Raspberry Pi fanless case with good thermals

[Will]

When I migrated from A RPi2 running OH2 to a RPi4 running OH4, I started looking for a case.
I wanted it fanless.

I found various recommended models I did not go for and decided to test one that looked interesting due to the design of the multiple heatsinks:

I got it from Amazon for 11.99 € which I find really good for the quality.

I even wrote a review because I found this enclosure really nice. Here is an extract:

First impression was: This is heavy !
Sure this is not copper but this is far from cheap either.

The case is Manufactured very clean. It is nice to see the accessories you get:

• 2x sets of thermal pads
• 1x thermal plate
• 4x screws for the box
• 1x extra screw and the wrench for them

This case is nice because it does connect all the chips that need dissipation.
I did NOT use a thermal pad for the cpu but the thermal plate and thermal paste on both side.

When assembing the box, make sure to remove the protective films on BOTH sides of the thermal pads… I almost forgot despite the little cards mentioning it.

The dissipators appears to connect properly, at least I could check with the cpu and there is a good adhesion.

3/4 screws went in like butter. The last one however was a little off.
When assembling the case, I would suggest screwing FIRST the screw next to the output output. This one was a little off for me but it worked out.

The case is clean and looks nice. The sd card is ourt of reach and flush with the case.

Would I recommend that case ? 100%
For the price of 12 € ? 200 % !

I am running this case for a while now and I have more data to share.
Here is almost 30d of data:

You can see:

• RPi temperature
• load
• indoor temperature
• outdoor temperature

The case keeps it all running at a steady ~40°C.
The bump around April 23rd is when I moved the RPi+case from my (low altitude :)) cool desk up to the shelf close to the celing, with quite some racks / switches and NAS (it is hotter there).

[binderth (Thomas Binder)]

@Will

First impression was: This is heavy !
Sure this is not copper but this is far from cheap either.

I also have that same casing and it rocks! at least for Raspi4. it really cools down efficiently and you don’t need a fan anymore (my fans regularly died on me, my cellar is not dust-free I guess! ).

[splatch (Łukasz Dywicki)]

Can you check how it works under sustained 100% load? I remember my plays with various radiators and found out that, with 100% load, some of them start to loose their ability to cool down whole thing. Sure, we do not run 100% load all the time, but that’s how you can measure if passive cooling is actually working or you just rely on thermal capacity of case.
I did tests with other device using sbc-bench.

Below graph with cpu speed (100% means no thermal throttling), %cpu is load and yellow is temperature.

It was 35 mm heat sink with thermal pad (no paste), put on my desk. I might look for picture how it looked.

Example with 19 mm hat sink which didn’t pass the bar:

You can see on above picture that around 17:16 (less than 10 minutes from test start) system began to drop cpu speed and hit thermal throttling. It didn’t throttle much, it was throwing at most 25% of cpu frequency.

[binderth (Thomas Binder)]

@splatch

Can you check how it works under sustained 100% load?

I had to take my Raspi4 out of openHAB for that tests, so I’m afraid I can’t do that presently.
FWIW, this is my typical load of my openHAB Raspi4, 4GB with only openHAB4 running on it with around 1200 items, around 100 of them changing every second at least and subsequently a sh…load of rules as I’m running openHAB both an an EMS and “normal” smarthome.
shows only what you said: the CPU utilization is around 10% with some peaks.

I’m a bit perplexed, that Zabbix Agent2 won’t cover CPU temperature, but:

~ $ vcgencmd measure_temp
temp=40.4'C