Skip to content

v2.0.0: 2024 Post Comp Release

Latest
Compare
Choose a tag to compare
@AarjavJain101 AarjavJain101 released this 08 Aug 06:09
· 43 commits to master since this release
96d53a8

What's New

This release marks the final changes on the car when we drove in production at FSGP and ASC during the summer of 2024. The changes reflect necessary adjustments to meet scrutineering requirements as well as to improve driver QoL in relation to accelerating.

Below are the specifics of this release

  • Our previous over temperature thresholds were faults at 45C for charging and 65C for discharging and a trip at 45C. However, the scrutineers require us to fault every time we exceed a manufacturers threshold (so the trip at 45C is a no go and since the manufacturer specifies that 65C is the limit for discharge we need to add a fault state for that as well). Thus, we added logic to first check if we are charging or not and then from there check for the appropriate thresholds. See PR #151 for more details.
  • ADC noise on the pedal spiked just under 4000 at times which meant that the previous MAX_THROTTLE threshold would be passed and resulted in a fault message but more importantly throttle = 0.0 would be sent to the MDI and then to the MC. PR #151 fixes this issue by setting the limit to 4000 (still some room to distinguish a short circuit which would be closer to 4095).
  • PR #153 fixes an issue where the ECU triggered a fault yet no LED or telemetry information indicated this. We discovered that the issue is because noise is generated on the fault GPIO pins from the master board to the ECU (only at high currents with the solar arrays plugged into the battery). To solve this problem we took advice from a Scrutineer to use a spare GPIO (in our case the BAL GPIO) and set its value high initially. Then while the system runs, and there is high irradiance, if we actually get a BMS fault then the FLT pin is set to high and the BAL pin is set to low. If and only if this condition is met do we consider it a fault. Otherwise, if the BAL pin still showed high (due to noise) then we do not consider that a fault occurred. Hence we essentially added simple logic to do a checksum on our FLT data. Note that we do not send COMM or OT FLT info to the ECU because the FLT GPIO already encapsulates these. See PR #153 for more.

What's Changed

Full Changelog: v1.1.0...v2.0.0