what do you think about my swerve?

Hello everyone
I’m currently working on swerve drive
I wanted to hear the opinions of people with experience.
I would appreciate your advice
Thank you


https://grabcad.com/library/swerve-drive-cad-2

1 Like

I havent had a chance to open the CAD file but how are you attaching an encoder to the azimuth?

Also, I feel like you could probably cut some weight from that top motor mounting plate.

Looks good though! I can see the inspiration taken from 2910.

That one of the problems there is at the moment.
I think I’ll connect an encoder with a printed gear to the part that rotates the entire model

Great design! Some things to think about (if you haven’t already):

  • You may want a thrust bearing above the smaller bevel gear to reduce friction with the bearing it’s constrained to.
  • What kind of machining capabilities do you have? I see a belt that doesn’t seem to have a tensioner. A lot of teams do this but you will want to make sure you have the right tolerances.
  • I haven’t looked at the cad yet but it doesn’t look like the motor mounting screws are easily accessible. You may want to make some holes on the bottom plate as well so you get get an Allen key through to the motor mounting screws if you need to remove one. Basically, try to think through how you would assemble/repair the module in different situations. Drivetrains need to be as robust and easy to repair as possible above all other requirements.

Good ideas. Do you have a way to check that the assembly and dismantling of the system is good enough?

Which bearing did you use for the azimuth? Also, why did you decide to use a BAG Motor + VP for the azimuth? I wonder because the 2910 module uses a NEO with 18:1 reduction that seems to be enough for them and the 1323 (WCP SS Swerve) module uses a 775pro 56.67:1 reduction, both result in around the same unadjusted rotational velocity of around 2000deg/s according to JVN in full throttle. Moving from a VP to something like a NEO with that reduction can probably save some weight.

Also, do you actually plan on building this? If so, how do you plan on getting the SDS bevel gears (or something equivalent) in Israel? I’ve been wanting to get them as well but there isn’t a way to do that AFAIK

The Bering we are currently using: https://www.swervedrivespecialties.com/collections/mk1-module-parts/products/bearing-3-5id-x-4od

We are still looking for a cheaper place to buy it

As for the bevel gears we will first try to produce it in our CNC \ 3D printer

If this is not accurate enough, we will wait for one of the members of the group to fly to the United States and we will buy it there.

One of the main reasons we use BAG is that we already have at the workshop and it reduces the cost of the project

How much does the module weigh currently? What’s the unadjusted linear speed of it?
It seems that I forgot to write that in my first reply but the module looks really cool! It’d be awesome to finally see swerve on the field in Israel

Right now we’re working on weight loss holes (I guess we’ll post a new version of this CAD later)
Before the weight loss holes we evaluated and weighed a little bit less than 2 kg

Have you talked to @PatrickW about the possibility of shipping to Israel? SDS is a pretty small company, so you might be able to work something out for a special international shipment.

I have my doubts that 3D printed bevel gears will stand up to the rigors of competition as the last stage in a swerve drive. It might be a good off-season project to test it though. If you can find a good local source for bevel gears (even if they need a bit of post-machining for FRC use), I’d love to hear about that.

2 Likes

+1

1 Like

I updated the files in grabcad
Right now the weight is 1865 grams (according to SolidWorks)

What I am doing for the swerve drive I am working on right now is just making an in-depth assembly guide. The idea is that if you can explain exactly how each part goes together and in what order to someone else, then your design is probably sound in that aspect. Also, it helps to have other people look at your design and assembly instructions to proof read it. Definitely make sure the people who will be manufacturing this or working in the pits approve of the assembly/maintenance process you have planned.

1 Like

What are the pros and cons of this and the SDS module? Looks very similar. Also, Alibaba will get you the bearing for cheaper if you RFQ. Be prepared for lots of spam email, though.

Looks pretty good! However, how do you plan on securing the large bearing to your main plate? I think you’ll need to be able to constrain both the inner and outer races of the bearing to your robot. If you look at 2910’s swerve as an example below. The outer race is constrained by the “ledge” in the main plate on one side and the button head screws on the other. The inner race is constrained by the “ledge” in the machined pulley on one side, and the wheel forks on the other side.


That’s what you mean?

Yup, also don’t forget the ledge/lip in the base plate that the screws hold the race against, I can’t tell if you have that in your model. Scribbled in red here:

2 Likes

now i am working on placing the encoder for azimuth
At the moment I placed this encoder:


But I understand that this encoder is also recommended

What do you recommend?
I’ve just connected it with a conversion of 44:20. The conversion that the sensor should get is 1: 1. The gears are 3d printer, and the gear is pressured on the axis of the encoder
Will it work well enough?

Why not use the encoder that fits in the planetary stack? We’ve always had trouble side loading an encoder with gear like in the picture.

The problem is that you want the encoder at 1:1 in order to have absolute knowledge of the module’s position. It’s possible to use a limit switch and a quadrature encoder, but that requires that you spend a second or two zeroing at the start of the match.

1 Like