I’m having a hard time understanding why the rocker is needed as opposed to 1 wheel on the ground.
The only thing I can think of is to create clearance under the robot when not driving sideways?
It ensures a strafe wheel is always touching the ground. Another common way to accomplish this is by having one wheel on a spring loaded suspension so that it’s always pushed into the ground.
1 Like
Yes this makes sense, thankyou!
The 3 most important parts of your robot… the drivetrain, the drivetrain, and the drivetrain.
Consider what any drivetrain beyond the kitbot costs in terms of resources vs. the benefit to gameplay. The majority of teams would be better served focusing on game piece manipulation.
5 Likes
To expand on pkrishna’s point about keeping a strafe wheel on the ground - neither your robot frame nor the carpet or floor are ever actually flat, and assuming they are will lead to unpleasant surprises later.
Can’t be pushed by what can’t hit you
If looking for overviews, two of the presentations here are about various drive trains:
See, I don’t classify swerve strafing as “Going sideways”. Mecanum and H-drive, when they move in a direction that isn’t forwards or backwards, drive slower and sacrifice maneuverability. Swerve is a rare exception, where you’re basically just turning the “Forward” direction of travel to another side of the robot, not going sideways.
However, the inability to go over obstacles easily, motor usage and weight of swerve is a big enough turn-off for me to stick with WCD.
Can’t score under defense if you can’t push. Remember that this game is fundamentally broken, in that a dead robot sitting next to your rocket shuts down half of the scoring opportunities for said rocket if you can’t push them out of the way.
The biggest problem with H-drives is ensuring both your forward wheels and your middle wheels have enough traction to drive. The rocker does this with no additional actuator.
That’s a strange definition, and while you’re welcome to it, most people consider swerve moving sideways.
1 Like
We ran a four-CIM octocanum for several years that was heavy, but no more of a motor hog than any other drive train. (This was back before they let you have six CIMs on your robot!)
It was grotesquely inefficient, but relatively easy to build, program, and drive–and only used four motors and one solenoid. (If I had to do it all over again, I’d use four pancake cylinders, one on each corner, but still drive them off the same solenoid.)
My students loved it – we were fast and maneuverable when we wanted to be, could beast through just about anyone in traction mode, and were very difficult to push in any direction if we didn’t want to get pushed. (We couldn’t turn in traction mode, but we didn’t need to – you use the low-gear traction wheels to push enough so that when you switch to mecanum you can zip away.)
One of the things I liked about it was that your gear-shifting was built into the mode-shifting. By changing from 6" wheels to 4" you get a .67 factor reduction, which you compound one step further by going from a small sprocket on the 6" wheel to a large sprocket on the 4" wheel. IIRC we were geared for about 16 fps on the mecanum (9:1 Banebots p80s) and 6 on traction, without the need for an additional shifting gearbox of any sort.
Here is a video series detailing how we build our drivetrain. We’ve been slowly evolving it since the summer of 2014 and have basically used the same design since.
3 Likes
All of 2363’s recent videos are great resources. I know they’ve been widely watched among the students on my team.
1 Like
My team has done west coast with pneumatic wheels for a couple of years, and they’re great for several reasons. One they are easy to build, two the pneumatic wheels provide excellent traction, this year we were able to push just about anyone around. Third they are very easy programming wise, the code is very simple compared to a swerve or even Omni.
This drivetrain video and your gearbox has been a tremendous benefit for our team. We made slight changes to your gearbox design, and are using HDPE plastic for the plates and it works incredibly well. We didn’t have a drivetrain issue all season. Your explanation on the chain-in-tube was spot on. Our kids modified it slightly and we ran 1x2 tube with #25 chain on 18 tooth sprockets with no problems all. Thanks to 2363 once again for an incredible resource.
1 Like
Thank you guys very much, our team has decided to go for a VexPro 6WD Custom West Coast Chassis which, if funded, will be built in the fall with NEOs
Ok, now that OP has made a decision, I’d like to help teams at the lower end trying to make the decision to step up, or not. If you used anything other than a KoP or Versachassis or WCP drivetrain with mods only to the wheel size and gear ratio:
- What did you build?
- What build resources did it take beyond a drill press and a chop/horizontal band/cold saw?
- How many hours did it take over and above KoP/VC/WCP modifying only wheel size and gear ratio?
- How much of an improvement did you find this to be over a KoP/VC/WCD modifying only wheel size and gear ratio? Please be as specific and/or descriptive as possible.
If you have different answers for different years (or different for official season vs post-season), please pick one or separate them clearly.
2 Likes
OK, to answer my own questions for my time mentoring 3946, 2013-2018.
- 2013: Let me just put the takeaway that every wheel on an FRC robot should be driven.
- 2014: Built a mecanum drive train. We had already decided against the KoP (for reasons which weren’t really valid). It only took a few dozen extra hours, evenly split between mechanical and programming. It did not pay off at all, because we had analyzed the game as being far less defensive than it wound up being.
- 2015: Built an H-drive. Wound up dropping the slide wheel and just doing a 4-wheel omni by the end of the season, but built a suitable pneumatically-engaged strafe wheel a year later.
- 2016 STRONGHOLD: Built a 10 wheel drive, with overlapping wheels. The drive itself worked quite well at crossing defenses, but we tried to use KoP holes for chain runs, which caused a lot of chains going off-track. Bottom line: don’t change infrastructure from inches to millimeters (or vice versa) and expect kit/prefab parts to keep up.
- 2017 STEAMworks: used the KoP, then chopped half the chassis to a 4 wheel robot. Awesome; would do again (and have done again) when short wide robots meet the need.
- 2018 POWER UP: Stock drive train. Unfortunately, the manipulators didn’t work. Wound up dropping back to a 2017 (4 wheel short and wide) drive drain for post season, which worked MUCH MUCH better.
2 Likes
Our team has tried almost all of the main groups of drive trains over the years. Unfortunately, because we are a low resource team, we have never been able to make the mighty swerve drive.
• 2015 - Four inch mecanum. This was before my time, but it is the only time our team has ever advanced past the quarterfinals in an event, so it probably worked pretty well.
• 2016 - Tank drive. And when I say tank drive, I mean the REAL tank drive: 3.5 inch wide 3-4 foot long homemade tank treads on each side in the typical tank tread robot configuration. These things were awful to put together, as each tread was two sides of 25 chain connected by cut and filed 1x3.5 inch plexiglass slats and so, so many small bolts. While they were very much not worth it to make, the treads on that robot allowed us to brute force our way over every defense making us one of the few bots that could do everything. We still have that robot (in some state of disrepair), and hope to put a tennis ball cannon on it to make it a true tank.
• 2017 - Six inch mecanum. This worked pretty well, we made the frame out of a bent a roadsign and AndyMark toughboxes (like every other robot so far). It gave us very high manuverability, but whenever another robot got in our way, ours was useless at pushing past it.
• 2018 - Stock chassis. Relatively fast assembly, though the center drop gave us tall robot syndrom, which was very not good.
• 2019 - H Drive. We used vex gearboxes for the first time instead of the AndyMark toughboxes, and it was so much better. Also with this, we put a simple suspension system on the center wheel that gave us high and constant traction for strafing. This led to our best season ever (ranking 9th at our home event and 12th at the infamous Kentwood event). If we had finished our climbing mechanism (curse you, bag day!), We could have done much better due to the manuverability of our drive train.
Overall, would highly recommend H drive.
This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.