Quote:
|
Originally Posted by DougHogg
We learned that if a battery isn't allowed to complete it's charging cycle, it can show a high voltage because of a "surface charge" but still not be fully charged.
3) In previous years, we had used velcro or zip ties to secure the two halves of the Anderson connectors during a match (something that we learned from other teams), but it didn't seem to be a problem this year, so we didn't bother.
Doug Hogg
|
Battery charging is a perennial problem. Managing a number of batteries on a hodgepodge of chargers by picking the one with the highest voltage can leave you very surprised. Having dealt with this problem for a number of years, we bought a 5 independent channel, 6 amps per channel, three stage charger this year and it does a great job of keeping up with 6 batteries (one being in the robot). The charger is
http://www.batteryweb.com/interacter_x5.cfm
using the LS1206 charger (5 of them) in the unit. It is expensive, but it keeps up with the battery charging even when going through the finals and has reliable indicators for the state of the charge cycle. We also test our batteries for capacity with a constant current load tester before the season and remove any that do not measure up from competition service. Our used 2004 ES batteries had about 50% more capacity than new 2005 EX batteries, this year.
You gotta keep the religion on that velcro strap for the power connector. Keep spares in case the one on the robot is lost.
Now for at least some of Team 1280's lessons learned, no shortage here:
* We must have the record for the most broken axles at the Sacramento and San Jose regionals. We were using heat treated O-1 steel that was far too brittle. By changing this to heat treated chrome-moly (4130, or 4340) we got axles that are tough enough for outboard wheel use. After any welding is done using rod of the same alloy, heat treat, quench in oil and then temper at 800 degrees. Such axles are very strong but still bend instead of snapping when hit with a sledge hammer.
* The PWM style connectors used on the RC controller have a bad habit of working their way off when robots interact strongly. An open loop
PID control system is not a pleasant thing. You need to add a means
means of keeping these connectors in place.
* Setting your drive victors for coast when the power goes out is good for that last foot or two going home when the clock runs out, but is a real disaster when a robot is giving you a shove. You can set up an RC controlled relay to get "brake" during normal operation and "coast" when the robot is disabled at the end of the match.
* A cast iron worm gear is very incompatible with a hardened steel worm, always use a brass worm gear in such a gear box. We learned this just prior to ship when we found the cast iron worm gear in the bottom of the gear box as black dust.
* It is important that the couplings to pots in a PID control system are absolutely bullet proof. Glued on rubber hose can break loose and slip,
leading to potentially destructive a loss of control. If the hose is flexible,
like latex tubing, it flexes a bit and you lose precision in the control.
* Your robot had better be built to take a high speed collision from all sides, or the weak side will likely get bashed in.
* Your robot had better not be too close to the size limits, or any
change in the dimensions of the frame due to collisions might result
in a robot that is "out of the box." It is not fun to be shortening a robot
with a sawsall in order to pass size inspection at the Nationals.
* After spending countless hours, and much money, building a good robot to go to a competition; take your scouting very serously or you will make serious mistakes in alliance selection. If you get the opportunity to go to the nationals, take your scouting even more seriously because you have no "corporate memory" for teams arriving from other parts of the country, and you don't see a given robot as much in the seeding matches due to the larger number of robots in the division.
Doug, We hope to see you at the Cal Games next year!
Eugene