Week Review 2/15/2024

Making the Float PCB

This week I finally milled the PCB for the float. My first attempt worked with the small problem that the mounting holes were slightly too far apart. After I remilled the PCB everything seemed to work, but when I plugged in the cable to drive the motor and give the 12v of power the profiling feature stopped working.

Milling the PCB was more of a challenge then I was anticipating. KiCad does not output .brd files, and nobody in the shop has used anything other than .brd. Thankfully Henrik in programming had used KiCad to mill PCBs at Artisans Asylum using their PCB mills. He showed me how to use Gerber files and Drill files.

You first import the front copper Gerber file, then you import the back copper, the outline, and the drill holes. It is important that the drill file is not Gerber. I do not love the Bantam Tools PCB milling software, but it does get the job done. Single sided PCBs have been outdated for decades, but it is the best you can reasonably do with desktop PCB mills.


Ben has made significant progress on the enclosure of the float and sealing the antenna. Although it is not a guarantee we are hoping to get the float in the pool after break.



I made the assumption that somehow the signal pin on the hall effect sensor signal pin was being tied to ground somehow, but some investigation showed that the connection between the ground pin and signal were not continuous. I isolated the variable of the connector by supplying 12v from my bench supply to the proper pins and measuring with a multimeter from the other pins.

Under these conditions the PCB worked without any problems. I then took a closer look at the connector, and after removing the heat shrink I noticed there was a short in the cable.

That allowed me to piece together what was happening. When a profile started there was a short circuit between 12v and GND. This caused the voltage to drop in the battery pack. The low voltage caused the MCU to crash, stopping the flow of electricity to the short.