A week later than we’d hoped, we finished the drivetrain for our 2007 robot. We have video of it driving that we’ll likely put up on YouTube or similar soon.
It’s a flat 6WD with custom, single-speed gearboxes geared at 17.5:1. Its speed sits somewhere around 6 ft/s – slower than what we’ve run for the past few seasons, but purposefully so. The wheels are 6" diameter Kitbot wheels on the center with AndyMark omniwheels at all four corners. Traction is great and it turns on a dime.
Overall weight as shown – minus the battery and the plate it sits upon – is 41 lbs. and change. The electrical board is 7 lbs., each gearbox is also 7 lbs., and the wheels, chain, axles, sprockets and frame make up the difference. The front and back rails haven’t been lightened yet as we’re awaiting a finalized design for other components and want to ensure we can securely mount them to the frame. Lightening those parts will reduce the weight another 2 or 3 lbs.
nice setup. the organization on the electronics looks superb. why didn’t y’all go with the mecanum drive this year? it seemed like the full range of motion would come in handy for navigating around the rack… i guess this means that you guys have an awesome mechanism under construction
Fred (fsayre here on CD) will be thrilled to see that people think the electrical board is neat. He’s made that a priority this season and done a pretty good job of it.
We opted not to go with mecanum wheels for a few different reasons. Primarily, our prototype never was tested with anything beyond open loop control and so we were uncomfortable committing to a mecanum drive without the utmost assurance that it would work flawlessly. Secondarily, it doesn’t jive with out strategy well, nor do I think that omnidirectional movement is the answer to the challenges of placing ringers on moving spider leg.
Thanks for the compliment. Currently, there are no plans to put anything outboard of the wheels for protection, but if we have some weight available to us, I’d consider it.
I’m sure you already know this, but for the benefit of others like me who are admiring your work, I’ll point out (anyway) that you have 15 lb. of standard <R37> bumpers available if you choose to use them. A little more traction and protection are often helpful.
Also, I really like this traction set-up. My team is planning a conceptually similar layout using four aluminum AndyMark omniwheels and two slightly modified AndyMark FIRST (2007 KoP) wheels driven by four CIMS. Your set-up will have all that ours has, and a bag of chips. Very nice work.
How well do the side panels take the weight of the gearboxes and motors? With all the weight reduction you’ve done on the side channels, it seems as though you’re reduced their torsional stability as well. We’re mouting the 2x CIM adapter into the Banebot tranny, and we’re mounting that in a similar fashion. The Banebots, with the two motors, weigh 9 lbs a piece. I have a feeling that we’ll need to support the back ends of the motors. Have you come across a similar issue?
Overall, I LOVE this setup. It’s very similar to what you used last year, but the robustness is so much greater. One additional feature I would add is a row of 1/4" holes (spaced 1" apart) along the tops of all the rails, similar to the kit chassis. This allows you to mount whatever superstructure you want, where ever you want. We’ve done this with our sheetmetal chassis this year, nothing could be easier!
It also makes lots of ugly (imho) hoels in your chassis. I’d rather mount the superstructure I want as opposed to any superstructure I want. If you design it all out first, you don’t need any uncecessary holes, and in my opinion that is elegant.
Very neat, organized work. You may want to consider lessening the front and back bumpers (didnt know exactly what to call them) because they are perfect candidates to get hooked and cause a nasty ordeal.
Thank you, also, for the compliment. We’re all really proud of this drive.
Unfortunately, it looks like the other mechanisms that sit atop the frame won’t allow for us to take much advantage of bumpers this season. While the bumpers aren’t included in the 28 x 38 sizing requirements, the extra distance required to deploy things beyond them is and we’re not sure if it’s worth the sacrifice.
We’ll be mounting some bracing across the top of the frame to accept the rest of the robot, so that should go a long way toward returning torsional strength to the frame tubing – but so far, we haven’t seen anything that makes us too worried about how well this will perform or how long it’ll last.
Great job 488. I knew when I saw the pictures of the parts made that it was going to be an insane drive base. I think I saw an annotated drawing of that back in November. What’s the footprint?
Now, I am intrigued about why you chose such a low designed top speed. My normal assumption when teams have slow design speeds is that they wanted more pushing power, but I know better than that with 488. Especially considering you’re still using a drive-train designed for precision turning, and are in the 5’, 110 lb. weight division.
It’s not a concern primarily because a defender ought not be wasting their time with us and an offensive machine’s time is better spent trying to score (or outscore, as it were) our robot than to get in its way. We can switch between one or two pairs of omniwheels in about a minute, so we have some flexibility in determining how we want the robot to perform.
Are you speaking of the fenders at each of the four corners? If need be, I can mill those to be a vertically oriented ‘wedge’ that will help to stop other machines from getting caught on or in the channel.
Thanks! I posted an annotated rendering of the mecanum drive base that we built in the off-season, but I don’t think I’ve posted any other images of this particular drivetrain. In any case, the overall dimensions are a slim 27" x 36" – while the wheelbase is approximately 24" wide and 28" long.
Nope – it’s not a legal COTS component, and it’s heavy and ugly. It just happened to be a convenient platform for support the battery while testing.
As I mentioned above – we can easily switch a pair of omniwheels for something that has more traction – so this is a very versatile design. Defense will be important in this game given how imbalanced the effort to score is as compared to what it’ll likely take to prevent someone from scoring, so I geared for a slight pushing advantage over most teams. We’re making no effort at pushing those with drivetrains designed expressly as tanks because we don’t see those types of robots spending much time hassling us.
We’re planning some ability to score ringers in a limited number of locations, but we don’t imagine that we’ll ever be considered so much a scoring threat as to warrant defense through a match. We are expecting some attention, however, and we’re hoping that we’ve reached a happy balance that’ll allow us to push through some defense and maneuver around others.
This robot is slower than those we’ve built for the last two years, so for us, we consider this to be slow and deliberate. It’ll be interesting to see if a majority other teams end up faster or slower than our chassis – especially considering how this season’s kitbot is remarkably less well-suited for defense than in years past.
