Reserved for co-op mentor Daniel Puratich and co-op
past references
braintstorming
Combine what autonomy wants onto the zp4 board. If auto wants a full raspberry pi on there we should get the chips. Or if we sticking with lte hat on pi, should zp4 be an rpi hat for ardupilot essentially? Does that exist? putting a full rpi on there would be epic.
...
Table of Contents
| Table of Contents | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
The Big Idea
Copy the Pixhawk’s functionality except:
a device with more compute for autonomy to do airside processing (an RPi 5)
support more CAN transceivers for our push to CAN infrastructure and remove all extra unneeded connectors
Single Stack for the drone’s LV processing system to minimize assembly effort
flight controller would be a RPi hat
stack would be: RPI5 + LTE hat + anything we want + zp4
Leave the HV stuff separate from this stack
this allows for us to do weight balance and wire length optimizations still and keeps our system somewhat flexible.
This project is far out (as of May 2024)
We should wait for this to become more necessary as time keeps slipping into the future.
If EE has available to start now (or as soon as possible) we should because this is a large endeavor
As time goes forward we will get more new features and ideas so no point in the huge rush.
Add more CAN transcievers
ensure we have at least four? keeps the datarate high enough for our new CAN infrastructure, pixhawk only has two CAN transcieverss
Optimize for our platform
from a high level this has all been done before, we need to optimize for our platform once we define our system architecture
References
Development
Implementation Details
We need to be extremely cautious on the mounting of the IMU and barometer ICs as they are in the pixhawk
We should pick the most powerful ardupilot supported microcontroller from the list (STM32H755 or something).