I think we were using 2.5" Colsons last year. The magic wheels are 3". Printing the hubs allows us to get a larger diameter wheel squeezed in without actually raising the height of the robot compared to machining the hubs like we did last year.
Admittedly, a bigger wheel diameter would minimize some of these inconsistencies. At WMRI we drove on magic wheels and we tuned our paths once and didn’t have to touch them again. During the season, I was constantly making adjustments between matches.
Many teams have a “default” drivetrain style that they use most years, while staying open to changes depending on the game and field. For my current team, I’d say it’s a 6 wheel West Coast drivetrain with 4" grip wheels and a drop-center. This year we deviated from that in some ways, and hopefully I can give some insight into our design process:
-Last year we had mild trouble with the endgame platform, and got high-centered a couple times. This year had a similar endgame platform, so we decided to use 6" wheels instead of 4". We figured 6" wheels would help a little if we tried to climb to Level 2 or 3 as well - to drive onto a platform you have to get the center of your wheels above the surface of the platform, and starting with the center of our wheels an inch higher is one fewer inch we need to elevate the bot. The tradeoff is moving our center of mass ~1" up, which we decided we were okay with.
-We wanted more maneuverability than we had last year, and considered doing a drivetrain with side-movement (H-drive, swerve drive, etc). But the tradeoff with those would be more time needed to program them (and thus less driver practice time), and they would take up more space on the base of the robot where we wanted to put electronics and other mechanisms. As a compromise, we decided to put omnis on the front and back, and have only the middle two wheels be grip wheels. This also made the drop-center unnecessary, as the omnis allow us to turn well without it. Keeping grip-wheels in the center made us a little more resistant to being pushed around than we would have been with all omnis.
-We considered using pneumatic wheels instead of grip wheels to help us survive driving off Hab 2 at the start of the match, but ended up not doing it. Towards the beginning we were thinking we’d have a good autonomous mode and starting on Level 2 would mess it up because we wouldn’t know where we landed/what direction we were facing. We didn’t consider starting on Level 2 until we got to our first regional and didn’t have working auton, and at that point we didn’t want to make big, untested changes to our drivetrain
-We ended up playing defense for a good chunk of our first regional and turned out to be really good at it. Between the two regionals we debated swapping our omnis for grip wheels to give us more traction for pushing opponents. But because we hadn’t designed a drop-center, we wouldn’t be able to turn well at all with all grip wheels, and we decided the dexterity of the omnis would serve us better than all grip with no drop-center
These decisions all started from the premise “if you want to deviate from [default drivetrain], you need to have a good reason” and I think that’s a good way to approach it if (like us) you want to get the drivetrain designed and built quickly so you can move on to more interesting parts of the build. If your team finds drivetrains really interesting and wants to invest more time in developing an optimal one or a cool one, maybe a more “blue sky” approach would serve you better.
Just like the very old quote attributed to Abraham Lincoln concerning how long a man’s legs should be – a robot’s wheel diameter should be sufficient to reach the ground.
2056 got to Einstein last year with some wheels on their drivetrain that didn’t touch the ground (under normal driving conditions)
I guess a person with extra legs can afford to let some of them hang idle, some of the time…
Is it ok that we upgrade to this Kit and add one omni wheel in the center “H” position? Thus this will allows the center wheel to move the robot side to side.
Sure, but you need to make sure that your H wheel is spring-loaded into the ground so that it always maintains traction. Sometimes the floor isn’t perfectly flat so you can’t try and set the wheels at the perfect height.
My team did an H drive in 2017 and we didn’t realize that. We ended up switching over to a 6WD tank drive because we could never get the H wheels to reliably transfer torque into the ground.
Then did you do the side moving with this 6WD tank drive?
No, we were not able to move horizontally when we switched to a 6wheel tank drive.
The point is, make sure that you either have 1 spring-loaded wheel or two wheels on a rocker to make sure that you are always forcing a wheel in to the ground. That is crucial to make an H drive that actually works.
I see. Thanks!
I asked about that, apparently done intentionally to help avoid beaching and make the climb less likely to rock backwards, same as the slightly forward leaning elevator. Details details details…
I don’t actually think this will work, the kit only changes out the corner wheels, which will improve turning, but the center wheels are still traction, so you won’t be able to do an H-drive.
As mentioned, if you did not assemble the drivetrain properly, that could be why it came apart and quit working. It can be tricky to get all the pieces in the way they should be, we usually have trouble with it, especially with students who haven’t ever done it before. If you’re a rookie team, then you’ve never done it before, and might have missed something. It’s easy to do.
You might want to spend more time trying to get what you have, working as designed, and practice driving with it a bunch, before deciding you need to upgrade.
We’ve found that we build a more effective robot when we spend more time on trying to improve the parts of the robot that move the game pieces. The kit drive train has worked well for us the past several years. (after we fix the mistakes we made putting it together, that is)
Yeah, we are rookie team. The kit drive train has worked well for us as well.
Now for us it’s off season, we think it might improve our robot if we upgrade the wheels.
I think the gentle hint here is to live with the existing drivetrain (when fixed) and focus on either trying to build a more advanced object-handling mechanism, or perhaps refine your existing mechanisms. I recommend the latter, to see if it is possible to achieve really good results with the original concept. If you push yourselves with a specific target in mind during the off season, you will learn more design, programming and sensor usage. These will serve you well in coming seasons, allowing you to perfect what you have originally built.
There are very, very few elite level teams with all of the resources required to fundamentally redo the entire robot mid-season. The rest of us just have to constantly refine what we have already built.
For instance, if you struggle scoring any hatches, try to either make or improve your hatch mechanism with specific times in mind. If top teams are acquiring hatches in ~3 seconds from the driver station, aim for 9 seconds and see if you can push your times even lower. Likewise with hatch placement, stick with low scoring locations and focus on speed and accuracy of hatch placement.
Improving these metrics will force learning and improvement of skills and knowledge that you’ll be able to apply to next year’s robot.
To see the results of this kind of activity, review matches of 1241 this season. Bear in mind 1241 is a past World Championship team and still top 5 in the Ontario district. Their initial Deep Space robot design is mostly unchanged, but they relentlessly improved it throughout the season from pretty mediocre early on to a point where they stand a fair chance of being an alliance captain at worlds.
That’s the kind of improvement you should be aiming to accomplish in-season with more experience.
Hi, Guys! Thank you!
Based on your suggestion, after discussion, we decide to change the front two wheels from HiGrip to Omni.
Hope this will improve the performance of our robot!
3946 modified a U2 kit chassis to H drive in 2015.
As noted above, you must not have a solid wheel contacting the floor. We solved the problem then by removing the rubber tread from the center wheels. If doing this with the 6" setup, you could just swap the center wheels with 4" wheels and allow them to “float” most of an inch off the carpet, though they would give you some “fun” as you climbed onto the HAB if you’re planning to use this for an offseason robot.
Also as noted, the transverse-mounted wheel must be on a spring or pneumatic mount, or other system which ensures that a steady amount of weight is supported by the center wheel. We tried a variety of altitudes, and nothing worked, even on level carpet. We found that the amount of weight on the wheel varied as a function of acceleration, making control of the robot significantly worse than a simple skid steer. As such, we removed the slide wheel after several matches. Post-season, we did mount an omni slide wheel on a pneumatic pressor, and it worked great once we got the pressure adjusted to the right value (about 25-30% of the weight on the center wheel for us). Note that the post-season slide was on a different robot chassis.
This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.