Foxpoint logo

Work log 2023-09-03 - Git clean up & fix for depth sensor
2023-09-03

The workshop

What did we do?

Fist thins first, code clean up from last session that ended in the middle of the night. Verified that the balast tanks works via the gound control UI with the SPI-bus implementation. Before we experienced issues with the depth sensor MS5837 and made workarounds for how to work with depth off set. Turns out we were initializing the wrong sensor type in the code, switched to 02BA and it works like a charm. All we need to do now is to store initial air pressure value when we start the depth sensor node and then subtract that value from all coming depth readings.

Finally we made a roadmap on what to do the coming weeks, where propeler control is a priority for us to make diving algorithms and evaluate dynamic diving with the propeler alone. Once logic is in place we need to find a indoor pool preferably longer then 20 meters to do verifications.

Learnings

  • Cleaning is nice, room for more code cleaning session in the future
  • Take time to investigate issues more closley before building work arounds
  • Need to add a DNS server configuration to modem wwan connection

Next up

Coming Sunday we will deep dive in the propeler control logic, there needs to be a PID (P) algorithm controlling the propeler verically where pitch will be the control signal. We then need to take into consideration the yaw that can occure mostly when we turn. Yaw value is provided from the magnificent BNO055, with this value we then need to compensate the propeller position accordingly. More on this next week.