Building on the success of 2017 Swerve 4.0, we’ve adapted the module for VersaFrame and went with Mini-CIM’s this year over 775pros for drama free performance. Also adapted to Colson wheels for longer tread life and more consistent motion profiling.
Looks a lot like the swerve I’ve been working on. Are you guys releasing CAD this time?
Do you have more pictures or a CAD of this?
Yes, In this thread here
Nice! We love the VersaFrame and swerve combo we’ve had the past two years. This one looks amazing!
What do the slots for the mounting bolts help with?
CAD on it’s way…
Previously posted renderings…ignore 775pro’s…we are using Mini-CIM’s.
Picture of the assembly from slightly above:
Picture of the assembly from slightly below:
From what I see, it allows them to make the plates independent of particulars on their frame setup, like if the VersaFrame holes are in different places on the left or right side due to a not-exactly-on-the-inch cut. Instead, since you know there’s a hole somewhere along every inch, they just throw enough bolts at it to let them constrain the module.
That is some real slick double stacked gears you have there. A bushing on the output shaft of the motor for the final stage before the input pinion, which also acts as a collar to hold everything in place. I’m sure finding the right gears for spacing and ratios was fun. Very, very clever.
Can you talk us through your steering gear process? Looks (from the single picture) like the undercutting is a little lacking, so I’m curious if that’s from design, or production.
How did you manufacture the steering gear, and other plates?
How much does a module weigh?
Just to clarify the design a bit, the output shaft of the motor has both a 20DP gear (which meshes with the steering gear on the turret of the module) as well as a small plastic gear. These are the gears that you see stacked in the picture. The small plastic gear meshes with the larger plastic gear on the left. That larger plastic gear is mounted on a continuous rotation potentiometer (you can see the “D” shaft of the potentiometer in the hub of the gear). The gear ratio of the plastic gears is the same as the gear ratio between the 20DP gear and the turret gear. That way, the potentiometer makes one 360 degree revolution at the same time that the turret makes a 360 degree revolution. This gives us absolute steering angle feedback for our control system.
We turned down the end of the 3/8" hex output shaft from the steering motor to the right size for those plastic gears. That way, we could mount the 20DP gear on the 3/8" hex portion of the shaft and the plastic gear on the smaller portion of the shaft. The collar is a belt and suspenders approach as the plastic gears actually have set screws to secure them to the shaft. But we did not trust the set screw, so we added a collar for extra security.
We waterjet cut all the non-COTS parts for the module. The gear teeth in the turret are cut by the waterjet. We have found that it takes only a small amount of running before the gears lap in and run smooth. We used the 20DP profile for the teeth in the CAD model for the parts and then the waterjet path is offset by the kerf. The tolerance is not as tight as what you will get from a 20DP gear that you buy from AndyMark or VEX. But since the steering is relatively low speed and low load, we have not had a problem with this mesh. We used to include a small amount of adjustment (slotted holes) for mounting the steering motor to be able to adjust the mesh, but now that we are on version 4.1, we have found that the teeth on the turret are consistent enough that we do not need that adjustment any longer.
6.8 pounds fully configured for PowerUp.
Very cool, thanks for the answer!
Brilliantly set up system. I know the AM modules use a similar system to get the correct 1:1 steering rotation but the ratio is a lot closer than yours. Love the design, and keep iterating!
Update on the CAD?
Looking at the exploded picture. Your using a rectangular cut for the delrin bearings. Why din’t you use a 45 degree groove? Did the edge contact wear the balls?
Unfortunately that is just a shadowing artifact of rendering. The groove is in fact a 0.125" radius grooving tool in a parting/cutoff holder. We mount the plates to be grooved on a lathe fixture and cut the groove a nominal 3/32" deep at the apex…leaving a 1/16 float gap between plates.
See here… https://youtu.be/vE3397650W4
Sorry for the delay…Our lead CAD student just finished exams and promised them any moment now…right Evan?
It’s unacceptable that school grades get in the way of robots!..Oh wait, did I get that backwards…I keep forgetting. :o