From model to reality:
Originally it was designed all to be inline to be inserted into aluminum under water, but for the box these two halves are folded.
Because it was supposed to fit inside tight space, the model walls are especially thin. And used the copper strips instead of wires to keep width and height small as possible.
Under the petg end cap there is a 200A fuse and pico fuses for 14 balance wires.
Each bank ~20mohm
Total for 12s3p batter 40mohm
Each battery ~9mohm
each connection ~0.9mohm
Comparison of 0.15X8mm pure nickel resistance vs same length copper 0.3X10mm
from battery to battery:
Used Fnirsi SWM-10. Conclusion - these kinds of products suit only non pure nickel strips.
My conlusion for next time - to solder copper strips by making holes in the strips first (and freezing the cells for any case).
Charger - ToolkitRC M6D - charging dual at 9A, excellent charger till far. Make sure to calibrate it, and then you will get accurcy of < 0.01v. Even after discharge most of the cells within 0.001v.
I recommend this internal reistance instrument - YR1030
Capacity:
12s3p of Molicell P42A
460Wh of actual in the water till AFAIR 2.9v
Potting compound sealing one end
The other end additionally marine silicone around the wires. Underneath these is a ring holding wires together. I should have instead sealed from inside the petg cap and made the wall around wires larger
Message me if you want to receive a link to the Onshape design. It is parametric - for example I generated the 32700 holder
Model of the holder: