I really like the design - especially the cutouts in the metal - but I see two issues.
Firstly, this setup would use 12 PDP Slots. :ahh: If you’re fine running your robot with only 4 slots left then props to you - I surely wouldn’t be.
Second, the encoder positioning looks a little worrisome. I’m sure the modules will most likely be protected by the bumpers and chassis, but I’d still be worried about something banging up the encoder, especially if you fully rely on the encoder for your software to work correctly.
But props to you for designing this. It takes quite the skill.
Looks familiar (but more sturdy)…
I agree with JJ that the pros of 2 775’s per module may not be worth it, unless you only use 3 pods per robot. (See 5817 2017) The encoder and pinions/gears seem a little exposed for my liking , but it shouldn’t be hard to make a cover for both either with 3d printing or polycarb if you hit problems down the road.
The design looks really cool, but if you can find a way to make this dead axle the pods will get stronger around that point where you have hex shaft right now. Maybe using a versawheel and bearing bore sprockets would do the trick, but it seems pretty strong already so that may not be necessary.
While good code can avoid motor wires getting messed up by limiting rotation etc, if you do actually go and make this be sure to use some harnessing so your encoder wires can be bunched togegher with the motor wires.
All in all, great design!
Indeed. Looks very similar.
1 Like
For pretty much every game in the past 5 years (since I’ve been doing FRC), you can have a decent robot with a solid drive train plus 4 motors and pneumatics. It might not be able to do everything, but with a good driver I’m convinced it could perform at a mid-to-high level.
2012: 1 for intake (pneumatics to deploy), 2 for indexer, 1 for flywheel, pneumatic bridge tipper
2013: passive human feeder intake, 1 for indexer, 2 for flywheel, 1 for climber
2014 (easy): 1 for intake (pneumatics to deploy), pneumatic catapult
2015 (easy, but stupid): passive human feeder intake, pneumatic tote elevator, 1 for step pusher (3 for ballast)
2016: 1 for intake (pneumatics to deploy), 1 for “utility arm”, 1 for flywheel, 1 for climber
2017: 2 for floor intake (1 for roller, 1 to deploy), 1 for climber (1 for ballast)
Thanks JJ
Yes I agree 12 PDP slots is a lot, but keep in mind that I am not designing this to do anything other than drive. My team does not have the budget to build something like this nor the time, so I decided to CAD this for fun, and so I can learn more during the off season.
If I was building this for a regular season, I would look at changing the motors to something like cims or mini cims, so I don’t take up as many PDP slots.
Now regarding the encoder, yes I agree, I would like to protect the encoder a little bit more. I am looking into making the encoder sit on the inside of the pod, and make it super compact. Let me know if you have any other ideas for the mounting of the encoder.
Paulius Pace
Team 3161
Limiting yourself to fewer remaining robot for simplicity sake- probably a good idea. Using less cause you physically can’t use more motors is a tough design constraint to work around.
You could also easily go over 4 motors for most of the games you listed with a fairly simple robot.
2013: 1 intake roller, 1/2 intake arm, two on a shooter (either to adjust angle, more on the shooter wheel, or as a kicker), plus if you wanted to climb above level one
2014: pretty easy there
2015: 2 for intake, 2 for elevator, 2 for can grabber
2016: 2 for flywheel/catapult, 1/2 for intake arm, 1 for intake roller, 2 for climber, 1 for defense manipulator, plus another to feed the ball depending on rest of design
2017: 1 for gear intake rollers, 1 for intake arm, 2 for climber I guess makes four, but not if you want to do any shooting.
Some of these might be a bit of a stretch, but the point is, while 4 motors is feasble its not a longshot for a design that works best to need more motors. I’d hate to pick a prototype that doesnt work as well because I am limited to 4 motors. This isn’t FLL where you can easily make crazy attatchments and stuff, and 4 motors is a big limit to lots of potential designs.
This changes things… I’ll shut up now
Oh I definitely agree. This would be a design choice that I think could only be done well by a select few teams. I’m definitely not saying that it’s hard to use more than 4 motors (in fact it’s pretty easy). I was simply saying that if you have a super strong drive base, it is possible to have a decent robot in the past 5 years with only 4 motors for manipulators.
Interesting design choices you have here. For the style of swerve module you’ve made, this is a really good design.
That being said, I do wonder a few things:
- How would this module be supported? If it’s just by the hex shaft up top, you may need to rethink the way you’re doing bearings.
- Is that 20fps free speed or adjusted? Either way, going to 18fps or so will serve you well, even with 8 775s.
- I like the slotted sideplate design for the plate up top, but how would those slots be made? Most manufacturing processes for these things have a radius of cut. Also, consider using a “jesus nut” for securing it, rather than relying on the tolerances of the standoffs.
Overall, a pretty modern take on a classic design. I personally prefer coaxial setups, but I could see myself using this if I had to in a number of ways.
Thank you asid!
-
yes i was planning on using a hex shaft that goes through the chassis and would have a belt pulley system to rotate the pod.
-
it is 20.22 ft/s adjusted (a 12 tooth press fitted onto the motor, meshed with a 100 tooth, which has a 16 tooth sprocket driving a 22 tooth sprocket.)
-
This was designed to be used on a waterjet. I have personally used slotted plates before and its super easy to implement and super robust at the same time. Its nice alternative to sheet metal bending, and allows you to build some pretty neat things out of 1/4" and 1/8" sheet metal. My team has been able to use this method with wood as well. We are looking into building a lot more of our robot out of wood next year because our school has a nice wood shop but no machine shop.
Paulius Pace
Team 3161
Had this idea recently (possibly in the shower)
I was going to use gears for the final stage but the sprocket makes the spacing much more manageable…
Good work! Have you thought about doing 2, 2" wheels? Could be even tinier and allow for a less aggressive reduction.
Thanks AJ
I used sprockets because I do not like the idea of having the final stage being gears cause they are more likely to sheer. At first I was going to use belts and pulleys, but there were no belts that would contain the pod to the size I wanted.
I have not thought about 2" wheels. Vex pro has 2.5" wheels that I recently looked at. If I was doing this again, I would consider it.
Paulius Pace
Team 3161
Would changing the ft/s to 18 really make a significant difference?
Also what do you mean by “jesus nut”?
Paulius Pace
Team 3161
Since the most important part of a robot is your drive train, and if you have the motor room for it, I would not mind using something like this, if we have the budget and capabilities to pull it off. That being said, I cannot see my team going with anything like this, and I would not recommend it to anyone unless you are fully confident in your teams capabilities to design and build a chassis.
Paulius Pace
Team 3161
If the 1/2" hex shaft is the only thing holding the swerve to the robot, it will probably bend/break. I would use a bushing similar to our swerve with all the shaft stuff removed. Light, easy to make and no breaky, breaky. This also gives you some place to pass the pile of wires needed to get down to the motors and encoder. I assume you have an idea how to manage wire wind up.
18fps is only a 10% decrease, but considering that you’re not shifting that’s 10% more headroom before you brown out your drivetrain. Going even slower can help too, but at that point you’re starting to waste the power of 8 775s.
20fps can definitely work, but your software will need to be that much better than if you geared lower. I know 2451 had good success with 8 775s most of the year, so you could look to them for ideas on gearing.
Jesus nuts are described in this Instructable in step 4: http://www.instructables.com/id/How-to-Build-your-Everything-Really-Really-Fast/
It’s essentially a captive nut that alleviates some of the problems associated with cutting radii (even in waterjetting) and attaching things at right angles.
+1 to Mark’s point about using a large bushing; this is similar to what the Team221 Revolution Pro swerve does. Alternatively 2 large bearings like the 6812 is not too expensive and offers a lower-friction solution.
As far as 775s in a drive train, a current limiting scheme needs to be implemented. With 8 775s you’ll have issues blowing the main breaker if you don’t limit their current. My understanding, 2451 did it through the roborio and were very successful. We did it through the Talon which is how I would recommend doing it. Set it and forget it… we were also successful.
Could you expand more on why drive trains brown out? My team has had issues with browning out before, but we fixed that by adding limits to our speed. I’m wondering what causes a drive train to brown out, and how can you prevent it.
Ahhh those, I call them t nut slots. I really enjoy using those in addition to finger locking slots. We used them in 2015 and 2016. I was thinking of adding it to the pod, but I thought that the Hex standoffs were enough, and adding them would be overkill.
I will look into adding a bushing. I thought a hex hub would have been enough as long as it goes through tubing and has a bearing on either side.
Can you expand on a current limiting scheme? How did you guys use talons to do so?
Paulius Pace
Team 3161