|Ideas||Rott-bot was designed so that its drive wheels projected through both sides of the body, so that being flipped over would not be disabling.||Rott-bot 2000 uses the same principles. The overall shape of the robot is almost the same.|
|Frame||I started with pine 2x4s; when these proved too heavy, the frame was (rapidly) rebuilt with pine 2x2s. Wood is extremely easy to work, but it's heavy and not terribly strong. One of the frame members was nearly sawn through by the arena saws.||Steel. I initially bought 3/4" square tubing, but mindful of Rott-bot's weight problems, I used 1/2" tubing instead, with a 1/16" wall. I cut up the tubing with a miter saw, and took it to a friend who welds. One day of welding later, and I had a frame that weighed only ten pounds that I could stand on.|
|Armor||Rott-bot wasn't exactly armored, but it did have a plywood skin roughly cut to shape and bolted on. The piece in this picture is post-Long Beach, and the labels indicate the robot responsible for inflicting it.||Rott-bot 2000 has 1/16" aluminum tread plate on the bottom and sides, and 1/4" acrylic (not polycarbonate!) on the top. This is the belly plate, the largest piece.|
|Weapon||Rott-bot was supposed to have circular saw blades at the front that would chew up opponents. Unfortunately, due to miscalculations of the weight limit I had to work with, I couldn't include both the extremely heavy drive motor and the battery for powering it. So the saw blades were purely decorative.||Rott-bot 2000 uses a smaller, lighter motor for driving the saw blades, and can afford the weight of it.|
Since I'm interested in electronics, I built my own speed controller.
It's based on a PIC 16C63 microcontroller, and takes two channels of
joystick information and turns it into two-wheel PWM.
Due to poor time management, I didn't get the controller past the prototype stage, and I ended up stuffing a breadboard into the robot. It's a miracle nothing was shaken loose.
I also couldn't get a four-FET H-bridge working in time, so direction control is done with a DPDT relay, and the speed control happens with a single FET on the ground leg.
I'm still using the home-brew speed controller, with a few
improvements: the PWM rate for the motors has been doubled to
60Hz, the control circuitry has been fixed so that the FET is
turned off while the relay is switching to prevent arcing, and
I've had a very nice PCB made. I didn't even bother to try
to eliminate the relays, they worked just fine.
And the speed controller software is now also capable of blinking some LEDs.
As in Rott-bot, the power electronics are optoisolated from the logic electronics.
Rott-bot had two 12V 4Ah lead-acid batteries. The original idea
was that one battery would run the saw, and the other battery
would run the wheels. When it came to light that Rott-bot was
overweight, the saw motor was removed, but the battery was left
in, and, for my second match, was wired in series with the first
battery to give 24V to the drive motors.
I also used an ordinary 4-cell pack of NiCds to run the electronics and the radio receiver, and a 10-cell pack of rechargeable alkalines to act as a gate drive for the FETs on the motor drive.
Almost the same -- I'm using two lead-acid batteries, but when
I went to the battery store to get spares, I found that the
technology has advanced to the point where a battery of the same
size has a capacity of 5Ah, not 4. I have two of these batteries
wired in parallel (so 12V, 10Ah capacity) for running both the
saw and the drive. This is enough for a five-minute match, even
with spirited driving and moderately heavy use of the saw.
I had hoped to eliminate the 15V alkaline pack, but, for reasons I haven't yet figured out, my FET driver wouldn't work without it. And the logic now runs from a 9V battery run through a voltage regulator.
|Drive, Wheels and Gearboxes||
I used the perennial surplus kiddie-car gearboxes that so many
lightweight robot competitors use. They're pretty strong and
very inexpensive ($6 each). The Mabuchi motors in these are
rated for 6V; many competitors run them at 12V. For extra
speed, I wired my batteries in series and ran them at 24V.
The motors mostly survived, but with my crazy driving, the metal pinion on the motor tore teeth off of the first nylon gear in the gearbox.
My main problem was in using hard plastic wheels, even worse than lawnmower wheels, that didn't get much traction. I could see the wheels spinning in the arena. I tried scoring the plastic with a file, but it didn't help much.
You'll also note on the main picture that Rott-bot's axle isn't held in by anything except friction. In practice, I had to drift it out with a hammer and another length of drill rod, so I wasn't too worried that it would be knocked out in combat.
I'm using the same motors and gearboxes in BBSF2000 (well, same
model of gearbox -- the old gearboxes are trashed). I'm
running them just at 12V. I picked up some nice pneumatic rubber
tires at Harbor Freight, so I don't think that traction will be
The axle is held in place by being sandwiched between 1/8" plates of steel.
I used an Airtronics VG4R FM radio. I think that there was something
wrong with my receiver -- when the transmitter was shut off, and the
receiver was in an electrically noisy environment (like my computer
room at home, or the BattleBots arena), the receiver would put out
random pulses. I almost failed safety inspection because of this.
Radio receivers normally have some susceptibility to noise, but this seemed excessive.
|I found an Airtronics PCM radio for sale on eBay, so Rott-bot 2000 will not suffer from the same interference problems. And the failsafe of the receiver means that all the code I painstakingly wrote in the PIC to detect loss of receiver data will not be used. Ah well.|
BBLB was double-elimination, so every robot got to go into the
ring at least twice. Rott-bot lost both matches, but both of
them went to the end and had to be settled by a judge's decision
or an audience vote. My two opponents were Mike Regan, driving
driving Gorange 3.
I was pleased that my first attempt at building a robot survived its two rounds against seasoned competitors.
Rott-bot was twitching madly during the Kilobot Rumble, so I was (properly) asked by the safety people to remove it from the ring.
|We've been asked by BattleBots to keep the results of the matches secret until after the Comedy Central show airs. Unfortunately, the current story is that the lightweight events will not be broadcast at all, so I don't think Rott-bot 2000's performance will be on TV.|
A summary of the problems with Rott-bot:
Rott-bot 200 has a new and improved list of flaws:
Since a lot of people ask about cost, and these robots are near
(or slightly below) the minimum necessary to enter a competition
like BattleBots, you can
use these figures as a rough minimum cost. All costs are approximate.
Keep in mind too that a fair bit of the stuff, like the battery rechargers and the radio transmitter, is not in the arena and is not at risk.
In addition to the material actually used in the robot, I had tools like a decent Wintel computer ($1500), a PICstart Plus programmer for burning the microcontroller ($200), a small drill press ($150), a Makita chop-saw for nice square cuts ($150), a hand-held Skil-Saw ($75) and a cordless hand drill ($75).
Rott-bot 2000 was a little more competitive and prettier.
Again, some of these parts are not at risk from combat, and most of them survived anyway.
I had all the shop equipment I used in Rott-bot, plus I bought a small tablesaw ($150), and a grinder ($60). I used the chop saw for cutting the steel frame members, probably shortening its life. I didn't have to spend $2000 on a TIG welder, because I have a friend who has one, but I may be buying myself a small MIG for future work.
With both Rott-bot and Rott-bot 2000, my actual costs were higher, due to false starts and purchases of spare parts.