Team 1259 Experience & Review- (Swerve Drive Specialty modules) off season swerve drive project

As a team, we’ve wanted to dip our toes in the swerve drive realm for a few years now and with the 2020 season ending early resulting in our 2020 funding rolling over to the 2021 season, a replay of the 2020 game for the 2021 season, as well as the BOM removal; we thought it would be perfect to make this dream a reality.

In this post, I would like to share our teams’ experience and some difficulties we had as we started on our swerve drive journey using the Swerve Drive Specialties MK2 Swerve Modules. Starting off with the initial purchase, we had been watching for the SDS modules to get back in stock once it did the processing time was fairly short about 3 days from initial order confirmation to being marked shipped.

When it arrived we promptly shredded the box to bits like a child on Christmas morning. From the initial package (Standard kit all steel gears) & motor spaces, we were happy with all the parts and the quality of the machining. Some things to keep in mind are the extra pieces of hardware you’re going to need, like the Vexpro wheels, absolute encoder, etc. All of these things can be found on their website but in addition to this, our team (now having to work off school campus) have decided to make our own BOM of the things we would need for assembling and build the drive train frame; trying to mitigate any issues we might run into during the actual build process.


After we finished staring at all the shine aluminum parts, we started our assembly process with guidance from the MK2 Assembly video. We did have to pay careful to the length of the fasteners that were being used as we opted to buy the motor spacers instead of cutting of .3" off all 8 of the motors, as there was no separate assembly video using the motor spacers. Although it wasn’t too much more difficult than if we were to cut the shafts, it did certainly required us to slow down and ensure we were using the correct bolt length, either the bolts from the module kit or the bolts provided in the motor spacer bag.
Side note: I helped to assemble a pod with two other of my peers and we though (2-4 people per module) was a good size to get everyone involved and staying involved during the entire assembly process.

As we made our way throughout the rest of the video we did forget to add Loctite on various occasions as we were overly concentrated on the actual assembly. I would recommend to other teams that they should watch the assembly video a few times beforehand to speed up the assembly process and to have a better understanding of how this swerve module works and how to repair/diagnose any issues if any arise in the future. I would also recommend doing a dry assembly to make sure you understand how everything goes together as well as serving a learning opportunity for the underclassmen without having to constantly use Loctite. Even after assembling our first pod, we did manage to mess up the assembly of the second pod in 1259 fashion.

Which brings me to another point, DON’T EVEN THINK ABOUT USING RED LOCTITE! The video also tells you to use blue Loctite and for very good reason. If you mess it up, it might mean a possible disassembly of most o the pod which would otherwise be a pain if red Loctite was being used.

The one major issue we ran into.
After finishing the first two pods we had a mentor point out that there was an binding issue when rotating the VexPro wheel. He pointed out that there was a noticeable amount of Resistance at certain points of the wheels rotation. The first module experienced about 80% noticeable friction/binding and the remain 20% of the rotation was perfectly fine. The second module experienced the opposite, 20% noticeable friction/binding, 80% perfectly fine during its rotation. After trying to troubleshoot the issue we found the culprit of our wheel binding. Upon further inspection of the first two pods, we noticed that the large bearing that rotates the entire wheel wasn’t entirely flush with the surface of the module plate but they were fully seated in the bear groove. We verified that the 5 screws holding the large bearing were fully bottomed out. This now uneven surface caused the two “towers” that screw down to the main model plate to not be entirely flush when fully screwed down. Although this wasn’t noticeable upon any visual inspection, it caused the tower to not be exactly perpendicular to the module plate.


The bearing and main module plate


The location of where the bearing interfered with the “towers”

Solution
A remedy to this was to add a .001" (aluminum flashing) shim to correct for the slight angle in the “towers”. Once added this solved the binding issue that we were having for the two modules as well as preventing any binding issues in the last two modules we built. We doubt the shim solution would impact any future reliability or wear issues but remember I’m just another kid on the internet so attempt at your own risk. We don’t know if this was a batch issue or if it was just our specific kit but as a team, we’re curious if anyone had this very issue? If so, did this solution help you out? We would love to hear your experience with SDS MK2 modules.


Picture of the shim we used

Summary
We are happy with this kit and would recommend this to others that are interested in pursuing swerve drive in the future. Just remember to take your time, enjoy the process, and to get as many of your peers involved in building a swerve drive. This definitely was a fun start of our first 2020 offseason project as I go into my senior year and eventual and the 2021 game.

8 Likes

When there is a periodic bind when rolling the wheel it is almost always happening with the bevel gears. This is caused because the screws attaching the 60T bevel gear are not evenly torqued. The tighter those 6 screws are torqued, the more the 60T bevel gear will actually defrom the plastic of the versa wheel and emboss itself into the wheel. If the screws are torqued unevenly, the bevel gear can seat slightly off kilter and cause the periodic bind. The nominal design actually accounts for the deformation of the versa wheel, however, this can result in more aggressive than ideal bevel gear mesh if the screws are looser than expected.

The SDS billet wheel eliminates the variability of mounting the 60T bevel gear because it seats very repeatability on the metal wheel due to it not deforming when the screws are tightened.

From your pictures it looks like the large rotation bearing has fully seated in the main plate. I do not think there is an issue with the bearing or main plate. I believe by adding the shims between the wheel mounts and base pulley, you have effectively reduced the gear mesh of the bevel gears slightly to where the wheel can spin freely through out the full rotation.

You should not need the custom shims for the module do spin freely. Please reach out and we are happy to get this issue resolved. If there are any manufacturing defects in the parts, I would be happy to send replacements.

8 Likes

That makes sense, I appreciate the response. We’ll be sure to try and reattach the bevel gear and the wheel when our team meets up. Again, great product!

1 Like

The first set of MK2 modules we bought in 2019 moved perfectly smoothly. The second set bought in early 2020 had binding issues with the wheel drivetrain. It turned out that the wheel spacers were slightly too short. @PatrickW helped us identify the problem and sent out replacements right away. Our second set of modules run smoothly now as well. I assume the wheel spacer issue has been resolved in recent production kits, but tossing this out just in case it might be of help.

We ran the Versawheels in competition and broke every one of multiple sets on the generator boundaries. We didn’t cross the boundaries at very high speed or very often. Our bot was 118 lb without battery or bumpers. They also wear extremely quickly. If we hadn’t been breaking them, we probably would have had to replace them just as often for the tread wear. We could not “afford” the billet wheels because we were right up against the BOM cost limit after already optimizing for low and no cost items and taking some things off the bot. However, with the BOM constraint relieved for 2021, we will almost certainly be moving to the billet wheels if the boundaries remain for the 2021 game. We had modified AndyMark Hi-Grip wheels to fit and were planning to stress test those for use during the remainder of 2020, but we should no longer need to go through with that low-cost, somewhat painful option.

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