Rott-bot vs Rott-bot 2000

I've made two competition robots so far, Rott-bot and Rott-bot 2000. Their designs are similar to one another, but the techniques I used in construction improved over the year.

ROTT-BOT
Rott-bot
ROTT-BOT 2000
Rott-bot 2000
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. Rott-bot 2000 frame 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 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 armor 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 weapon Rott-bot 2000 uses a smaller, lighter motor for driving the saw blades, and can afford the weight of it.
Speed Controller Rott-bot speed controller 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.

Rott-bot 2000 speed controller 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.

Power 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.

Rott-bot 2000 batteries 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 a problem.

The axle is held in place by being sandwiched between 1/8" plates of steel.

Radio Gear 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.
Performance 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 Deathtrap, and Grayson DuRaine, 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.
Flaws A summary of the problems with Rott-bot:
  • Wheels: hard plastic wheels that didn't grip. The most significant flaw by far.
  • Gearboxes: delicate teeth shorn off by spirited driving.
  • Radio gear: too vulnerable to electrical noise. The organizers warned against using FM, but would I listen?
  • Armor: plywood is weak and heavy, and the sides of Rott-bot were completely unarmored. If the arena spikes had been slightly higher, I'd have done even worse.
  • Weight: using heavy wood and not measuring carefully enough resulted in a quick frame rebuild, and a non-working saw motor.
  • Technique: Rott-bot was still not running when I loaded the carcase into my car for the drive to Long Beach. The first time I drove it was in my first match. You can imagine that my control was not perfect.
  • Appearance: ugliest robot at the event.
Rott-bot 200 has a new and improved list of flaws:
  • Gearboxes: geared too low (I had great torque, but a top speed of only about 2 MPH), and they're fairly delicate... sudden direction changes can tear teeth off of the plastic gears.
  • Speed controller: should have soldered the integrated circuits directly to the printed-circuit board rather than using sockets, or failing that should have glued them into the sockets. Also, the speed controller was uncovered, so any conductive debris that got inside the body (and there was a big slot where the saw drive belt came out) could easily short out the speed controller. It should have been wrapped in something, or maybe dipped in plastic.
  • Armor: delicate brittle acrylic for the lid. Next time I splurge on polycarbonate.
  • Casters: held on by two #8 (4mm diameter) bolts, and a good side hit can shear them off. I should have welded some plate steel horizontally further inboard, and used four bolts to hold the casters to that.
  • Recharge time: I should have had semi-exposed power studs so I can recharge the main batteries without opening it up. I should have had a bigger recharger. I maybe should switch to nickel-cadmium batteries rather than lead-acid, still investigating that.
  • Construction: the whole thing should have been easier to open and close. I had too much "layering", so that to (say) remove the speed controller circuit board I had to remove the lid and the front plates in order to get a wrench onto the nuts holding the belly plate, and then remove the belly plate to get at the nuts holding the circuit board to a piece of wood inside. Should have used more tapped holes and fewer nuts. Also, not enough room in the electronics compartment. I wanted to add a car horn as a gag (probably would have been inaudible anyway), but I ran out of room.
  • Weapon: A 5/8" piece of threaded rod that is essentially being used for a battering ram should be supported at more than just the ends.
  • Technique: I'm a terrible driver, should have practiced more.
  • Appearance: Fit and finish on the armor... the front plates were added last, and they look terrible.
  • Achilles Heel: Power switches are exposed at the front, and a well-aimed hit could have simply turned the machine off. What I really wanted was those switches they have in military and space equipment, with the switch flanked by two semi-circular pieces of metal, but I couldn't find any at the surplus stores.
Cost 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.

Item Cost Source
Radio Gear (Airtronics VG4R FM, new) $150 Sheldon's Hobbies
Wood for frame and armor $50 Local hardware store (OSH)
Electronics (PIC microcontroller, MOSFETs, transistors, relays) $100 Digi-Key, Jameco and JDR. The last two are local to me, but all three do mailorder.
Miscellaneous hardware (nuts, bolts, washers, casters) $60 Local hardware store (Dale)
Weapon (threaded rod, nuts, washers, pillow blocks, saw blades) $60
(mostly for the pillow blocks)
Local hardware store (Dale)
Batteries (two 4Ah lead-acid, one 4.8V NiCd, ten rechargeable AA alkalines) $120 (including rechargers) Lead-acid: local electronics store (HdB in Redwood City); NiCd: came with radio gear; AAs and rechargers: local computer/electronics store (Fry's)
Wheels, gearboxes and motors $40 Wheels: local hardware store (Dale); gearboxes and motors: C&H Surplus
Total $580 my bank account

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.

Item Cost Source
Radio Gear (Airtronics PCM, used) $150 eBay
Steel for frame $30 Local hardware store (Dale)
Tread plate aluminum for armor, aluminum angle for corners $60 Local hardware store (Dale)
Acrylic lid $20 TAP Plastics, a west-cost plastic shop
Electronics (PIC microcontroller, MOSFETs, transistors, relays, LEDs) $100 Digi-Key, Jameco and JDR. The last two are local to me, but all three do mailorder.
Printed circuit board fab $150 ExpressPCB. They only make boards in pairs, so the $150 was actually for two boards.
Miscellaneous hardware (nuts, bolts, washers, casters) $60 Local hardware store (Dale)
Weapon hardware (threaded rod, nuts, washers, pillow blocks, saw blades) $60
(mostly for the pillow blocks)
Local hardware store (Dale) (actually scavenged from Rott-bot, so real cost was $0)
Weapon motor $15 Surplus, not sure which store
Batteries (two 5Ah lead-acid, ten rechargeable AA alkalines) $120
(including rechargers)
Lead-acid: local electronics store (HdB in Redwood City); AAs and rechargers: local computer/electronics store (Fry's)
Wheels $20 Harbor Freight
Gearboxes and motors $20 C&H Surplus
Total $805 my bank account

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.


This page works best with Lynx.
mail me or go home.