/tech/ - Soyence and Technology

>>31994 (OP)
obviously they wouldnt release the proper designs, thats why they had to make the open source shit slightly worse
>3d scanner
too inaccurate

>too inaccurate
Meds, even cheap ones can do 10 microns

>>31994 (OP)
I tried to 3D print a fan way back when, but the issue is tolerances and getting a good motor and ball bearings.

Plus you have to sand the damn thing in order for it to look decent.

>>32082
still have the CAD files if anyone is interested

>>32082
What fan did you base the blades and housing off of and what fan did you use motor assembly from? I would think it would be hard to find ones that aren't too shitty or too expensive to justify hacking into a new fan.
I think noctua probably sinks most of their cost into high quality bearings + motors and a plastic that let's them get really tight tolerances. Probably difficult to make on your own.

>>32095
The best I could probably think of is using drone motors and something stiff like nylon. I'd say do resin for the surface quality but its too weak

>>32083
Yes

Does adding a ring fixed to the outside of the blades kill performance or something? It would fix tolerance

I don't understand why computer fans have hubs. You're very obviously fitting a circle in a square perimeter. At least 21.5% of the fan is wasted space, but yet even more space is carved out to fit a motor in the middle of it. Why not do this but power it with some rollers in the corners?

>>32891
Because that would lead to greater friction, so reduced efficiency and more noise. Thicker fans are the way to go and only 4reddit Noctua fanboys are mad about Phanteks making a better fan simply by making it 5mm thicker.

>>32097
Here. Original file extension is .stl but the sharty wont allow me to upload .stl files.

>>31994 (OP)
I have access to some industrial 3D scanners at work (actually accurate enough), but I don't have any noctua fans. Which ones did you want scanning?