Minutes of the PARTS meeting on May 3rd, 2003

President Pete Skeggs opened the meeting by announcing that the bulk of the meeting would be devoted to PDXBOT.03, our robotics exposition and competition scheduled for May 25, 2003. Pete reported that he had videos of our past events that he could loan to anyone wanting to study mini sumo robot behaviors. Pete passed around copies of the PDXBOT.03 promotional poster and encouraged members to visit the club website, download and print several copies of the poster, and display them in schools, in libraries, at work, etc. Pete also passed around images showing the development of the PDXBOT.03 tee shirt design, and encouraged members to buy tee shirts on the day of the event. Pete recognized club member and Tektronix employee Monty Goodson who arranged for the donation of a Tektronix digital storage oscilloscope as a raffle prize, and who also got a $237 grant from Tek to help offset our cost of producing PDXBOT.03.

Pete explained the seven competition events: Japan-class (3 kg) sumo, mini-sumo, micro-sumo, beginner line following, advanced line following, walker robot course, and robot talent show. He said that prizes would be awarded in the competitions, and that we would have raffles for robotic equipment. Pete asked for members to bring technology demonstrations to PDXBOT.03.

PDXBOT.03 event coordinator Larry Geib urged everyone to read the rules for the events (posted on the PDXBOT.03 area of the PARTS website) they plan to enter. He reminded all mini- and micro-sumo competitors that sticky tires and sharp scoops are not permitted. Robots entered in the advanced line following competition need to be less than six inches high to avoid an overhead obstruction. Entrants in the robot talent show should plan on a maximum of five minutes to demonstrate their robot’s unique abilities. Larry urged everyone to pre-register online to save time on the day of the event.

Pete introduced Bill Harrison of Sine Robotics who showed two of his Japan-class sumo robots. Bill has been the driving force behind sumo robotic competition in the United States, and has traveled to Japan to compete with his robots. Bill is interested in increasing the number of U.S. and Canadian-built Japan-class sumo robots, and making them more competitive against the top robots from Japan. Bill thinks that U.S. builders can start to select from more and more refined designs for scoops, motor drives, tires and traction technologies.

Bill showed how he built a vacuum traction system into a robot . He used a small surplus membrane vacuum pump driven by a DC motor. The pump creates a vacuum of a few psi (4 or 5 psi is sufficient) within a rectangular chamber gasketed with Teflon. The vacuum is applied ahead of the driving wheels, and behind the scoop. One advantage of a vacuum system is that it will work with any smooth surface. There are a few drawbacks, however. The vacuum pump often requires more powers than the drive motors, and once vacuum is lost your traction boost is gone, too.

Bill then showed a robot that uses magnets from a hard disk drive head positioning mechanism for traction boost . He said that working with magnets of this strength requires precautions. They can be brittle, can pinch fingers, and will magnetize drill bits used to drill mounting holes. Bill said that mounting the magnets is the biggest challenge; if they are already attached to mounting plates, leave them attached. If you need to mount bare magnets, Bill suggests using an epoxy that is not brittle when cured. To be most effective, magnets must be mounted as close as possible to the surface of the floor. Unlike vacuum systems, magnetic systems don’t require any power for their operation, and they immediately re-apply their downward force if the robot is momentarily lifted by an opponent.

Bill cautioned that traction-boosting systems cannot be added after the fact to an existing robot; the design must include the effects of the additional downward forces that will be generated. Friction that we’re used to under normal gravitation is quite a bit different under the additional downward pressure provided by either vacuum or magnetic systems. These robots need to be built very rigidly; the base of Bill’s robot was made of 0.250" T6061 aluminum, the minimum thickness he feels will stand up to the deflection that results from vacuum.

Bill had various other suggestions. Bill uses shim stock for his scoops, but says they need to be supported from behind to make them beefier. On four-wheel drive systems, Bill positions the front and real axles as close together as possible to improve pivoting speed. He has intentionally raised the rear wheels on one of his robots to convert it to two-wheel drive. If the front of the robot is lifted by an opponent, the rear wheels will contact the ring and start providing additional traction. This gives him fast pivoting capability, and additional power when required.

Bill said that the wheels on each side of four-wheel drive systems should rotate at the same speed. He prefers a drive system with a single motor mounted between the wheels. The motor output gear drives both wheels at the same time. Bill uses gear motors with about a 10:1 reduction at the motor, and selects driving gears on the wheels that provide another 2:1 (or perhaps 4:1) ratio at the wheels. At least 3/16" tooth engagement is necessary to prevent damage to gears; gears thinner than 3/16" will be damaged by stress. Bill said that 0.250" or 0.375" rod is sufficient for wheel axles, if rigid bearing mounting is possible. Bill usually starts with a gearbox and finds a motor that will work, rather than the other way around. He has had good success with American-made Clifton Precision motors that provide about 20 ounce-inches of torque. Bill said that he shoots for at least 30 pounds of pushing force and forward speeds between 200 to 300 inches per second; he said that some driver detuning may be necessary to achieve a good balance among torque, speed and motor control.

Bill suggested that builders of Japan-class robots use NiCad cells for power. He uses multiples of 6-cell batteries that yield 7.2 volts. He’ll typically use twice the rated motor voltage for performance, although this shortens motor life. He has used lead-acid batteries for higher-voltage systems, but reports that lead-acid batteries cannot produce the necessary current.

Bill said that there are maybe a dozen "really competitive" Japan-class robots in North America. Of those, perhaps only six could beat the top Japanese robots. He said that there are about a hundred Japan-class robots that regularly show up at competitions, and estimates that there are about a thousand Japan-class robots under construction in the U.S and Canada.

Pete thanked Bill for his insights, and the Show and Tell session began.

Gene Collins showed the latest improvements to his robot . He has added a rotating/tilting turret on the top; the turret holds a laser pointer and a video camera. Gene hopes to implement a parallax rangefinder. He has added movable feelers on the front as touch sensors, and a Sharp IR sensor for distance measurement. He has added a pendulum that allows him to determine if the platform is tilted.

Shawn Marshall showed his bread boarded PKE meter circuit . The problem he had with twitching servos was cured with a larger capacitor across his timer chip power connections. He had questions about building an internal battery charger.

John Hurley reported that he had downloaded Mark Gross’s JAL compiler for the Mk III controller board, and it works fine.

Bruce Wiedemann reported that he is working on another Battle Bot. The titanium skin is complete. Bruce had questions about using the Sharp IR sensors for weapons deployment.

Mark Medonis showed the updated vacuum-formed chin that he made for Maxwell . He also brought several molds and forms that he uses to vacuum-form the polycarbonate. Mark plans to offer Maxwell kits at PDXBOT.03.

Karl Kuchs showed his new colony robots . He is using the tank chassis from Budget Robotics , and dual DIOS controllers from Kronos Robotics. He uses an overhead CMUCam to track the positions of his robots. He recommended using the expanded PVC sheets sold by Budget Robotics for prototyping. He’s building a line following robot that uses a CMUCam connected to a BASIC Atom. He’s also building a micro-sumo robot that uses micro servos for propulsion.

Karl Boe reported that he bought a DIOS Ultra board from Kronos Robotics. He’s very happy with the board and the development environment; he likes it better than the BASIC Atom that he had been using. He plans to mount it on his robot, "Blockhead". Karl also got a Zilog Z-8 Encore promotional board with a C compiler for only $50. Others in attendance suggested getting one of these boards before the promotional pricing ends. The boards are available from Mouser and Digi-Key.

Monty Goodson showed his new micro sumo controller board with line sensors . It is able to sense lines at a distance from 1 to 5 mm.

Tim Weaver showed his new AVR controller board for his mini sumo robot . Tim’s design features twin gearmotors, wheels with polyurethane tires, and the circuit board forms the bottom of the robot. Tim has used LINUX development tools for all of his programming.

Warren Leach reported that the current issue of Circuit Cellar magazine has a project to build a microprocessor-based cable tester that uses JAL.

Joey Triska showed his Mk III mini sumo robot with Devantech sonar detectors. It performed well in the after-meeting competition.

Daryl Sandberg showed Boxter 4.0, his Japan-class sumo robot with magnetic traction assist . Daryl and Bill Harrison had a "robot pull" in which each placed a robot on opposite sides of a sheet of 0.250" steel. Bill’s robot with vacuum assist had more power than Daryl’s robot, but Daryl’s robot had greater attraction for the steel plate than Bill’s magnetic-assist robot.

After the meeting Bill spent time teaching PARTS volunteers in the finer points of judging and refereeing a mini sumo match .

PDXBOT.03 is Sunday, May 25 from 10am until 5 pm in the Smith Center Ballroom.

The next PARTS monthly meeting is June 7 at 10:30 am in PCAT 28.