| |
NAVIGATION
|
|
|
|
|
Aftershock is the successor of Shockwave. Originally, I was going to
call it "Shockwave II," but my brother (Peter) suggested the name of
Aftershock, I liked it, so it stayed. Aftershock was built so that I
would have a robot that could defeat Grant Mckee's Enders Wraith.
Aftershock is very fast, and has a great deal of traction with its
homemade tires. Unfortunately, I fell into the dreaded trap of
finishing your robot at the last minute; actually, it was finished a
few hours before we left for Canada (blame it on a term paper, not me
;-)! Had I finished it sooner, I would have been able to fix its
numerous weaknesses that have become apparent to me. So the end result
was Aftershock failed to defeat Ender, and also lost to Solo, another
of his mini sumos. Although Aftershock was a miserable failure (in my
humble opinion...) it was quite educational. You learn from your
mistakes, and I made quite a few - more than from it's
predecessor, Shockwave! If you want to know what
they are, here the are:
- Finish your robot on time! Otherwise you put too much
stress on yourself.
- Mini sumos are subject to considerable forces, strongly
support your frame. Aftershock is made from 0.025' 2024 aluminum, and it got
bent pretty well. I should have stuck with steel like I did with Shockwave.
- Don't use an algorithm where the microcontroller pulses
the infrared emitter and listens for an echo, it takes too long and isn't very
reliable. Aftershock's front sensors are almost worthless.
- Put a lot of effort into making quality printed
circuit boards, each trace can easily be a point of failure. If you
accidentally peel off a trace, don't take shortcuts in fixing it. Thank God
this happened *After* the competition!
- If your robot is performing timed maneuvers, be aware that
the battery voltage changes as the robot runs and is recharged and thus alters
how much your robot turns, backs up, etc.
- If you are building
your own battery pack, make sure that each of the cells are at least
one-hundredth of a volt within each other. My reward for not
doing this is a ruined battery cell that had to be traced and removed
(apparently, it had been cycled under 1 volt!). I was very fortunate
this happened
*After* the event!
- Make sure your scoop support can keep the leading edge on
the floor, even after several impacts.
- Ensure that your wedge has no aberrations on which the
opponent's scoop can catch on. (this is what cost me in the finals, I ended up
in a pushing match with you-know-who and then one of my edge sensors acted up).
- The most important sensor on your robot is the edge
sensor. I found out that it is better to have an auto-calibrating edge sensor
than manually setting the threshold. Also, shield your sensors from ambient
light. If you saw my robot at the games, you may know that all the bouts I
lost were due to driving off the ring. Yup, thats right, I would have
swept everybody if it weren't for my edge sensors!
|
|
|
|
Western Canadian Robot Games 2005, mini sumo, advanced division,
second place! Technically third, but 1st and 2nd were won by the same
person (congratulations, Grant Mckee!). However, WCRG doesn't more than
two medals and prizes to the same individual.
Robonexus, mini sumo, nothing special. No event was held, but I did get
to face some other opponents, a sumovore, and a Marvin Slyder. Oh, and
I *could* have faced Mr. Calkin's mini sumo, but I was to caught up in
enjoying the antweights that I postponed it until I had to leave!
Stupid! =( Oh, well, sorry Mr. Calkins, I guess your robot won't get
schooled until 2006. ;-) Or am I mixed up? I don't know for sure...
Anyway, Aftershock was pretty messed up at the Robonexus, (it was
running at a lower voltage, the front sensors failed again, the motors
were starting to die, and I "misplaced" the programming cable) so I probably
would have lost anyway.
|
|
|
|
Class: mini sumo
Status: being reparied
Size: 9.9 x 9.9 x 6 cm
Weight: 445 grams
Drive Power: Battery
Motors: 2 overvolted Sanyo 75:1 gearmotors
Wheels: 2 Solarbotics RW2 hubs. Rubber is replaced with foam and silicone (I am using urethane tires now)
Drive type: 2 drive wheel differential steering
Battery: custom made 9.6v 1500mAh battery.
Frame: Integrated. 0.022" Brass base, 0.025" 2024 aluminum sides, spring steel scoop
Brain: Basic Stamp 2
Sensors: 2 GP2Y0D340K for sides, homemade IRPD front, same cheap Lynxmotion line sensors.
Circuit Board Construction: Homemade printed circuit board
Weapon(s): ineffective wedge, driving force
Weapon Power: Driving force
Time to build and design: 2 months
Estimated cost: $90 (most of the parts were taken from Shockwave)
|
|
|
|
 Here
is the top of Aftershock with the top circuit board and circuit
board-standoff screws removed. Towards the right you can see part of
the battery pack in a triangle formation. If you look at the image
closely, you can see part of the other battery pack under the circuit
board. Both of these battery packs (3.6 volts and 6 volts) are put in
series and power both the electronics and motors. Note that the inside
of the robot isn't painted. Those blue wires you see connect the motors
to the "motherboard" (as I like to call it). Underneath the 3.6 volt
pack are the motors. The whitish thingy-majigers is the background. ;-)
No really, its the silicone coated tires. (I'll get to the wheels in a
second, depending on how fast you can read. 8^) Oh by the way, look
between the motor drivers, you can see the remains of an exploded capacitor. It
blew up when I accidentally powered up the robot while it was being
charge, creating a voltage too high for the capacitor to handle. By
God's grace I removed the plug fast enough to save the rest of the
robot. This was two days before we left for Canada so that was just too
close...
 And
here is the scoop and "wedge" of Aftershock. The three holes are for
the sensors to see through (more on that later). Because they
camouflage with the paint, you'll have to look carefully, and if you
do, you will see the edge sensors, which are connected to a circuit
with servo connectors connected to it. This package is the Lynxmotion
line sensor. I have been utterly disappointed with these as they cost
$15 and are very unreliable. In fact, some of the packages have an LED
driver (comparator) that can't even fully turn off the LED! I have also
found that it even burns out the infrared emitter with too much current sometimes!
Disgusting. The sides of the scoop are bent up (or down) and have holes
for screws to hold the it to the rest of the body. The holes are
enlarged to allow me to adjust the scoop to keep it flat on the ring
and then tighten it down. Big mistake because during big hits, the
scoop would shift raising the scoop off the ring. I should have coated
the surfaces with neoprene rubber to discourage this...
 Front
view the scoop. The bluey chinsy-thingy-majig at the tip is blue spring
steel. Aluminum is too flexible and soft to be used as a scoop. I glued
it on with silicone to make the scoop somewhat flexible hoping it would
be better able to get under the opponent. Due to the high angle of the
wedge, this "added flexibility" hardly makes any difference at all.
 These
standoffs (purchased from Radio Shack) play an important role in the
frame of this robot. First of all, they hold the circuit board in place
and provide room for the 6 volt battery pack. the sides of standoffs
are drilled and "tapped" (I didn't tap them really, I just forced the
screws through the hole - ugh). The screws are what hold the brass base
plate to the shell. This setup is one that I am really proud of 'cause
it saves so much room. The
holes you see on the top were originally going to be used to mount a
polycarbonate top. Unfortunately, some of the electrical components are
too tall to allow for the Polycarbonate. To allow the side sensors to
see out, I cut out a rectangular hole in the body (left of picture).
 This
is the right side of Aftershock. From here you can see what an elegant
solution the standoff was. Notice that on this side I cut the aluminum
too short. Now about the wheels. The hub came from Solarbotics.
Replacing the rubber, antistatic foam is placed around the hub. The
foam had to be "treated" to make it softer. Since antistatic foam isn't
very sticky, I coated it with silicone. I hate silicone. It only keeps
its bite for a few hours and then becomes a normal rubber that is very
difficult to peel off the wheel. However, while it is still sticky,
Aftershock had so much traction that its Sanyo 75:1 gearmotors
running at over 9.6 volts (full charge) could barely spin the tires (if
it was going against something it couldn't push)! Nonetheless, next
mini sumo is going to use polyurethane for rubber. One of the side
sensors are looking at you in this picture... Notice how nicely it fits
within the rectangle; this isn't the case with the other side.
 Here
is the rear view of Aftershock. Unfortunately, It lacks a rear opponent
sensor, and thus is subject to getting hit from behind. The screws go
into the drivetrain which is foam taped (visible in this picture) to
the base plate. Not much here.
 ...And here is a peek at the drivetrain which came from Shockwave.
It is made from 0.01" shim stock steel. The sides and back are bent up
to for the motor faceplate and screw holes respectively. The motors and
the screws that hold them in come from Solarbotics. I had to break off
the 3.6 volt battery to take this picture, more repairs are needed.
 This
is the switching regulator module. The red pushbutton the reset/
start. Originally, I had a start button and a reset button, but
during the competition, while I was under tremendous pressure, I twice
made the mistake of pressing the reset button when I should have
pressed the start button. So, off it went. When the reset button is
pressed, the robot writes in a certain EEPROM address, and then sleeps.
The next time it is pressed, the controller erases the EEPROM
memory and starts the 5 second countdown. Thus, I can also use the
power switch for the same purpose. The red dot is an unlit indicator
LED that tells me the countdown has begun. The clear dot is the power
indicator. The black spot you see are two switches I was going to use
as strategy switches. Unfortunately, I never had the time to make some
additional algorithms.
 Front
view of the robot, wedge removed. In case you are wondering why
the motherboard is so crooked, its because the screws that hold it down
are presently missing. =( You also get a good view of the front
sensors. They work by pulsing a 38 kilohertz signal from the left
emitter, then the receiver (in the middle) checks for an echo. The
process is repeated with the right emitter. If both signals return, the
opponent is in front. If only one returns, the robot turns towards that
side. If none returns, the robot explodes. Haha, no, it continues
searching. The board is covered with antistatic foam to prevent
infrared signals from leaking into the receiver. I'll bet Aftershock
uses more antistatic foam than any other sumo 'bot! :-)
 And
finally, here we have all the circuit boards stacked together, the
front sensors, the motherboard, and the switching regulator. I actually
borrowed this idea from the Solarbotics Sumovore. The front sensors are
attached to the mother board with solid 22 AWG copper wire. I should
have used right angle connectors... So, how do you like it? :-)
|
|
© 2005 All Rights Reserved.
Create a
free website at Webs.com
|  |
|
