Update the code to handle the new board

[GoogleCodeExporter]

We'll need to at least update the following:

 - The pin mappings will need to depend on the board type

 - radioIn.c will need to be updated to offer a larger number of real PWM inputs when available

 - servoOut.c will need to be updated to allow outputting to a larger number of real PWM outputs 
when available

 - no hardware switches on the new board

 - other things I'm forgetting?

Original issue reported on code.google.com by benjie on 20 Apr 2010 at 11:40

[GoogleCodeExporter]

Bill I sent this one to you just because I thought you'd want it.  :)  I'm 
planning to work on this one too.  

One note: In radioIn.c, I think we could have all the pwm input interrupts call 
one common function to share 
some code...

Original comment by benjie on 21 Apr 2010 at 1:29

[GoogleCodeExporter]

I think we need to restructure the code for the new clock speed.

When we do that, let's make it so that all the timing parameters automatically
accomodate any changes in clock speed. I have seen some good example of this in 
the
Microchip documentation and am happy to find them again when the time comes.

Original comment by peter.ho...@gmail.com on 21 Apr 2010 at 1:47

[GoogleCodeExporter]

Original comment by wpremerl...@gmail.com on 24 Apr 2010 at 1:56

• Changed state: Accepted

[GoogleCodeExporter]

I suspect that we will need to reconfigure some of the code to be a shared 
library of
functions between two quiet different versions of MatrixPilot because the new 
board
is going to allow new features that use up more RAM than the old board can 
support.

We will probably quickly want to move away from waypoints.h towards a flight 
control
language. And the new board could easily support real-time updates a) of Control
Gains in a special test flight option and b) real-time updates from the ground 
of new
waypoints, and some other flight variables (e.g. control gains again). These 
types of
features are unlikely to be supported by the old board RAM. So we will need to 
think
carefully about the source code architecture.

Original comment by peter.ho...@gmail.com on 12 May 2010 at 4:11

[GoogleCodeExporter]

The basics for this one are mostly complete.  Still outstanding are UDB4 
magnetometer code and UDB4 OSD code.  I think everything else works for the new 
board.

I think we should continue to use the same code base for all boards, and add 
new UDB4-only features, flight-languages, etc. as new files that only get built 
as part of a UDB4 build.  But if we can continue to share most of the code, 
then users of the old board will get the core improvements and bug fixes.

If we want to change how the whole core of the UDB4 DCM code works, then it 
might be time to rethink this... :)

Original comment by benjie on 30 Oct 2010 at 8:20

[GoogleCodeExporter]

Team,

When we get the UDB4 into production, I would like to start with a fresh piece 
of paper on the firmware. There are several things I would like to do which 
would not fit very well with the present structure:

1. Raise the clock to maximum rate.
2. Change the way the DCM is computed to either add 2nd order Taylor terms, or 
use the exact equation instead of the one that we are using. This is a must 
before we can go to the next higher performance level.
3. Increase the frame rate of the DCM calculation.
4. Support digital servos.
5. Consider using floating point throughout.
6. Rudimentary operating system.

The changes that I am considering would not work on the UDB3. So, we would be 
"finalizing" MatrixPilot version for the UDB3.

Original comment by wpremerl...@gmail.com on 30 Oct 2010 at 9:01

[GoogleCodeExporter]

Original comment by benjie on 1 Mar 2011 at 1:30

• Added labels: Milestone-Release3.1
• Removed labels: Milestone-Release3.0

[GoogleCodeExporter]

Original comment by peter.ho...@gmail.com on 1 Aug 2013 at 11:02

• Changed state: Started

[GoogleCodeExporter]

Changed to fixed.

Moved Bill's wishlist to separate items

Original comment by uavflightdirector on 3 Aug 2013 at 6:06

• Changed state: Fixed