I recently set up PIDs with the IME which is pretty accurate, but want to move up to using tracking wheels. I only have access to 4-inch Omni wheels and the old 3 wire encoders. I wanted to gear up the tracking wheels because we only have 4-inch wheels, but this would introduce slop into our system because of the vex gear tolerance. Does anyone have experience using the 4-inch Omni for tracking and some tips? In my testing without changing the gear ratio, it was accurate but could be more precise. I only have two meetings next week with the field before our next competition. I don’t want to waste time there experimenting. Does anyone have any advice to make my tracking wheels more precise?
My advice is with 2 weeks, don’t implement tracking wheels. Last minute tracking wheels are worse than a well implemented internal encoder setup. Tracking wheels benefit from smaller wheels(2.75") and lots development time. Do some research and development with them after competition. I find they can be used to make some incredible autonomous runs.
Can you elaborate on how this fact alone justifies a gear reduction on your tracking wheels? Is it because the wheel will rotate less per unit of displacement due to the larger circumference? In which case, right on –– you can gain some resolution with a reduction but your tracking wheels will be even more massive than they already are.
Completely disagree. The math doesn’t change, only where you’re receiving the position data changes. You get dramatically improved autons as a result in virtually every sense. Tracking wheels don’t necessarily imply odometry. If @BowRow was asking how to implement odometry in two weeks, well I wouldn’t even encourage a veteran team to do that.
As for advice on improving resolution, well you can’t without a reduction or without using smaller wheels. If you were using two tracking wheels to calculate the angular displacement of the robot, then you could spread the wheels out as far as possible from the center of rotation to increase the resolution (think about the geometry of that change). Maybe I’m missing something, though.
TL;DR use a reduction, nothing wrong with that other than space efficiency and arguably slop
I set up the tracking weels already and they work slightly better than IME, but my only limitation is the old encoders, but mainly the 4-inch wheels. I’m just wondering if anyone has tried gearing it up, and had it work better.
Oh and it has been done before –– took forever to find this image
So I guess I have been eating a little too much odometry pudding. I forgot to consider there are other uses to tracking wheels. If you are wanting to track distance then tracking wheels make sense as they are not affected by wheel slip or object impact. for linear measurement, a 4 inch encoder wheel can work fine as you about .03" resolution from the wheel in an ideal world based on a quick napkin calculation. Whatever you do, don’t try implementing odometry in 2 weeks. It will probably not work well with that short of a time frame.
Thanks man for the help, thats what I was asking about. Im researching setting up an odometry system, but am not ready yet. I just was looking for tips on improving the tracking wheels. I think he thought I was talking about full on odomtry. Thats what a lot of people think about when they hear tracking wheels.
Hey, were those thin roller wheels discontinued? Im a senior in highschool, and started when v5 came out. Ive only seen them on a couple bots.
Thanks for the tip, thats all I was wanting to do with the tracking wheels.
All good, I could’ve easily made the same mistake
Those are 3.25" Omni wheels cut in half. I would recommend you do the same on the 4" wheels so you not only get two tracking wheels from one full size wheel but you also cut the size of the wheel in half so you can fit a gear within the footprint of the full wheel.