This is a CAD (AutoDesk Inventor) exercise with a purpose. The purpose is the creation of an updated drive-train lesson for the team. Our key learning during the 2010 season was with our first pivot drive robot. Each (of 4) Pivot drives was independently driven and steered. The team was very pleased with this design’s performance.
From a CAD standpoint, this was my first experience making cut-away views for improved illustration & education. Cross-sections are normal (for engineering drawings), but this had a different mission. Results are quite okay.
I like a lot of things about your modules this year. Two chain reductions instead of a transmission, the triangle gussets on the sides of the module for stability, the dead axle with extra nuts instead of spacers at the bottom, and using an axle as a structural member of your module. Nice job!
I am very surprised that you did not have trouble with the cantilevered side-load on the Denso motor shaft. The worm gear in that motor is a plastic gear captured only by the thin sheetmetal lid on the housing, and is not designed for any side-load whatsoever. You might want to consider supporting the end of that sprocket if you can. We have found that a 5/8" OD ball bearing fits nicely into the center hole of the plastic Denso adapter to provide end support.
I very much like the chain reductions and overall design of your module. Very similar to our own.
Inventor was used to model the whole robot. The chains are #35. We have not had any problems with the steering motors from the sprocket not being supported on 2 sides. We have had major problems with the Denso window motors. The locking pins caused some serious drag when not run at full speed. We removed the locking pins and the denso motors worked much better in this steering application. Even with the pins removed, we had random motor shut down. We removed the jags and substituted victors. Since we went to victors we have not had any problems with the Denso units. Denso and jags do not mix. The over all design has worked very well. However we are working on several design tweeks in the off season. One thing that gave us problem were set screws constantly loosening. The next iteration will not rely on sets screws.
That observation seems worthy of a tangential discussion. We haven’t used window motors for anything lately, so I can’t comment, but I’m very curious. Are there any other teams out there who had a similar (or contradictory) experience using them with Jaguars?
There is a thread on the Jag and Denso problem. http://www.chiefdelphi.com/forums/showthread.php?t=83973
We have not responded to this thread since removing the locking pins. Even with the locking pins removed, we experienced random Denso shut down. For IRI we swapped out the jags for victors and the problem has disappeared. The window motors are great without the locking pins and victors. We have no hard data to help with why the Denso’s overheat while under very light load with the jags. I think it has something to do with the high switching frequency of the jags interacting with the PTC causing it to self heat. I would like to see First allow the removal of the locking pins. Its easy and improves the performance of the motors. However, The bypassing or removing the PTC is much more questionable. It’s an actual modification of the motor. So if you use the window motors, use victors.
We have not had any problems with the cantilever loading on the window motor sprocket. We are still using the original (4) window motors after (6) competitions (on & off-season) and (4) demos. We’ve got (L&R) spares, but have not needed to replace any originals yet.
We embed a magnet in the end of the sprocket. A Cherry absolute magnetic encoder is mounted on the plate below the sprocket. This encoder provides information on the sprocket angle and (since the drive sprocket on the motor and the driven sprocket on the pivot are both 15T) therefore the pivot angle. This makes mounting the bottom end of the drive sprocket problematic.
Regarding building chains in Inventor, I have a process which works well (albeit a little tedious). I will write this up (it’ll be short).
