A Look at the Little Green Machine
August 2003
Fig. 1: The Little Green Machine.
Introduction
The Little Green Machine (LGM) was entered in the Trinity Fire Fighting Contest in 2003.
He is the most complex robot I have built to date. From this project I learned a lot,
and I though I would show off the robot a little. This note shows how it was
constructed and what it looked like.
Fig. 2: Top and side views of the LGM.
As you can see in Figure 2, the robot was pretty simplistic in design. It used differential
drive with modified servos. The wheels are 5.5" in diameter and approximately 1/4" thick.
The main frame piece (the big green circle) is 11" in diameter and 1/4" thick. The robot
had one omni-directional caster which was mounted at fixed height with a few pieces of
C-channel. There was a tail made of spring stock under the back to keep the robot from
doing wheelies.
From the top view you can also see most of the sensors, note that the bottom of the
picture is the front of the robot. There are two GP2D120 sensors
vmounted on the top of the deck, there is one on the outside edge lined up with the wheel
axis. Two forward facing GP2D12 sensors are lined up in a crossed pattern on the from of
the robots deck. The UVTron is the small beige rectangular PCB just behind the forward
IR sensors. The PyroElectric sensor is the white cone that is on the end of the snaked
wire in the side view.
The little mounts that the IR rangers are held on with are small pieces of L-angle
aluminum, drilled and painted.
Fig. 3: The processing power.
Figure 3 shows the Dios Ultra board from Kronos Robotics. The smaller board next to it
is a Kronos Robotics SN754410 board that I used to drive the fan. The Dios Ultra was
powered by the 7.2V RC battery you see next to it, the board has a 7805 regulator
built-in. The fan was powered by a 9V battery which is mounted underneath the frame.
Fig. 4: Underside of LGM and a closeup of interface board.
The headers on the Dios Ultra were wired to the interface board which is under the
frame. This board supplied the IR sensors with their own regulated supply. There is
a board underneath the interface board, which you really can't see, which mixed the
servo power with the servo output of the Dios board, and carried a window comparator
for the Pyroelectric. I had designed for the robot to have 7 IR sensors, and a
pyroelectric, which would have used the 8 analog inputs, however I found out that
one of the analog channels is broken in the chip design! I quickly wired up a
window comparator for the Pyroelectric sensor. In the end I never actually got around
to interfacing the final IR sensor (which was to be a long range analog one facing
forward for exiting rooms) and could have avoided using the window comparator.
There is a set of 4 AA NiMH batteries that you can see in figure 4 that
powered the servos.
Fig. 5: View of motors and caster, closeup of servo and wheel.
I made the wheels from 1/4" thick PVC on a lathe. They were turned down to diameter
and given a groove in the edge, which allowed the O-Ring to sit on it. As I said, the
diameter was too big, they stressed the motors a lot when starting and stopping.
About a week after the contest, I happened to see that for about $3-4 you can
get the same thing, in a smaller diameter from Budget Robotics (it was a new item).
They even sell replacement O-Rings. The servos were bought from there, because
they sell already modified servos and the price is good. The servo mounts also
came from Budget. They are a very nice design, and real sturdy.
|
|
