Update of the 775pro drivetrain I’ve been working on.
changes:
Added perpendicular omni crossbeam. Going to have to play with the ratio on this one to see what gives a good balance of speed for strafing as well as not overloading the motor
I realized I don’t need to make the entire drivetrain side plates out of sheetmetal if I can just figure out how to get all the motors to power one shaft. So I made a little housing that holds the three vp’s, and allows them to be chained to a single output shaft that can then be put through a 2x1.
Added the wheels, all omni’s for right now, can put a traction wheel at the back and add a piston to jack it up when you want to strafe.
Still under 25 lbs without counting the wheels and chains.
Total height from floor is under 5" without wheel, 6" with.
Center wheel needs to have some type of device to push it down on the floor, or it will not strafe. Basically reiterating BREAD’s response, but putting it in lay man’s terms. check out 303’s 2015 bot for inspiration.
I have thought about putting springs over the bolts where the parallel omni crossbeam mounts to the side 2x1. In-between the L-brackets and the top face of the 2x1.
if you have a piston with a small wheel at the end you can jack up the back of the robot slightly which would put pressure on the parallel omni as well.
At this point I just don’t understand the point of using VPs in your gearbox. You already need to make a plate with a relatively high tolerance to hold the VPs in place. At that point, why not just use WCP’s gears or belts to make a fully custom gearbox? It will be a lot lighter and cheaper. I don’t see the upside here.
If you mount the motors directly into a plate you need to create arc-slots for airflow, and I can’t do that on a manual mill, I could make straight slots but then they wouldn’t line up with the holes in the 775.
Also we’ve never made a custom gearbox before, and the 775’s and vp’s are very easy to change out if something breaks.
You are already plasma cutting all of this, so making a couple tensioned belt reduction stages shouldn’t be too hard. You have 8 months to figure out how to make gearboxes.
Why does the center wheel have to drop? Why can’t it be on the floor the entire time? Isn’t that the idea of omnis? (we are putting this on an FTC bot for off season so I am very curious)
Because with variation in the height of the field floor you need to push the omni against the floor so it makes contact and has enough friction to be useful and move the robot.
Speaking from experience, use some sort of dynamic drop (springs, pressure, ect.), my team made the mistake of not adding a dynamic offset on our 2015 H-Drive robot and regretted it. We ended up modifying the center wheel to have a fixed ~1/8" drop (which made it rock around the center point in an odd way) and even with that the strafe only worked about 65% of the time.
The issue is weight distribution; even on a completely flat surface, if all the wheels are touching the ground, your center wheel only has 14% of the robots weight on it, meaning it will have a very hard time pushing the robot sideways (Think of what would happen if a 21lb robot tried to push a 129lb omni wheel robot sideways using a single motor, omnis or not, it’s still going to be hard to push).
The piston will help with weight a bit but the bigger benefit is compensating for uneven floors. If you want more weight on the wheels, remove the center 2 wheels on the sides and that 14% changes to 20%.
