I finished building the 3DR quadcopter about a week ago and encountered several issues in the assembly documentation, so I figured I would post my issues and recommendations here with the hope that they could be incorporated into the official documentation.
The main problem regarded the connection of the motors/ESCs. Based on tracing the signals through, testing, and finally flying, I claim that the instructions are flat out wrong and the copter will not fly based on the documentation. Phasing is always an issue in control systems and the arducopter is no exception. This is an area where extreme care should be taken during the initial assembly, and then extensive testing should be performed to verify the assembly. The ground testing worked very well to verify the system, but the assembly documentation basically didn't address this at all.
When all was said and done, channels 1 and 2 were swapped compared to the documentation, so the PWM channels have to cross the PDB This seems to be due to the direction the PDB is installed (down vs. up).
Rather than trying to explain exactly where the issues were, I would suggest the following:
- Add silk screen markings for each channel (1-4) to the PDB where the 4-wire harness is connected and where each channel goes to the ESCs
- After the user installs the PDB onto the baseplate, have them write on the baseplate which channel is which. If they don't do this, it's not obvious which channel each ESC should be connected to.
- When connecting the arms, have the user mark which channel each arm represents. This way, it is easy to connect motor 3 to channel 3 on the PDB.
- Add a note that says the ESC PWM cables should be installed with the black/ground wires towards the center of the frame. It is not possible to see which direction they should go once the PDB is installed inside the frame.
- Add a note to ensure correct polarity of the 4 wire harness from the PDB to the APM. After it is properly connected, put some white-out across the connectors to help with plugging it in in the future.
- Ensure that the polarity is properly described and maintained for the 2 wire harness for the PDB. Also use red/black for the wires to simplify connections
- Provide mounting holes in the 3DR plate that actually line up with the APM. The diagram in the instructions is incorrect. It is not possible to mount using all 3 APM holes in the "+" orientation.
- Provide better documentation regarding ESD safety. I'm surprised that more boards have not been fried due to non-chalance regarding ESD.
- ESCs shouldn't be zip-tied until all of the ground testing is complete, since signals may have to be swapped.
- Pitch direction of the transmitter is swapped in the ArdupilotMegaPlanner software. There is no documentation that says the GUI should move opposite the stick movement, but that's what I read in the forums and that's what I had to do to get it working.
- The testing that involves listening to the motor speeds with the props off is very difficult, since its tricky to hear small changes in motor speeds when they are all running. I wound up bailing on that test and moved on to the equivalent tests with the props and and was immediately able to verify the directions of all the axes.