 1.5cm wide, just like the CPU. I cut it in the length so I would have 1.5x1.5cm base that perfectly fits the chip. Then I used a thermal pad instead of paste to keep things clean. It's true that the thermal properties are little worse but not much, and the pad is a little bit sticky so it helps to keep things in place. It looks like this:
![[Image: IMG_20160810_234957.jpg]](https://s10.postimg.org/7a07lmbud/IMG_20160810_234957.jpg)
![[Image: IMG_20160810_234602.jpg]](https://s10.postimg.org/h9v41in3p/IMG_20160810_234602.jpg)
![[Image: IMG_20160810_234509.jpg]](https://s10.postimg.org/nou4y6tth/IMG_20160810_234509.jpg)
With everything in place I ran the stress utility and the temperature started to rise from base temperature of 40C to 50C and then slowly (1 degree each 10 seconds) up to 70C where it settled down. Clearly, the heat-sink itself doesn't have enough cooling capacity to radiate all the heat away, but this was somewhat expected with such a small form. However, I'm impressed with the thermal pad which seems to nicely conduct the heat from the CPU (I didn't measure heat-sink temperature but it got pretty hot. Maybe next time I'll try to measure it.)
When the temperature rise stalled, I tried using fan (the big one for room use because I didn't have anything else to test with at hand) from fair distance of ~50cm at low speed, which created a light breeze. The temperature dropped to 50C in half a minute and continued to drop slowly. If I ever had problems with temperature again, I would probably mount a small 5V fan and let it run if the temp gets too high.
After killing the stress utility, the temperature immediately drops from 70C to 60C and after 5 seconds is at 50C. Just as a final note, the heat-sink is really good thing to have. I first tried the stress utility without it just for reference and the temperature quickly rose up to 80C where the board shut down.
I have the board for little longer and didn't have temperature problems so far doing only basic stuff (the highest I got was during some compilation), but your post inspired me to give my board a heat-sink today.
I looked around and found a little heat-sink that was precisely 1.5cm wide, just like the CPU. I cut it in the length so I would have 1.5x1.5cm base that perfectly fits the chip. Then I used a thermal pad instead of paste to keep things clean. It's true that the thermal properties are little worse but not much, and the pad is a little bit sticky so it helps to keep things in place. It looks like this:
With everything in place I ran the stress utility and the temperature started to rise from base temperature of 40C to 50C and then slowly (1 degree each 10 seconds) up to 70C where it settled down. Clearly, the heat-sink itself doesn't have enough cooling capacity to radiate all the heat away, but this was somewhat expected with such a small form. However, I'm impressed with the thermal pad which seems to nicely conduct the heat from the CPU (I didn't measure heat-sink temperature but it got pretty hot. Maybe next time I'll try to measure it.)
When the temperature rise stalled, I tried using fan (the big one for room use because I didn't have anything else to test with at hand) from fair distance of ~50cm at low speed, which created a light breeze. The temperature dropped to 50C in half a minute and continued to drop slowly. If I ever had problems with temperature again, I would probably mount a small 5V fan and let it run if the temp gets too high.
After killing the stress utility, the temperature immediately drops from 70C to 60C and after 5 seconds is at 50C. Just as a final note, the heat-sink is really good thing to have. I first tried the stress utility without it just for reference and the temperature quickly rose up to 80C where the board shut down.