yoos / tricopter Goto Github PK

The rotor cylinder is rotated around the global Y axis regardless of what the DCM looks like.

Accidentally started to plug in the power to the electronics via the wattmeter today. Big spark made me stop and pull back, but this makes me wonder about three things:

1. Does the wattmeter have some sort of short circuiting to protect the important electronics on its output lines?
2. Is this what happened last year when 955 was using my wattmeter?
3. What is the reverse-voltage behavior of the ESCs? Do they let any current through?

Miscalculation. Tail arm is longer due to rotating mount. This affects balance and control.

Running make in flightcode with a default install of avr-libc returns the following error:

/usr/lib/gcc/avr/4.5.1/../../../../avr/include/avr/pgmspace.h:944:38: error: 'uint_farptr_t' has not been declared

I've been editing pgmspace.h to #include "../inttypes.h" instead of #include <inttypes.h>, but this is a hack. There is probably a Makefile configuration I'm missing.

Currently, I've an arbitrary scale factor hard-coded into itg3200.cpp. This needs to be figured out, as it is contributing to DCM drift.

Accelerometer weight should change based on how much total acceleration the tricopter is subjected to. AeroQuad seems to have a simple one that does this.

Can I do this and still use rxplot on it?

For a correctional rotation vector, the accelerometer is responsible for the global X and Y components, and the magnetometer is responsible for the global Z component.

Does this hint at data corruption elsewhere?

For example, the tail rotor tilt PID should depend on the current throttle level because the vector component magnitudes change.

Serial output on laptop is nice, but I'd like for everything to be also dumped in a text file to be plotted later, if wished.

There should be separate text files for each run instance, perhaps named according to date and time.

Make it possible to read from /dev/js0 instead of (and in addition to?) Arduino serial stream.