Date
Attendees
- Ben Hyman (RIT Student)
- Billy Siegener (RIT Student) (bill)
- Dante Sivo (RIT Student)
- Yevgeniy Gorbachev (RIT Student) (gorb)
- Chloe Clark (RIT Student)
- Aaron Chan (RIT Student) - grafana expert
- Kenny Kim (RIT Student)
- Will Merges (RIT Student)
- James May (RIT Student)
- caw5317
- Alec Paul
- Jonathan Russo (RIT Student)
- Paige Elias (RIT Student)
- Adam Lehman
- @joe
Goals
- Lay out the electrical and physical standards for a modular flight computer
- The modules should be electrically independent
Advantages
- HITL/SITL testing
- Breaking out work - fewer people per board, each individual board is simpler
- Modular
Disadvantages
- Each board will need an Ethernet chip and microcontroller, adds electrical complexity
- Physical connectors are physical failure points
- We have way more hardware people than software people
- Mitigated by HAL (will is this right?)
- Requirements can change enough to obsolete parts of the architecture, requiring significant work
Discussion Topics
Protocol
Most protocols other than Ethernet and CAN (I2C, UART, SPI) would reduce to the same thing with extra work or would be too slow.
- need central shared bus
- any protocol with a master/slave structure is disadvantageous as it breaks "functional independence"
- speeeeeeed
Board Separation
- 70cm RF is on the GPS board, not the telemetry board, because that's the only use case for 70cm RF
- Battery board is separate because it's a large failure point and should be very customizable
- Logging board should listen to literally every other board
Connectors
- Ribbon cables may make mechanical integration easier
Danger will robinson
- scope creep can get to us
- this could get very large and very complicated very quickly
- important to be diligent in designing for now, not digging ourselves into holes, not making unnecessary modules, etc.
Action items
| Time | Item | Who | Notes |
|---|---|---|---|