#robotics Logs

Jun 02 2020

11:13 AM veverak: hmm
11:19 AM veverak: how to cool down raspberry in waterproof box?
11:19 AM veverak: (Box is DIY and we can choose materials)
11:20 AM polprog: maybe a heatpipe to outside?
11:24 AM veverak: that's problematic for mounting
11:24 AM veverak: I was thinking that even in closed box
11:24 AM veverak: one 40mm FAN could make all the difference if the box is from metal
07:42 PM mrdata: veverak, surface area is your friend. how flat is the box?
07:44 PM mrdata: and is this water proof because it will be immersed in water? if so, the water is an excellent heat sink
07:44 PM mrdata: internal fan will only circulate the warm air inside
07:45 PM mrdata: so i'm thinking: a flat box with outside fins will increase that surface area
07:46 PM Tom_L: dunno about inside a box but this keeps it around 49c under loads: https://www.amazon.com/Metal-Raspberry-Heatsink-Thermal-Adhesive/dp/B07VMD25SK
07:46 PM Tom_L: no fan
07:46 PM mrdata: if it's just the cpu that gets hot, then your thermal management can focus on that
07:47 PM mrdata: but perhaps you have motors to worry about
07:47 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/cnc/Rpi4/Rpi4_bottom.jpg
07:47 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/cnc/Rpi4/Rpi4_top.jpg
07:48 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/cnc/Rpi4/Case_top1.jpg
07:48 PM Tom_L: bottom side of top
07:49 PM mrdata: did you know that with satellite technology, half of all the engineering effort is heat management
08:00 PM mrdata: as you know, metal, especially copper and aluminum, conducts heat much better than plastic. https://www.engineeringtoolbox.com/thermal-conductivity-d_429.html
08:01 PM mrdata: copper is 2000x better than acrylic
08:06 PM mrdata: and 15,000x better than air
08:06 PM mrdata: with air, convection is your friend
08:22 PM mrdata: i had to think of this when i made lift-lock control boards for trucks
08:26 PM mrdata: https://transfanon.fandom.com/wiki/Long_Haul_(TTM)
08:27 PM mrdata: ^ almost like this