Here's why you should raise your case
23.3.2020
Translation: machine translated
Modern enclosures also draw in fresh air from below. Depending on their height, they can draw in more or less air. How high should your box be to draw in the optimum amount of air?
"Maybe in the future I'll put my feet up higher for better ventilation," I thought to myself as I tested the Fractal Design Meshify S2.
With its feet, the case is only 17 millimetres off the ground. This already seemed low to me during the test. Now I want to know exactly whether this is enough for optimum airflow.
Test parameters
Although the setting for my test is very specific, I consider it made for such a test. At the bottom of the case, I attached a 280mm liquid cooling radiator. Two 140 mm fans draw air through the radiator into the case - they are therefore mounted in a so-called 'pull' configuration. This constellation makes it difficult for the fans to draw in fresh air. In most other cases and configurations, the fans will find it easier to draw air into the enclosure, as they will have less resistance.
Here are the components for the test:
The system is cooled by a custom Corsair Hydro X series liquid cooling radiator. Two LL140 Corsair fans are mounted at the front. The 280 radiator also has two LL140s. The 360 radiator above is vented by three Thermaltake Riing.
For the test, I run the Heavy Load and Fur Mark test programs. Heavy Load requires all processor cores to be running and Fur Mark pushes the graphics card to its performance limits. I set the Corsair fans to 'Extreme' in iCue, Corsair's proprietary component control software, and leave the Thermaltake fans on the standard bios profile, as they are responsible for the top outputs. The Corsair fans run between 1350 and 1500 rpm.
I run the stress tests for 20 minutes and take the temperature of the CPU, GPU, motherboard, southbridge and coolant before the test and then every five minutes. I do this with HWMonitor from CPUID.
I do the whole thing several times. First with the PC's feet 17 millimetres above the floor, then by raising the PC in 5.5mm increments. To do this, I use carambola counters, which I place under the case feet.
The results
During the tests, it's around 23°C in my office. The PC is running all day and the 3D printer is currently running at full speed, hence the high temperatures. As the tests increase the temperature further, I open the window after each test, so that the room is 23°C again.
I first do the test with my feet 17 millimetres off the floor. Here are the values obtained:
| Temps | Température CPU | Température GPU | Température liquide de refroidissement | Température carte mère | Température southbridge |
|---|---|---|---|---|---|
| Avant le test | 37 | 28 | 27.82 | 32 | 64 |
| 5 minutes | 67 | 46 | 32.41 | 54 | 65 |
| 10 minutes | 69 | 48 | 34.73 | 56 | 67 |
| 15 minutes | 69 | 49 | 35.88 | 57 | 68 |
| 20 minutes | 70 | 49 | 36.36 | 57 | 70 |
It's time to place a token of carambola under each foot and repeat the test at 22.5 millimetres above the ground.
| Temps | Température CPU | Température GPU | Température liquide de refroidissement | Température carte mère | Température southbridge |
|---|---|---|---|---|---|
| Avant le test | 38 | 29 | 28.85 | 33 | 65 |
| 5 minutes | 68 | 47 | 33.3 | 55 | 66 |
| 10 minutes | 69 | 48 | 35.29 | 56 | 68 |
| 15 minutes | 69 | 49 | 35.88 | 57 | 69 |
| 20 minutes | 70 | 49 | 36.51 | 57 | 70 |
Not much has changed. The values are more or less the same as in the previous test. Sometimes even a little worse. This may be because the coolant temperature was slightly lower at the start of the test. I put another carambola stone under each foot. The case is now 28 millimetres off the ground.
| Mesure | Température CPU | Température GPU | Température liquide de refroidissement | Température carte mère | Température southbridge |
|---|---|---|---|---|---|
| Avant le test | 37 | 29 | 28.55 | 33 | 64 |
| 5 minutes | 67 | 46 | 32.58 | 54 | 64 |
| 10 minutes | 68 | 47 | 34.17 | 55 | 67 |
| 15 minutes | 68 | 48 | 34.43 | 56 | 68 |
| 20 minutes | 68 | 47 | 34.61 | 56 | 69 |
It's already better. After 20 minutes of testing, CPU and GPU temperatures are two degrees lower. The coolant temperature has also dropped by 1.9°C. I'll add one more counter: 33.5mm above the floor.
| Mesure | Température CPU | Température GPU | Température liquide de refroidissement | Température carte mère | Température southbridge |
|---|---|---|---|---|---|
| Avant le test | 38 | 28 | 28.27 | 32 | 64 |
| 5 minutes | 67 | 46 | 32.51 | 54 | 65 |
| 10 minutes | 68 | 48 | 34.63 | 56 | 67 |
| 15 minutes | 68 | 48 | 34.73 | 56 | 69 |
| 20 minutes | 68 | 48 | 34.78 | 56 | 69 |
The CPU and GPU temperatures are now 1° Celsius higher. This difference only became apparent in the last five minutes of the test. The coolant temperature has also increased. The difference is only 0.15° Celsius. So the fact that the CPU and GPU are hotter than before may also be a coincidence. I put a fourth stone under the case's feet. Now the Meshify S2 is 39 millimetres above the ground.
| Mesure | Température CPU | Température GPU | Température liquide de refroidissement | Température carte mère | Température southbridge |
|---|---|---|---|---|---|
| Avant le test | 37 | 28 | 27.82 | 32 | 64 |
| 5 minutes | 67 | 46 | 32.41 | 54 | 65 |
| 10 minutes | 69 | 48 | 34.73 | 56 | 67 |
| 15 minutes | 69 | 49 | 35.88 | 57 | 68 |
| 20 minutes | 70 | 49 | 36.36 | 57 | 70 |
It is now clear that going higher is not useful. The temperatures for four carom counters are similar to those with two counters. So according to my tests the "sweet spot" of the case would be 28 millimetres above the floor.
Just for fun, I repeat the test on four stones without the bottom filter. So, after twenty minutes, the CPU, GPU, motherboard and southbridge values are the same as with two stones. However, the coolant temperature is 34.27° Celsius and is therefore the lowest in this constellation.
The higher the better - up to a point
I should have guessed: at some point, there's no point in lifting the case any higher. But I didn't expect an extra 11 millimetres to be necessary. In my opinion, this is a design flaw by Fractal Design.
What does this mean for you? I have up to two degrees less on the CPU and GPU side. As mentioned at the beginning, my fans have a very hard time getting cool air into the case, because in a pull configuration they have to draw air in through the radiator first. If you don't have a radiator, there can be a huge difference in temperature if you place your case higher.
Fancy trying it out for yourself? Use my methodology and share your results in the comments.
Kevin Hofer
Senior Editor
kevin.hofer@digitecgalaxus.chFrom big data to big brother, Cyborgs to Sci-Fi. All aspects of technology and society fascinate me.
