Minutes of the PARTS meeting on s, 2003

President Pete Skeggs opened the meeting by announcing that the election of officers for 2004 through 2005 would be held during the meeting, and that a proposed budget for the next year would happen at the December meeting. Pete offered each of the candidates for office an opportunity to briefly talk about their vision for PARTS.

Presidential candidate Monty Goodson said that he is impressed by the expertise and talent among PARTS members (and he’s impressed by the cool robots they build!) He would like to help encourage members to build robots, and hopes to be able to provide visits from component vendors and robotics manufacturers to PARTS monthly meetings.

Vice-presidential candidate Tim Weaver proposes paying for additional web space so that we can maintain an archive of photos, articles, and robotics information.

Incumbent Treasurer Dana Weesner said that his re-election would bring continuity to the office. Dana ended his stump speech by encouraging everyone to vote for the other candidate, Larry Geib. Larry said that if he is elected he would like to see our club’s non-profit status formalized. Larry felt that PARTS could receive more support and donations from corporate sponsors if our non-profit status was in place.

Secretary candidate Paul Burkey said that he has been a long-time PARTS member since the days when Marvin Green organized the club.

Following these brief "campaign speeches" the election of new officers was held. The twenty-six voting members present elected Monty Goodson, Tim Weaver, Larry Geib and Paul Burkey to two-year terms. They will take office beginning with the January, 2004, meeting.

Following elections, Pete offered attendees an opportunity to purchase professionally-produced DVDs of PDXBOT.03 at the club member price of $15. Both +R and –R DVD formats were available.

The education session for the day was "Circuit Board Design", presented by Steven Minichiello. Steven said there are many considerations when designing an etched circuit board: through-hole versus surface-mount components, board and copper cladding thickness, number of copper trace layers, circuit speed, etc. Steven demonstrated the concepts by using ORCad, but said that each board design tool will provide similar features.

Steven said the place to start is with a "netlist" of the circuit design so that design rules can be utilized. If you design a board without a netlist, you must manually ensure that all design rules have been satisfied.

Steven showed ORCad’s graphical representation of the netlist. It was represented as a straight-trace layout of all the circuit interconnections with all of the circuit components arranged in a vertical line. He showed how he could select and drag individual components to a desired location on the board; the traces between components would "rubber band" to give a first-level approximation of trace routing and crossovers. Steven showed how to define "obstacles" to prevent traces from being routed through certain areas (such as clearance holes or other cutouts) on the board. The goal at this stage is to keep interconnections as short as possible, and to reduce trace crossovers.

Steven said that component placement typically takes 20% to 30% of the time required to design a circuit board, and is the most critical aspect of a successful design. Optimum component placement can 1) reduce board size (and cost), 2) affect the number of traces that can be run between components, 3) affect the maximum possible efficiency of the design tool’s auto-router. Steve said that rotating a component by 90 degrees or 180 degrees can often help resolve problems with trace routing. Component placement has to be influenced by need for isolation of signal lines.

Steve listed several things to keep I mind while laying out boards. Most low-end design tools do not attempt to minimize crosstalk between signal lines, so designers need to minimize crosstalk by avoiding long parallel traces and by having traces in different layers cross at 90 degrees to each other. Traces that will be carrying high currents, such as motor driver outputs, should be kept physically isolated from digital signal lines. If the design includes any analog input signals, these must have good grounds; avoid narrow or looping ground traces. These are electrical issues, and should be dealt with by appropriate physical layout.

Isolation of signals can sometimes be improved by rotating a component, or by using a different foil layer for a signal in need of isolation. Use your design tool’s minimum trace width and trace separation design rules to improve separation. Isolation can often be improved by using a "copper pour", or grounded barriers between traces.

Steven showed how ORCad auto-routed a board design. He then showed several areas that needed to be manually tweaked to improve isolation. Steven said he was not aware of any non-product-specific books that describe the process of laying out an etched circuit board. Larry Geib recommended Al Williams’ book and CD-ROM, "Build Your Own Printed Circuit Board" (McGraw-Hill/TAB Electronics, ISBN 007142783X, available at Amazon for $19.57 plus shipping) as a good resource for learning and using EagleCAD light.

The "Show and Tell" portion of the meeting began with Daryl Sandberg, who showed his new "3 for $59.00" mini-sumo circuit boards from PCBExpress. Daryl has decided that his new robot’s name will be "Slick" because of its low-friction Delrin body shell. Daryl plans to use lots of sensors so he’ll always know where his opponent is.

Warren Leach announced that a friend of his was moving and decided to donate a complete Heathkit electronics training course and a powered breadboard system to PARTS for the lending library. Warren provided a complete description of the series to librarian Brad Lewis; Brad already has included the new materials in the lending library list.

Paul Burkey mentioned that he had a video of his new robotic lawn mower that members could watch after the meeting.

Larry Geib passed around his copy of "Build Your Own Printed Circuit Board". The included CD-ROM includes a copy of the "light" version of EagleCAD, and can be run under either LINUX or Windows. Larry bought his copy locally for $27.95.

Monty Goodson showed a video of Ted Greibling’s M&M sorting robot in action He also showed some nanos on loan from Scott Davis (former PARTS member; now in San Diego, CA), Casey Homes (Simi Valley, CA), and Bronson Silva (Campbell, CA). Nano-sumo robots must fit completely within a 2.5 cm cube, and must weigh less than 25 grams. Most nano-sumo robots use rechargeable Lithium-poly cells (3.6 volts, 140 mAh, about $5.00 each) to power small motors. Casey Holmes' nanos used two motors and gear trains from a couple of R/C aircraft "pico" servos.

Tim Weaver said that he was pretty much "RoboMaxx’ed out". He couldn’t get LandShark ready in time, but is looking forward to next spring.

Steve Davee brought some small blue and clear plastic boxes, suitable for storing micro-sumo and nano-sumo robots. He said that folks interested in getting started in nano-sumo robotics should check out micro-flight websites as source for Li-poly cells. Steve showed a sheet of "Dos and Don’ts" for working with Li-poly cells. Steve passed around the November 2003 issue of Scientific American that has an article about SWARM robots.

Dana Weesner said that he attended Robothon in Seattle on Saturday only. He said that they had competitions of antweight radio-controlled robots against autonomous mini-sumo robots. He said that the antweight guys had painted white lines around their hazards, and were starting to realize that autonomous robots have better control than human controlled robots.

Brad Lewis urged PARTS members to check the club library website for listings of lending library materials. Send Brad an email prior to a meeting if you want him to bring a particular item for you.

Bob Pearson said that he and son Bob are involved in LEGO robotics again this year. The course this year does not have lines that can be followed to lead the robot from one challenge to the next. Teams will have only six weeks from the time the challenge is officially released until the competitions begin.

Mark Medonis showed a $19.95 black and white video camera he bought from Supercircuits.com. The camera operates on a 12 volt supply and has a CMOS image sensor.

Steven Minichiello reported that the developer of the robot "Johnny 5" in the movie "Short Circuit" is working on a set of blueprints for builders who would like to build one of their own. Pete said that he would post details on the PARTS Yahoo group news list.

Shawn Marshall showed his full-size replica "proton pack" from the movie "Ghostbusters". Shawn and a friend built the pack in 1984.

Mike Dinhart showed the mini-sumo robot he built in Brett Nelson’s robotics class. The robot currently avoids walls, and can find an object. Mike wants to modify his code to change the robot’s behavior. Mike is participating in Science Olympics this year, and the goal is to build a robot that sequentially sorts billiard balls. He wanted some help with the design of the robot gripper arm.

Pete Skeggs showed a serial to parallel adapter cable for programming Parallax Basic Stamp 1 processors. This eliminates all the problems BS1 users have had with poor parallel port support under Windows. The cable hardware includes a Maxim MAX202 chip and a few other components. The software program runs under Windows. Details are available at www.picaxe.co.uk. Pete won a Parallax Toddler for winning the maze course at Robothon in Seattle. Pete said that there was a demonstration on Sunday of Dave Anderson’s balancing robot, nBot. Details, photos and videos are available at http://www.geology.smu.edu/~dpa-www/robo/nbot/index.html.

The next PARTS meeting will be Saturday, December 6th at 10:30 am in PCAT 28.