** Still in development. Won’t release STEP files before release **
[Update Sept 28]
I plan to use 3D printing to build a flexible, expandable, and heat-dissipating enclosure for the BPI R4. The goal is to provide good heat dissipation for the maximum load (two base-t sfp modules and wireless nic) through two 4020 fans, unified air flows and heat sinks. If anyone is willing to participate in this project, please tell me.
use the holes from the side on top of the case - the heat is going up. i replaced the top of the original case and now i have no heat problems anymore. guess the best would be to suck the air from bottom and push it out on top
Like your design! I guess it is pragmatic enough: 4 fans (2 left, 2 right) would probably be an overkill - also for power consumption… and of course flow from bottom to top would be better theoretically - but the board itself will not allow that, as it has no major holes, so it anyways would block all bottom to top airflow.
I also like your small form factor, just test it if i provides enough cooling and perhaps consider a variant with/without antenna board and antennas.
This design is not yet complete, my plan is to create a flexible enclosure - I plan to add attachable (maybe secured by screws or clips?) bottom I plan to add attachable (maybe by screws or clips?) antenna housings at the bottom, and manage the cables through these wire clips with small holes. The advantage of this design is that the antenna part is optional, just replace the wire clip and add the antenna housing. In addition, the antenna wire passing through the bottom will make everything easier and reduce interference with the airflow. Regarding the fans, I think four 4020 fans are a bit excessive and there are size considerations in my design, thanks for your suggestions.
I absolutely love it so far. IMO it will probably be OK with both fans on the one side. It is a small board and the fans are focused on the hottest parts aside from CPU, but most people already have an active CPU cooler already. Wont really know if it works well or not until you test it though. 10/10 aesthetic. Please keep us updated with your progress and files!
Why not to design top cover (replacement) for the official case? That will make air in and out.
At least official case is a metallic case and it works like a heatsink for wifi card.
The main problems of official case is that it has lack of ventilation holes for air in and out and one more problem is that it has no wall mounting holes. Even if I have active cooler on cpu heatsink it can not get fresh new air and can not push out hot air out. And it’s just mixing hot air with hot air inside of the case
PS
Here is a quick random possible design … you can put 3 coolers for air exhaust and air holes for for fresh air
I’ve considered that. The reason I eventually chose to do so was because the effect of passive cooling is limited, the official case is also hard to manage the antenna cables. And when the board is installed, it’s not that easy to access the boot select jumpers. Those reasons lead me to design another case.
of curse its up to u but as I already told before WiFi card transfers its heat to the main shell. And for eg in my case CPU temp is almost always lower then WiFi chips. And as for me I do think to change original top cover with own design to improve air exchange inside the case. I do want to make a temp (just use cardboard) top cover with different air holes designs to understand air flow. If it requires to add fans or it will remain temp in reasonable range in passive variant. O maybe use turbo fans (like for eg is used in R3 Mini) to exhaust hot air outside or use fans for air intake
PS
about antennas … it will take u only one time to manage cables … for eg it took me maybe 5 min to place cables inside case. It took me much more to screw SMA sockets as Ive 10 antennas (wifi + 4g modem)
Works like a charm, no heat problems anymore.
The wifi card uses the bottom for passive cooling, the cpu has its own little cooler from sinovoip (guess i could disable that and it still works) and the sfp+ must be cooled, otherwise it gets too hot. Not sure if its somehow possible to attach it more to the case.
With the closed top, even the two fans where not enough because of heat-build up.
I guess with a complete redesign, 4-8 holes in the bottom and 4-8 holes in the top, that should be enough to have a nice little airflow automatically generated under load (heat goes up)
Sorry, I think you underestimate the role of fans and airflows. A single 8k rpm 4020 fan provides more than enough airflow to cool a 60W heat source. This is common in modern consumer and enterprise products - see the design of flex and 1U server PSUs.
I dont underestimate the role of fans. Its just pointless to cool something from 45°C to 30°C when the max. working temperature is 55°C or when the fan’s can not blow out the heat.
But hey, put in as much fans as you want. But they wont do anything better. I prefer a silent system. I have a noctua on the sfp+ and the sinovoip oem solution for the cpu. I have 45°C cpu temperature on 20°C ambient with that open case. When i use the closed case, i get nearly 60°C and the sfp+ stops working - just by closing the top, with fan’s connected. So you can put in a dozen of fans that blow everything out of the little side holes, or use a top cover with small holes and use only 1 or 2 fans.
In any way: the default case from sinovoip is… bad and needs a redesign. So if someone provides some good alternatives, it would help the community and maybe sinovoip produces better cases in the future.
I just want to restate that the whole point of my design is only using two 4020 fans in the case. I expect there should be no CPU, SFP, or NIC fans. All of the rest of the thermal sources have tiny passive heat sinks (made up of copper, aluminium, etc.)
Assume the Noctua A4x20 is used as the 4020 fans in this system (Sunon would even provide almost twice the capability than Noctua if necessary).
According to Banana Pi Wiki, we may assume that the max thermal power is about 60W (DC-in 12V*5.2A=62.4W). The airflow of a single fan is 9.4 m³/h, for dual it would be 18.8m³/h, which is 0.00522m³/s. Also, we may assume the normal environment temperature is 25°C - which may use 1.184kg/m³ as the reference density of air.
Mass flow rate:
Assuming an air temperature rise (ΔT\Delta TΔT) of 20°C:
Specific heat capacity of air (Cp): 1005 J/(kg\cdotpK)
Cooling capacity (Q):
Although these are just some simple calculations, you can see the potential of this design. For 60W of heat dissipation, two fans are ideally sufficient. Even taking into account the effect of resistance, the dual-fan design can provide sufficient cooling capacity while tolerating up to 50% efficiency loss.
The primary reason leading you to believe that the two fans may not be sufficient is due to the limited heat exchange surface area. I recommend using a larger heatsink in this design or cascading multiple heatsinks with copper plates to improve thermal conductivity.
I tested a dozen of heat sinks in many dimensions, none worked because of the missing airflow. I even bought ones that have dimensions 15x10cm with fin height of 6cm (because i thought maybe its possible to cool the case more). Passive cooling is not possible.
Now, i use the sinovoip fan which already has a heatsink - also, i use heatsinks on the sfp port. On both (cpu + sfp) i have a tiny fan, one noctua and one from sinovoip. But even with those AND the heat sinks, it did not work. Now i replaced the top of the case and all of sudden, it works like a charm. It works that well that i believe, replacing the complete oem fan of sinovoip with a bigger heat sink could be enough. The holes on the side (oem case) are just bad because one has to suck in fresh air which leads to a bad mounting position when using regular fans. If you want to use the holes on the side and NOT create new holes at top/bottom, you should probably go with something like on laptops. Otherwise, i highly doubt it works or it must create such big amount of air flow, that its just… stupid because of noise.
edit: just noticed you updated your pictures. that should work as you go exactly the way i described.
But i am still not sure if that would be sufficient to cool sfp+ as this goes crazy hot. I burned my finger on it, so i would say it was about 80°C or even more - with heatsinks.
Continue your effort. If you find anything usable, publish it. I am out of this discussion now.
