FRC 4488 Shockwave - 2019 Robot CAD and Code

On behalf of team Shockwave, we hope everyone’s having a wonderful end of year.

Below is the link to the CAD for our 2019 season robot, Supernova:
GrabCAD link

And here is the code for 2019:
GitHub link

Best of luck this next season. Please let us know if you have any questions.



I have not been able to download the cad yet however. Just from what I’ve heard you used a stretched belt as the “teeth” of your turret. How did you calculate the center to center and the inside diameter of the belt. Also how well did your belts mesh. I just grabbed some belts and tried it and the belts didn’t seem to mesh the best.

For the stretched belt, you might be thinking of 1425. First of all, we didn’t calculate much, we ordered a 20in bearing and a large belt that just happened to fit over the bearing with a lot of clamps and manpower. To drive the turret, instead of using another belt around the outside, we used a custom made 54T delrin pulley with no flange. We had said pulley slide so we could tighten the meshing to account for any errors in calculations of the pulley size.

Towards champs as we were trying to speed up the turret, we found issues with the pulley skipping teeth during hard acceleration or deceleration. If we had to do a turret again, we would now use a bearing sandwhich or large X-contact bearing driven by a large custom made sprocket or gear. Belts are bad

We too found that two belts how you described, do not mesh properly which why we went this route.

Feel free to ask any more questions.

-Daniel Kassab Driver 1425

i’m intrigued…
i’m just curious, but why not just 3d print a large, custom pulley? how come making a custom gear or sprocket was avoided originally?

Making a custom gear was avoided because we did not have the capability to make one. Although our sponsors shop does have a water jet, they have no operator. Larger tooth profiles like 10 or 15DP would probably have a decent amount of slop. A custom sprocket and 3DP pulley were avoided because they would have been harder and we were running out of time with our way over complicated turreted elevator.

Team Shockwave also used a belt turned inside out to be the teeth on the turret. This was the Armabot Turret254 design. We based the pulley distance and tensioning system on the bracket that came from Armabot. What really helped the teeth fit together well was using two different belt tooth patterns. I believe the inverted belt was HTD5 and driving belt was XL5. The different tooth patterns on the belt matched up well when the HTD5 was inside-out and stretched tightly around the turret.

Your vision code is cool! I love to see other teams writing their own stuff instead of using a Limelight. Have you thought about using GRIP? I personally have had a very good experience with the software with making the initial pipeline.

Some of the programming students wanted to learn about vision processing code, so rather than using GRIP, they decided to use OpenCV in Python with a JeVois camera. We already had some previous OpenCV-based code in 2017 (based on team 254’s cell phone vision system at the time). This season, now that they understand the fundamentals, we are going to give the Limelight a try to see how it stacks up during build/competition season vs rolling our own OpenCV code like last season.

This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.