Design Decisions

• MPPT for each panel with a stacked board design so 4 MPPTs in parallel to each other

Some Things

• Batteries have a very set range of operating temperatures, so when the battery is mounted to the chassis, we need to prevent the heat transfer through conduction heating up the battery

Post-Comp Additions

Stuff that went wrong/right

• At least the boards came on time?
• Boards fit in the chassis + we only lost 3 solar cells
• ^Kind of, we had to sand down BMS and we want to avoid doing that again
• EPS layout needed (and still needs) a lot of work to function. Need to really focus on spacing and part placement next time
• Batteries lighting on fire was not good. We need to do a better job of researching and verifying the safety of the parts we use, it was all a little last minute this time
• Definitely need to invest more time into eps firmware next time cause that really hindered us from getting a demo working, I think the new team structure will help a lot with this
• We pushed deadlines a lot, there needs to be a superrrr conscious effort to make reasonable ones we can stick to next time- budgeting enough time for assembly, testing, multiple board versions, etc

Technical changes that should be made

• Encouraging ongoing ordering and testing of breakout boards would have been better than what we did, it’s relatively inexpensive in comparison to the funding we have, so not being afraid to make multiple versions as opposed to trying to make it perfect in the first go would be beneficial
• Involve more members in board review processes. Because it was close to comp we mainly had advisors and leads leave comments, but encouraging reviews from every hardware member in the future is great practice for them
• Side idea: It might help for leads who worked on the hardware for this comp to do ‘board workshops’ and film them, just going through and explaining all the decisions that were made/changes to come might be helpful and allows for new leads to ask questions in real time which wouldnt happen even with detailed documentation. Also new members are probably more likely to watch a video to understand a board than read through really long documentation.
• We should do FW and HW simultaneously next time (we did not, it was problematic) - and the FW plan should be super tight early on, we made a lot of late stage changes because of extra GPIO’s (need to keep track of these-may need extender), current sensing problems etc.. not ideal
• On the same note all important chips need to be picked out carefully so that they’re compatible, avoiding the i2c to 1-wire bridge in the future would be good, finding a compatible CAN chip for next time etc.
• Another similar note, firmware frameworks are also important. Keeping Zephyr and the blackpill limited us quite a bit so we will abandon that for next comp