|
[DPRG] DPRG ORG 08 Photos
Subject: [DPRG] DPRG ORG 08 Photos
From: Dean Hall
dwhall256 at gmail.com
Date: Tue May 13 23:15:40 CDT 2008
What worked
-----------
1) Using technology that I know and has proven itself: my 7-year-old
ATmega103 based microcontroller board with a library that I have
developed over the years and Larry Barello's AvrX microkernel that I
am familiar with.
2) Using some technology that is new to me:
- Pololu motor driver: http://www.pololu.com/catalog/product/705
- Digi (formerly Maxstream) XBee Series 2 zigbee module
3) Using a recommendation from a former colleague: foam-core paper
board is a rigid enough material that is very lightweight and easy to
work.
4) Placing my LCD on the hatchback so I can see the output.
5) Keeping a budget to track expenses so I didn't go overboard ($600
since November ~= $100/mo over the last 6 months ~= $3/day).
6) Using the ATmega's hardware peripherals and their interrupts to
handle as much as possible. This left me with plenty of CPU cycles
to run the high-level application task.
7) Talking out development issues on a mail list. Even if no one
else is doing something like I am, just trying to explain the
situation lets me see the error or silliness of my ways.
8) Mechanical construction is my weakness, so I outsourced that by
using the chassis and drivetrain of an RC vehicle.
What didn't work
----------------
1) Spending too much time on one detail: I spent a lot of time trying
to tune my PID. I got it to a workable, albeit not-so-great,
condition. I should have spent more time getting steering to work.
2) Putting a "roof" on my vehicle to make it look cool meant it was
very hard to reach my DIP switches for input & configuration.
3) The 4x20 character LCD heated up in direct sunlight and was all
black.
4) The magnetic encoders failed when the wheels exceeded a certain
velocity. I may be able to adjust their physical spacing to increase
operational range.
5) Not knowing a ballpark value for my ticks-per-distance off the top
of my head.
6) Setting PWM to 0 did not electrically brake the vehicle. There
may be a control bit setting to fix that.
Planned Changes
---------------
1) Find a way to allow dynamic input to my microcontroller so I don't
have the slow test/recode/flash cycle. Using the XBee radio and the
UART's RX is the first, obvious choice.
2) Find a way to drive solidly, reliably. If P[I[D]] isn't going to
work reliably, resort to just PWM or find something else that works.
3) Make sure encoders are reliable. This will improve dead-
reckoning, but will also feed into Planned Change #2.
!!Dean
http://www.deanandara.com/Argonaut
On May 13, 2008, at 16:37 , dpa wrote:
>
> I agree with Robert that there were lots of robots ready to
> progress to
> the next more difficult challenges. I think it might be really
> useful to
> all involved if each one can document what worked and what didn't,
> and what changes you have made or need to make to improve your
> robot's performance.
>
More information about the DPRG mailing list
|
