Date: Sun, 6 Nov 2005 23:00:32 -0600 (CST)
Subject: firmware development done
X-UID: 100

Firmware development is done. Now I have to draw up a test plan for
quality assurance. I'll do the QA tomorrow night. Really, this is a final
check to make sure that everything still works. After that, the
microcontroller will be inserted in the electronics board and this checked
out. After that, the entire electronics enclosure will be reassembled.

I'll publish everything once the robot is together and rolling. It should 
be useful as it captures in one place many problems I've seen scattered 
around the Internet in fragments.

I'm convinced now that C is good enough for this kind of development 
unless memory and timing constraints are absolutely critical. Much 
published sample code is assembly. It doesn't have to be. C will do the 
job as is much easier to read.

Also, piezo gyros are very cool things. I'm using a Futaba GY240 rate 
gyro. It is designed for helicopters. I find that the readings I get from 
it are pretty stable. The device has a natural period of 15 ms between 
output pulses but can be driven at a higher frequency, up to 100 to 120 
Hz. That's what I'm doing. The output pulse is always between 1 and 2 ms 
in width. Trivia fact - I read that Nikon optically stabilized lenses have 
a natural frequency of 20 Hz, corresponding to the human body.