View Single Post
  #35   Spotlight this post!  
Unread 20-01-2011, 11:17
JesseK's Avatar
JesseK JesseK is online now
Expert Flybot Crasher
FRC #1885 (ILITE)
Team Role: Mentor
 
Join Date: Mar 2007
Rookie Year: 2005
Location: Reston, VA
Posts: 3,674
JesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond reputeJesseK has a reputation beyond repute
Re: Never Do This, and Other Good Ideas Gone Bad

1.) Do not gear a drive train for >10 fps if it has less than 3 CIMs' worth of power. It does ok in a straight line, but even with a 6WD drop center it had some difficulty turning 145lbs.
2.) Do not use 80/20 for a triple-compound elevator that has 20lb worth of manipulator and game piece on the top stage. The stresses on the bottom stage will increase the friction in the 80/20 linear slides too much.
3.) Do not make a 'naturally articulating' drive train. The drive train needs to be rigid or controlled in all aspects at all times. Our 'naturally articulated' drive train pods worked 'fine' in 2010, yet they didn't quite perform as expected. By the time we were testing it, we didn't have time to change to a new design.
4.) Do not design a multi-configuration 'automode' that is configured via the little switches on a Classmate PC. The little buttons are just too durn small.
5.) Elevators that have multiple pre-programmed positions + override/tuning are great in theory, yet are very difficult to master from a controls perspective. We're going to do a multi-position elevator for the driver again, yet with a very nice twist to add some autonomy.