[simpits-tech] Airspeed (and air-core!) indicator success

[dabigboy at cox.net]

This week I finally decided to quit putting off the airspeed indicator project.....to great success!

http://www.youtube.com/watch?v=h472mXtVmkg

This one is from a Jetstream 31, and came out really nice. I'm also very encouraged about air-core motors.....huge potential here! The ONLY issues I am concerned about are:

1: Limited torque, though this can be aided somewhat by putting higher voltages to the motor (running 12v right now, but the L293D can take up to 36). This is only really a concern if driving heavy/draggy instruments, like artificial horizons and altimeters (I think you could totally do a real altimeter with one of these).

2: Slight "hang" every time you switch quadrants (every 90*, basically). I don't think this is a switching-speed issue with the L293D, I think it's just a natural physical property of the air core motor that will have to be dealt with in the code that drives it. But the "snag" is so minor that it's almost unnoticeable. Pretty sure I can fix it in code anyway. After all, the motors don't do it when they're reading RPM in your car, so it's obviously possible to fix. :)

It also seems like the needle position is a little inconsistent at times, but this got a lot better when I switched from 5v to 12v. At any rate, I'm now usually no more than 3 or 4 knots off with the airspeed indicator at most, and usually much less (many of my sampling points are dead-on).

I love the fact that you get 1024 steps per rotation with this setup...since the Arduino only drives the motor through 90* at a time. And since the motor is an analog device, you can actually get as much precision as your controller provides. There is ZERO "notchy" motion or visible steps like what you see with servos or (especially) stepper motors (although I realize steppers are typically geared way down, which makes them a lot more viable as precision instrument drivers). Even the digital nature of driving the motor with a microcontroller is smoothed out somewhat by the way the motor moves.

Matt