Low Cost, Traction Swerve

What is your slip ring rated for?

We can’t totally know for sure until we run this on the ground. In the test bed, it actually looked liked it snapped to it’s heading fairly well. But, I do agree and that why we’ve redesigned to include an SRX mag encoder. Should have that prototype up and running in about a week is what we’re hoping.

The gear is not a gear but an encoder disc laser cut. Laser cutter has an accuracy of +/-0.005". It was our IR sensor that required the slit to be 0.050" wide. Thus we only got the 2 degrees of resolution.

It’s a 4 channel slip ring with each channel rated for 30 amps.

We got it on Aliexpress, which is now my favorite shopping place.

https://www.aliexpress.com/item/Wind-Power-Slip-Ring-2-3-4-5-6-Channel-5-10-15-30A-Rotate-Dining/32888736198.html?spm=2114.search0604.3.106.37a62aa20Mzf1E&ws_ab_test=searchweb0_0,searchweb201602_4_10065_10068_10130_10547_10546_10059_10884_10548_315_10545_10887_10696_100031_10084_531_10083_10103_10618_10307_449,searchweb201603_16,ppcSwitch_5&algo_expid=8f98165f-90a8-4245-82f5-0cc7d05655c0-15&algo_pvid=8f98165f-90a8-4245-82f5-0cc7d05655c0&transAbTest=ae803_1&priceBeautifyAB=0

Yes it would be 3D printed. Correct me if I’m wrong, but I think because of the involute nature of a gear, it wouldn’t matter what pitch one chooses. I could see how practically this may not be as good as theory though. Given that there’s no load on this part, I would feel comfortable starting at 20dp at least (currently 10.)

With the involute tooth profile on a large DP gear you are engaging further from the line of center. This is reduced by either:

  • increasing the dp to bring the engagement point in closer to the line of center
  • switching to a cycloidal tooth pattern (a bit passe
    ) to align the tooth engagement with the line of center

In theory this all would provide better efficiency at the cost of max torque transmittance, but if +/- 2 deg is the target window and it’s comfortably being managing that I wouldn’t worry about my little rant in this post :rolleyes:

Does this matter in practice? Probably not a whole lot, so long as you are within the “realm of reasonable” and take care in making the parts.

Latest Update

Built 4 of these and put them on a chassis on the ground. Initially had 775’s with 20:1 gear ratio to a 4" diameter wheel. The gear ratio was a bit too low and we were stalling out the motor or it was spinning up way too fast. It proved really difficult to control. Still working on different control algorithms. But decided to re-gear the drive to a 80:1. Tons of torque but really slow so not really all that useful in competition but it proves a few points. Most pressing is its ability to climb over obstacles. Here is a video of the chassis climbing over 1.5" angle iron:

correct me if I am wrong, but if you were on a slanted surface, could you only go straight up the surface because of the weight of the robot? I could see it working on about anything under 15 degrees.

Now that you’ve got a chassis built, how well does it handle uneven surfaces? I know Stryke Force has had to implement mechanical positive traction to improve steering / create higher-performance module responses on all four corners with a fairly standard architecture - this design seems even more vulnerable to loss of traction for any reason…

I’m not sure where you get 15 degrees. It seems to handle climbing over uneven terrain surprisingly well. It conforms to the terrain it seems. At least it does at low speeds. Climbing a ramp should be no different.

15 is just a guess. a ramp is different because all of the modules are at an angle not just 1-2. would be interesting to get the robot on a large ramp and see the performance though.