Rookie Teams: Programming Examples

[Jeff McCune]

Personally, as a programmer I learn best through example. So, I wanted to mention to rookie teams, and any team for that matter, than I back up our team's code every night after we work on it (Sometimes more often) to a public webserver.

I'm by no means saying our code is spectacular or anything like that (Quite the contrary.) But we area a veteran team, and we're doing some things that from my experience working with other teams in Columbus might seem "complex" to a rookie team.

We chose to stick with PBASIC 2.0 this year rather than 2.5, but all the stuff here still applies. The actual syntax is just easier in 2.5. Logic is still the same.

Some things that are in our code base that might be interesting to your team. (Again, this is by no means the "correct" or even a "good" way to do these tasks. It's just how our team does them.)

1: Serial Communication to an external Basic Stamp, complete with a system to gracefully handle the external microcontroller getting jarred lose, and possibly coming back alive during the match.
2: Shaft Encoder routines to sense velocity and distance. (This is more hardware than software, as we have one hardware counter register attached to each of 4 shaft encoders on our robot.)
3: Single Joystick Tank Style driving. (Not so cool, since it's in the default code now.)
4: Low Pass Filter on all joystick inputs. (Makes motor acceleration smooth given "jerky" joystick movements.)
5: Momentary push buttons (Like the trigger) acting as a toggle.
6: A compass (Soon to come.)
7: Heavy use of subroutines to give the feel of programming in a higher level language.
8: Heavily automated transmission control.
9: Joystick Dead-Zone logic, which plays nice with the low pass filter. (If you combine the LPF example form the whitepaper on this site with the Dead Zone example from first, they don't play nice togeher... The Dead Zone will "overrride" the LPF the instant the joystick enters it.)

Stuff to be implemented as of right now:
1: Autonomous code. (It'll be there in a few days.)
2: GyroScope logic (For sensing top of ramp)
3: Optic Sensor logic (Also for sensing top of ramp.)
4: Completely automatic CVT handling (I.E. the robot will "Realize" that the wheels aren't spinning in high speed, but the user has the joystick all the way forward, and hence will automatically shut off the CVT to get more torque. This will be done within one pass through the main loop at max. Usually faster since our external stamp executes many passes through it's main loop for every pass through the RC's main loop.)

We really aren't doing anything special at all in our code that most veteran teams don't already do. I figured the rookie teams might benefit from it though. Anyway, it's there if you're interested, if it doesn't help, no harm done. If you have any questions specifically regarding the "what or why" of our code, just ask here.

Code with the RC tag gets loaded on the Robot Controller, 2P gets loaded onto our external stamp. And if it's not obvious, code with the most recent date tag (YYMMDD_HHMM) is the most recent.

So on that note, if other teams wanted to comment on our code, it would be most welcome! We always welcome constructive criticism, and with code there's *always* a slicker way to do things.

Take it easy, and good luck!

Finally, one last thing to mention. Only about half of our code is actually implemented at this point. Less than half even. There's a ton of empty subroutines and sections of code which do pretty much nothing. This is VERY much a work in progress, and as we proceed with programming we'll try to make it clear in comments which sections of code are for the most part complete and which aren't.

http://studentorg.eng.ohio-state.edu/~first/677/files/

[powercat]

Where did you find the Joystick dead zone whilte paper ?

[Jeff McCune]

Honestly I don't remember. I glanced in the normal places but couldn't find it. Sorry.

Take a look at our code. That piece of it is working at this point.