Thanks so much to our wonderful alliance partners: Teams 384 and 3168 at Rumble in the Roads, and Teams 1553 and 4828 at THOR.
Congrats on your wins!
One thing I have noticed though from going to 2 offseason events with our beta electronics + robot (which we didn’t win unfortunately ) is that the beta equipment can do some wonky things with autonomous mode. This year my team’s robot had a super consistent 1 ball autonomous routine; averaging our 2 regionals and champs I would say 90% of our shots in auto were accurate. When we went to our first offseason I noticed that the robot did not drive far enough forward to make the shot. (No drive length code was changed to my knowledge.) At our second offseason there were multiple beta teams such 1114 and 340 who were missing auto shots with what I think were similar problems.
Take a look at this video of 1114 and 1511 missing auto shots in Ruckus Qual 1. I know for fact that the error in our auto is that we didn’t drive far enough forward. What looks to be part of the failure of 1114’s auto here is the failure drive far enough forward (a minute distance I know.) to pick up the second ball. (This post isn’t to single out 1114 auto miss with us. They are a really awesome team and alliance partner; we were the beneficiary of one of their 2 ball routines at champs and I greatly appreciate the chance I had to work with them! It is merely to describes my observation that something with the change to the beta electronics messed with the drive lengths in auto.)
Another wonky thing that happened to us at our first offseason was after teleoperated ended we would spin in circles because one side of our drive train was still powered. This also happened to another non beta team, 250, so I assume it is some kind of FMS error.
This is NOT an FMS error from my reading of the threads on Teamforge (the place where the beta teams are posting errors and bugs and getting info). We had this happen to us as well at THOR. It is a bug with firmware for the RoboRIO and the Jaguars. They have released updates for both that are supposed to address this.
This also happened to 836 at McDonough in September, iirc, with their beta hardware.
Was the robot consistently driving a little bit short? Also, were you using the Victor SPs? If so, it might mean that the signal vs voltage curves on the Victor SPs are slightly different than those of whatever other speed controller you were previously using . Has anyone made a graph like this for the new speed controllers?
If rather your robot had a large variation in the distances it traveled in auto, then it is likely not the control system that is to blame.
Congrats! However, I’d call that more or less an advantage…that is IF you get it working properly. After all, it is meant to be a better version of the Crio which should give you some advantage over non-beta testing teams if you use it properly.
Yeah, the beta hardware has definitely been helpful. The main reason our success with it has been surprising though is because we still have a bot equipped with the cRIO and, since some of our code still wasn’t updated to fit the new functionality, we had been facing some small performance issues with our beta hardware bot at RITR. At THOR, we brought both the cRIO bot and the roboRio bot to compete, and it was just exciting to see the beta bot come out on top.
We have Jaguars on that robot.
Part of the problem probably is that many of our batteries have not been sustaining an acceptable voltage during matches.
That should not have any effect on autonomous mode though, since it is at the start of the match. Do you have any other theories as to why your autonomous was driving short?
Ah I see. That’s good to hear, and to see that the benefits of the beta system are evident. I myself look forward to using it in 2015, along with all the other new bits and pieces like motor controllers! Should be interesting.
We’re looking forward to it as well. It’s been an absolute pleasure testing this new system.
We actually finally saw loose PWM cables at THOR. It’s been something that other beta teams have had issues with but we didn’t encounter at RITR. Some tape fixed them for the remainder of THOR. I think we’ll definitely be using one of the various 3D printed retention clips that inevitably get made when it comes to the real season. We’ll also be doing a much neater job with cable routing this year to help with this as well.
The inconsistency of our auton shots at Ruckus was due to wear on the shooting mechanism over 6 events this year, not related to the beta test control system. We had it working consistently in the weeks leading up to Ruckus at our practice field with the system installed.
The default for PWM drive motors changed from Jaguar to Talon in a recent Beta software release. If you don’t account for that, your drive commands can end up being a few percent less powerful than you expect.
If you’re using CAN, this change won’t be a factor.