Differential Swerve

It has been a long time working to fix all the previous issues. I think I have one of the most final versions of the model. I’ve only made very tiny tolerance revisions after the final version rebuild so I think I’m mostly stable for the CAD. I’ve been printing out everything to finally build both pods at once. I’ve compiled changes that I’ve done only on one pod or changes that should have plenty of clearance to work. Right now the final shell is printing. I’ll probably work on getting both pods disassembled tomorrow and reassembling them with brand new parts and it should all be coming together quickly. I also worked a lot on the code. I made the robot field-centric and also started using odometry to help out with the accuracy. I haven’t done anything with auto yet but I hope to do something with Pedro pathing to have an auto pathing system for my swerve. I’m pretty excited, to build the final version of my project and have a final product so this can come to a close.

I finished doing the remodel with new parts. A lot of it was updating existing parts to fix what I noticed went wrong in the test build, which made V3. A lot of it was improving the clearances and tolerances between parts to reduce friction and increase efficiency. I also made a lot of changes to the software, which is currently uploaded to GitHub. My plans for the future are:Finish making the wheel grippier and make sure the tread doesn’t fall off.Make auto pathing software.Get a full drivetrain ready.

After finding out about the previous issues in the last version, I finally finished version 2. This one uses different gears that avoid clipping and I changed a lot of the other tolerances. I also found a major thing that I got wrong in my design. Apparently somewhere along the line I dragged a component, which was messing up all the spacings and everything was out of line. I had to go back and fix all of that.

I added a lot of convenience features for me to build this easier and made the pod shorter by a good 3 mm so I could actually add some aluminum protection on the ends to avoid the pod being damaged while driving. Overall I think there are a lot of improvements. Once I ship the mod and get the parts, I can finish building this pod and another pod to create a drive base. Hopefully I can finish programming it and hook it up soon so it works.

Just posted a devlog but forgot to publish my most recent timelapse and just wanted to give a couple updates. A lot of the tolerances for parts are highly dependent. When I’m inserting a bearing into a 3D printed part, I usually do 13.9 for a 14 mm bearing and 14.9 for the flange. Because some parts print really weirdly in a weird configuration or print vertically where the layers will collapse on each other, I had to make the outer race 15.2 mm and the inner race 14.2 mm, so 0.2 mm oversize instead of undersized. This gives proper tolerance.I also had a lot of weird print issues. I’ve been having a lot of stringing and I accidentally added supports on one of the parts so I’ve been losing a lot of material but it’s honestly taught me a lot as well.I attached a couple of pictures just to show you a difference from my last post. It’s been five minutes but it’s a slight difference. It’s just to show:- what I’ve gotten done- what I’ve assembled so far- what I’ve thrown away- what I’ve missed- okay this is an update after the 2-hour time lapse of me building an entire pod. There’s a lot of things but the biggest thing I noticed was that one of the bevel gears actually clips with the main drive ring gear, which ruins the entire model. This is the biggest issue I have to fix.

Other issues I noticed were that assembling was very hard because a lot of the holes were too big and things would fall out when they were flipped over. I also didn’t have the right size screws often, which led me to have weird spacings. Mostly issues can be ironed out through CAD and now that I’ve built it once I can remember what went wrong. For the second version after I fix all these major and minor issues, I hope it’ll go much better and I can get a working pod that’s in size pretty soon for version two.

I started recording and documenting this project halfway through the first revision. The first time lapse was me implementing the final preliminary designs, getting the encoder mounted, getting all the fittings correct, getting the bearing stacks right, and mounting everything in a way that it would technically fit in CAD.

The second and third time lapses are mostly printing, getting tolerances right, and getting fits right. My first 3D build will be complete as soon as possible and I can get started on making revisions and iterations to improve the design.

I think I focused a lot on trying to incorporate everything I saw. I looked at some designs online and reviewed everything that we needed to implement before hand. I think I’ve done a pretty good job of finding space for everything and putting everything in a nice compact fashion. I think after I get all the tolerances correct and finish assembling, this will be a great first iteration to improve upon. I hope this project can continue and I will get funding for my future prospects, which are getting this machined and getting higher quality parts.