Team 5406 2022 Robot Reveal

W robot. What were your team looking at for the climber design? What inspired you guys to go with the “windmill” climber?

We had a couple of climber designs on the whiteboard in week 1. We knew that any climb faster than ~10 seconds was probably good enough (because shooting the equivalent 8 balls in 10 seconds was probably out of reach*) and any of our concepts could probably be pushed to that speed. Given that there were multiple viable options, the team chose the windmill climb because it had the best “CR” ratio (coolness / risk) :slight_smile:

*this neglects a bunch of details about line-up time, and timing climbs with other alliance members which would make a faster climb valuable, but the long and the short of it is a 10s climb was our goal.

5 Likes

Just in the nick of time before our next event, check out this recap of Celt-X’s experience at Waterloo. Thanks to team 4678 and 4917 for building a great alliance. Thanks also to 2056, 610, 4069, 2200, 4907 and 1114 for some exciting playoff matches!

We’re looking forward to competing again at Windsor tomorrow!

7 Likes

5406 2022 CAD Release

Despite not getting to our shop until the end of January, and having an almost entirely rookie crew after COVID, Celt-X had a pretty great season in 2022! We’re very proud of our robot Tails, which had a reliable 5 ball auto, An OPR above 50 points, and an top-tier 7 second climb.

We may make some small improvements over the summer (in time for IRI), but otherwise Tails’ design is pretty much done. So this makes a good time to release our CAD, links to which you can find below. Feel free to ask any questions!

Solidworks Native and Step Files:
https://workbench.grabcad.com/workbench/projects/gcX8aHvArM7rSATL1ufR_TEycSc5gmg9I098dgOlp-7M4C#/space/gcSiKzBrTual35TxoJocXPFyZfEGxrVAFOpoG5Ep8LxPYb/folder/12297393

OnShape Import
https://cad.onshape.com/documents/f7b69a94a5d85ab0af5bbdcf/w/1e6493876e86aa31e870845e/e/e9bf5cdb599f90d0788f48fa?renderMode=0&uiState=6292daad32f8a44f54d72857

16 Likes

5406 2022 Code Release

To go alongside our CAD release, here is the release of our 2022 Codebase.

This year, our robot, Tails, features:

  • Five, Two, and one Ball Auto modes
  • Full Automated 8 second climb to traversal.
  • Vision Tracking using a limelight.
  • A mix between Timed-Robot and CommandBased Programming.

A massive shoutout goes out to our programming subteam who worked on our 2022 codebase, including:
Christian F., Kayla S., Cordeila K., Dylan V., Maja T., Colin M., Gurshaan S., Victoria A., Chris P.

4 Likes

Looking at your turret assembly, it would appear you’re using 2 UltraPlanetary gearboxes this year. One for a NEO550 to power/turn the turret, and another that gets backdriven and has an SRX Mag encoder on it for positional feedback.

Is this indeed the setup that was run on the comp robot, and if so, can you shed some light on the reasoning behind this decision/setup?

1 Like

Good eye. Yes, we use the second UP to turn a mag encoder (in absolute mode). The secret is that each of the stages in the encoder gearbox are flipped so that they reduce the speed of the magnet (rather than increasing it) when back driving. So even though the turret ring gear has many more teeth than the encoder pinion, the magnet on the gearbox “input” never turns more than one revolution.

We use the encoder gearbox to “zero” the other neo 550’s incremental encoder at startup, so we always know where the turret is pointing without a hard stop or having to home the turret manually before the match. Does that make sense?

3 Likes

Took me a while to track down a photo. Here’s the same geared encoder assembly on our 2020 robot. It’s a pretty easy cheap solution to the problem. I’m proud of the student (since graduated) who designed all the 3D printed adapters.

5 Likes

Seems to make sense to me! Was there a reason you went with this method rather than use a hall-effect sensor + magnet (such as a REV magnetic limit switch) to zero the turret?

The UP gearbox and MAG encoder at first glance seems like overkill and extra steps, but perhaps you found a hall-effect sensor wasn’t accurate or repeatable enough for such an application.

With a limit switch (or hall effect switch) you have to turn the turret some unknown amount of angle at the beginning of auto in order to find it, before you can zero, and then point the turret where it needs to be. That’s time you could be shooting. With an absolute encoder you can zero right away in whatever position the turret starts in. A 15 second auto runs out pretty quickly!

1 Like

You could swing your turret through the magnetic encoder in disabled and verify on the dashboard.

1 Like

Yep, for sure. But with an absolute encoder that’s one less thing to worry about pre match.

1 Like

The curved cable chain is an extremely cool idea! Why didn’t it make the cut?


Took this photo at competition of the cable management. Love how the constant force spring is packaged vertically. What is the cable sleeve you use?

7 Likes

The curved cable chain was very cool, but sometimes you have to walk away from a cool idea when a better one comes along. The sleeving-fairlead arrangement pictured by @Smtopps was lighter, stronger and packaged smaller.

We used cable sleeving instead of cable chain in 2018, 2019 and 2020 for the same reasons. This is the product we use: https://www.mcmaster.com/9284K615/. It cinches down on the cable bundle under tension, turning it into a nice, flexible “rope”.

6 Likes

is the cable sleeve attached to the shooter with just the two zip ties?

Yes, but they’re green so they’re stronger.

6 Likes