pic: Swerve Shifter



Swerve Shifter I designed as an off season project. It uses the same “guts” as the VEXpro Ball shifter and drives a VEXpro high traction 4x2" wheel. It has an output of about 12 and 5 feet per second. It is driven by a cim and uses an Andymark PG71 gear motor to turn. I’m open to any suggestions or comments.

This looks like a good start! I didn’t even think about shrinking my Swerve design by adding the ballshifter!

Is that drive chain exact c-c? If it isn’t, you’re going to need to look into tensioning…

Or some GT2 5mm timing belt with C-C :wink:

the only thing that concerns me is the module itself doesn’t look like it will survive missing two of four sides, butt you may have suppressed them to show the innerds. Either way looks great love the ball shifting to narrow down the gear box.

Is the whole weight of the robot supported by that one bearing? TBH that’s absolutely terrifying. I would’t be surprised if the jerky movements of the bot make that bearing pop right out.

How is your turning pulley mounted to the module? In this render it appears to be mounted to the drive shaft, but that can’t be right.

What’s holding your miter gears and sprocket in place? What keeps the gears from sliding on their shafts, and what keeps the top gear from sliding down?

How do you plan on making the turning module frame? Is a solid piece of metal or do you plan on making it in multiple parts?

Anyway, great job on getting the whole shifting gearset to be single stage!

-Adrian

Looks pretty darn good!

Personally, I would move the module rotation belt and stuff inside the housing, rather than hanging below like it is. That will allow you to put a thrust bearing between the module and the lower plate of the housing, which is pretty essential if this is going to actually support a robot.

I’d suggest taking a look at 1640’s swerve designs. Their design is very effective on the field, and they have been good enough to release basically everything you could ever want to know about it online, including CAD models.

I’m not saying “Scrap your design, go with theirs” but a lot of the challenge of swerve drive is finding the issues that you can’t forsee in CAD, and fixing them. And they’ve already done that part for you :wink:

Few comments:

-You need a sensor location for both the steering and drive. An absolute encoder is preferred for steering, team 1640 has excellent information on such sensors. It looks like you retained the US Digital drive gear from the ball shifter.

-You need a lot more support for the entire module. Specifically:
–Thrust surface on top
–Something to react module torque on the pivot - either a needle bearing on top, large bushing surface, or a delrin ring around the bottom which reacts to a chassis member.
–The drive axle should react no force except the torque it is driving

-The drive chain is fine exact c-c, you will have no problems with a chain of that length and sprockets of that size.

As of now it is, I realize that should be changed and will find some way to do so.

There are three holes drilled into the pulley and the module. It will be bolted together so they both move.

This was one of the problems I faced originally. My best guess is super gluing the key shaft, but I really do need to find a better solution.

Sadly, I do not have the resources to make this, but if it were to be made I would either use one solid piece or make multiple parts and try welding it together.

I like the use of the ball shifter in this application, but I’ve got some questions

It looks as if your drive shaft (3/8" of 1/2") is your pivot axle and is taking all of the axial, radial and shear loads. I would expect this shaft to fail in use.

What bears the weight of the robot (the axial load)?

How is the rotating pivot cage secured to the vertical drive axle? This will be a high-load joint, but I don’t see how this is done in the drawing.

A clever approach.