PIC: Looking for feedback on our drivetrain design

Yup those are just 2x1 aluminum with lots holes drilled in them (on 0.5” centers).

1 Like

Do you think linking the two gearboxes with churros would mitigate this issue? What about putting a strut between the two side rails if we were to keep the outer rails? The extra space in the middle is ideal but you make a good point about the stress on those gussets.

You can mitigate the issue, but it will continue to add weight and complexity compared to just a bog-standard Versaframe chassis like on the VexPro website. The more parts you add, the longer it will take and the more expensive the chassis will become.
If you’re not going to build a traditional WCD, just keep using the kit chassis and save the money for a better CNC, IMO. Maybe try making a standard Versaframe WCD in the off-season to see how you like it before making changes.


Thank you for your advice. I think we will take it. Me might design and test two inner rails. One that is full and maybe use a brushless more compact gearbox and the other similar to the current design with the jog so we can pressure test it and see what kind of damage the gussets face if any.

If you do decide to go the “true” WCD route with the gearboxes on the end, I suggest considering an “outside-inside” belt routing scheme. That is, have the belts from the gearbox to the middle wheels run outside the framing members, and the belts from the middle wheels to the other end run inside.

This is quite space-efficient and, most importantly, extremely easy to service. I can’t remember which team I first saw do this, but we’ve been doing it for a few years now on 449 and will probably continue to do so in the future.

I’m agreeing with most of the feedback, for slightly different reasons.

KoP pros:

  • It’s the easiest thing to assemble for FIRST. Almost any issue you have assembling it will be more likely to happen with a custom chassis. (The most likely exception is installing a gear backwards; when building an AM gearbox, be sure to put the gear bosses towards the bearings.)
  • It’s quite tough if completely assembled (including interior churros or other bracing which performs the function).
  • There are a significant number of COTS upgrades available. Most of these are from AndyMark, but the ThriftyBot 3 motor upgrade @Akash_Rastogi linked above is (IMO) the most elegant.
  • The width is fully selectable in 1/2" increments with no modification other than cut-to-length.
  • Three different wheel sizes (4", 6", 8"), and five gear ratios are available as fully supported COTS. With some creativity, you can do even better. (The main things to know are that the gears on each stage must add to 64 teeth, and that the cluster shaft is 3/8" and the output shaft is 1/2".)
  • Even sticking with front/rear symmetry, there are 6 different wheelbases available for 6WD (12.8" to 25.2"), all fully supported by COTS parts. Drop to four wheels, and you can make it as short as 6.4" (if you can figure out how to fit everything). I’ve built a robot on the KoP chassis with a wheelbase less than 8".
  • By default, provides two gear-driven wheels at the center of the robot, and four belt-driven wheels with no double-loaded belts. (Lose one belt, lose one wheel.) Even if all four belts break, the robot can still drive.
  • I literally saw a robot with swerve drive mounted in a KoP frame this year at Arkansas Regional. There must be more benefits. (I didn’t inspect that robot, but it didn’t look like this was a BOM/CAWST driven decision.)

WCD pros:

  • Cantilevered wheels allow you to move them out to just shy of the frame perimeter. This both reduces scrub forces and increases tipping stability in turns.
  • There is the option of placing the chain/belt inside of tube, totally protecting the chain/belt and gaining a bit of working volume between the tubes.
  • Dimensions are what you decide they are. (provided you can fit things in, of course!)
  • A great number of gear ratios and wheel sizes are available COTS, and basically whatever you want if you can manufacture gearboxes and wheels.
  • When designed as usual with one gear-driven axle near the middle, no combination of chain/belt failures will disable the drive system.

KoP Cons:

  • If you cut something wrong and didn’t buy a spare, fixing it is totally dependent on AndyMark shipping time. For me in southeast Louisiana, that’s only two days ground, but YMWV based on location.
  • As normally built, the outside edge of the wheels are ~2" inboard of the frame perimeter. This increases scrub forces and the likelihood of tipping in a sharp turn.
  • Some people report bending. I personally don’t see how this would be worse than a WCD unless the WCD used unusually thick tube, but it’s been reported.

WCD cons:

  • Cantilevered wheels can be a bit iffy. This is easily mitigated by using wider tube and moving the wheel closer to it.
  • When the wheels are really close to the frame perimeter, interference from the bumpers can become a concern, especially in collisions.

I may have missed a few, but these considerations led my low-mid resource team ($15-30k annual budget, excluding build space) to use a KoP chassis frame most years - and I consider the two exceptions to be a mistake understanding what went wrong our rookie year.


Word of caution: the Thrifty Bot 3-motor kit is just the inboard plate since it was designed for the AM14UX drivetrain. You may be able to figure out how to slap it together, but it may not be worth the hassle.

I think everything else I would otherwise say has been said here.

As it happens, I have an unopened U4 drivetrain about 8 feet behind me that I purchased for a local workshop I couldn’t attend due to a scheduling conflict. I have ordered a pair of the θB3 kits tonight (as well as some other stuff) and will include them in my fall 2020 workshop presentation, and report on the ease/difficulty of inclusion at that time. Based on my research done a few months ago, I expect this to be easy-peasy.

What are you referring to here? Unless it’s a typo, I don’t think I’ve ever heard of it.

Thrifty Bot 3 motor kits. I thought it was obvious given what I was replying to.

1 Like

From reading through this thread, it sounds like you will not be building this design as show. Still, it may be worth making this comment since you might run into this situation in the future with some other robot or robot part.

There is clearly a bearing on the outer side of the outer tube and one on the inner side of the inner tube on each side. It is not clear if there is a second bearing on each outer tube and inner tube (4 bearings per shaft) since the wheels and pulleys obscure that part of the tubes. If there are more than 2 bearings on a shaft, it will be “over-constrained” and will be difficult to assemble and will likely bind (see Section 5 on page 6 of this document). Several friends on different teams have built over-constrained shafts and they have all regretted it.

A second comment is that the configuration you show in your photo, where the axles are supported at both ends may be preferable when there are a lot of obstacles that one must drive over like in the 2016 game. Anecdotally, it seemed that all the robots with broken wheel bearings at the events I attended in 2016 had cantilevered axles whereas none of the ones with the axles supported at both ends experienced bearing failure.

1 Like

I’ve thought about doing this, but if you’re using 15mm belts and 2x1 tubing as the rails, doesn’t the belt prevent you from attaching anything to the top/bottom of the tubing in that half of the drivetrain?

1 Like

By “inside the framing members,” I mean inside the frame perimeter, not inside the tube itself.

1 Like

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