Difference between revisions of "Talk:DEWBOT XII"
From DEW Robotics
| Line 12: | Line 12: | ||
===Old school variable pitch V belt pulley solution.=== | ===Old school variable pitch V belt pulley solution.=== | ||
| − | :Different members of the team have looked at several 2 speed gear shifting solution. It became apparent that this would require a total redesign of our module. Things just don't fit in our current design. The real killer is the increase in the amount of machining required. I had thought of variable pitch pulleys for years but never found an off the shelf solution. Most available VSP are for 4L V belts and larger. Nothing would fit in our space constraint. In the spring of 2014 our supplier of timing belts and pulleys gave notice that they were dropping the timing belt and pulley line. In my search for a replacement supplier I came across the Torque Transmission company. They made the plastic timing pulleys that we purchased from QTC. In their catalog, I found a VSP pulley for 2L V belts. They also make 2L v belt plastic fixed pulleys. These parts would fit on our current swerve module with very little redesign. A VSPR10 pulley was ordered with a 8mm bore and 2mm | + | :Different members of the team have looked at several 2 speed gear shifting solution. It became apparent that this would require a total redesign of our module. Things just don't fit in our current design. The real killer is the increase in the amount of machining required. I had thought of variable pitch pulleys for years but never found an off the shelf solution. Most available VSP are for 4L V belts and larger. Nothing would fit in our space constraint. In the spring of 2014 our supplier of timing belts and pulleys gave notice that they were dropping the timing belt and pulley line. In my search for a replacement supplier I came across the Torque Transmission company. They made the plastic timing pulleys that we purchased from QTC. In their catalog, I found a VSP pulley for 2L V belts. They also make 2L v belt plastic fixed pulleys. These parts would fit on our current swerve module with very little redesign. A VSPR10 pulley was ordered with a 8mm bore and 2mm key way to match the CIM motor shaft. A 3.5" 3/8" bore pulley was also ordered. 50$ for the VSPR10 and 12$ for the 3.5" pulley. The VSPR10 pulley adds a significant cost to the swerve module. Many 2L belts were ordered and tested. The Gates belts in 18" length were found to be the best. Many, Many configurations of idlers and belt tensioners were tried. One important problem became very apparent. The first tries had several springs in the system. Primarily the .5, 1.5: stroke pneumatic cylinder and the V belt pulley spring. At different setting oscillation set up running away many times. After the initial testing, It became obvious that pneumatic actuation was not going to work. The current system as of 10/14/2015, replaces the pneumatic cylinder with a hobby servo. The first servo used was a Hitec 322HD. This servo work but barely has enough torque to move the tensioning pulley all the way to low speed. A Hitec servo just under the 2015 power limit was purchased and installed. Basically it works on the bench. To go forward it must be tested on a robot. |
Latest revision as of 21:08, 15 October 2015
Contents
CVT Swerve Project
Background
- Since 2010 1640 has fielded a swerve every year. For the last 2 years the swerve design has proven to be both durable and a strategic asset for the team. However, the 2014 game showed a weakness in our drive train. The modules have a fixed speed reduction. For 2014 it was evident that we needed both a lower high torque reduction for robot to robot interaction and high speed for sprints across the field.Several 2 speed shifting options were looked at. Team 1717 designed a 2 speed shifting swerve that was pure elegance. The cost, complexity and need for demanding machining resources made several of the designs unattractive.
My search for a shifting or CVT swerve solution was bound by the following design constraints.
- The design must use as much of our current design as possible.
- Any parts that are needed must be able to be machined in house or for very little burden on a machining sponsor.
- Cost must be minimized. Our swerve modules are not cheap and costs must be contained.
- Time. Additional time to manufacture, assemble, and source must be acceptable in the build season limits.
- Cots parts are preferred since they can be reused from year to year.
Old school variable pitch V belt pulley solution.
- Different members of the team have looked at several 2 speed gear shifting solution. It became apparent that this would require a total redesign of our module. Things just don't fit in our current design. The real killer is the increase in the amount of machining required. I had thought of variable pitch pulleys for years but never found an off the shelf solution. Most available VSP are for 4L V belts and larger. Nothing would fit in our space constraint. In the spring of 2014 our supplier of timing belts and pulleys gave notice that they were dropping the timing belt and pulley line. In my search for a replacement supplier I came across the Torque Transmission company. They made the plastic timing pulleys that we purchased from QTC. In their catalog, I found a VSP pulley for 2L V belts. They also make 2L v belt plastic fixed pulleys. These parts would fit on our current swerve module with very little redesign. A VSPR10 pulley was ordered with a 8mm bore and 2mm key way to match the CIM motor shaft. A 3.5" 3/8" bore pulley was also ordered. 50$ for the VSPR10 and 12$ for the 3.5" pulley. The VSPR10 pulley adds a significant cost to the swerve module. Many 2L belts were ordered and tested. The Gates belts in 18" length were found to be the best. Many, Many configurations of idlers and belt tensioners were tried. One important problem became very apparent. The first tries had several springs in the system. Primarily the .5, 1.5: stroke pneumatic cylinder and the V belt pulley spring. At different setting oscillation set up running away many times. After the initial testing, It became obvious that pneumatic actuation was not going to work. The current system as of 10/14/2015, replaces the pneumatic cylinder with a hobby servo. The first servo used was a Hitec 322HD. This servo work but barely has enough torque to move the tensioning pulley all the way to low speed. A Hitec servo just under the 2015 power limit was purchased and installed. Basically it works on the bench. To go forward it must be tested on a robot.
