Our design this year uses Neo motors, 1:1 without a gearbox, for both our indexer and shooter mechanism. We were having some trouble designing a way to mount the motors, and may have come up with an interesting design. Can this be done and/or is this the way that this sort of thing should be done?
Right now we have the motor directly screwed to our shooter side frame (maxtube). The outside part of the maxtube is drilled at .75" to locate the small lip of the motor, keeping it centered. We then have a small .5" 8mm keyed shaft to hex adapter keyed on the motor shaft, which is then connected with a hex coupler to the roller shaft.
I’ve seen many teams use a belt or chain in this sort of situation, so that you can have the motor not stick out ( the motor can then be mounted on the inside, next to roller) but this design seems simpler. I honestly can’t see these motors ever getting hit up on our shooter, and even if they do, they are going to be rigidly screwed on.
Can anyone see any potential pitfalls/reasons this sort of mounting can’t be done?
It looks like you’re directly driving the shaft with the motor. I wouldn’t suggest that because the vibrations and shocks that the shooter wheels and shaft experience are going to be directly transferred to the motor shaft and bearings. For example, when shooting, I assume there’s compression on the notes, so the shafts are going to be pushed outwards a bit. As a result, that could then push on the sides of the shaft and motor, leading to faster wear on the bearings. Of course, I’m a programmer and not that experienced in mechanical design, but that’s just from my experiences while prototyping and testing stuff.
You may be a programmer, but you’ve picked up on why mechanicals hate direct-driving stuff with a motor.
Can the mounting be done, yes.
Is it a good idea, NO.
Shock loads, as noted, are going to be an issue. Shafts getting nudged around, issue. I’m not seeing any screw attachments of the NEO, that’s a pretty major issue–though “they’ll be rigidly screwed on” could give either issues or no issued depending on what gets bent.
And you think the motors won’t get hit–oh, they will, trust me. Nowhere on a robot is safe.
You can get away with this for a short time. I would be highly surprised if you didn’t need to change it after your first event.
What I would do instead is to pick any 1:1 gearbox–actually, for this, I’d probably suggest the MAX 90 for packaging reasons, or a MaxPlanetary base unit–and use that. Or possibly a 1:1 belt/chain solution. The goal is to get something into the motor’s power path to protect the motor from mechanical issues coming back and messing it up.
Oh, and one other issue. NEOs at 1:1 are going to be insanely fast, even for a shooter. Please check once you build it to make sure it doesn’t make Note Powder.
You sure? Shooters which can shoot from a long range this year (which this bot appears to be able to do) require really high RPM to get a flat enough arc to increase the margin of error.
A NEO runs 5676 RPM at free speed. Maybe not as fast as a 775Pro at top speed… but let’s see just how fast you can expect stuff to come out of the shooter shown.
Looks like 4" wheels, 1:1 drive from the motor so the wheel is spinning at 5500 RPM (easier calculation plus allowance for load). C=pi()*D. Wheels spin at 69,000 in/min linear speed, translates to 96 ft/second exit velocity for the Note if it leaves at the same speed the wheels are spinning.
That’s 1 second of flight time to cross two FRC fields end-to-end, almost. This is generally considered pretty stinkin’ fast–if this were something heavier, it would probably be considered for “are we sure this is safe?” treatment by officials.
Shots aren’t gonna be exiting anywhere near the free speed of the motor because of the motors don’t output any power near their free speed and therefore will quickly bog down. Combine that with the headroom required to spin up the shooter in a reasonable amount of the time, and accounting for things having a bad day because this is FRC, and now you are shooting at much slower speeds. Lets say its 30% headroom and now you are shooting at 60ft/s which is only a bit faster than you’d want to be shooting for max distance shots.
Ahh, I guess this is why I haven’t seen a robot with a direct-driven shaft before. I had not really thought about the loads that would then be transferred directly onto the motor.
What we’re probably going to do now is mount the motor on the inside to a 1:1 gearbox, then chain forward to the rollers in a more traditional way.
I agree. We’re definitely going to rigorously test our final shooter, see what RPM works best, and probably even add a 1:3 module into the gearbox to give us a little more torque, as we won’t need all the speed.
Thanks for all the feedback! We probably would’ve have caught this issue otherwise until it was already too late (mid comp, in between matches when we’re desperately trying to swap out a motor with shot bearings and whatever else might’ve happened to it )
I think you misinterpreted what I said. Because of the inherit motor limitations (specifically low to zero output torque at high speed) as well as shots wasting a lot of energy due to it acting like a collision and because of friction, shots are physically unable to exit anywhere near your shooter’s free speed. Gearing the motor 3:1 down just runs into the exact problem but at lower shooter rpms and you will will run into the exact same issues but now you are doing so at lower velocity. For long range shots, you will need most of the speed of a 1:1 shooter because of these limitations
Chains don’t particularly like spinning at high RPMs, belts would be a much better option here
3 times zero is still zero, motors output progressively less torque as RPM increases and zero torque at their free speeds. You want the shot/recovery to occur well below the free speed so the motor actually outputs anything, with free speed shots being impossible
Flywheels with extra mass take several seconds to reach their free speed which is way too slow, and ilding the shooter at a lower rpm isn’t gonna help you as most of the spinup time would be spend near but not at the free speed. And even if you somehow solve that, you still lose a bunch of exit velocity from friction and collisions and slip
1293 ran basically this same setup in 2020 on their dumper, with fairly high compression to ensure the power cells didn’t eject before we wanted them to.
We only ran one event with that configuration before the world ended, but the motors were fine after Palmetto (where we seeded 8/63 and were a first-round pick) and got reused into our 2021 rebuild which had more run time since we did dozens of takes of our shooting challenges, plus the two-day SCRAP tournament.
Support the other end, mind your wires, and keep your roller sizes short when possible (we were running 1.25" OD VersaRoller).
For OP’s post, I think the 2" intermediate wheels are survivable–I’m more concerned about getting enough structure stiffening it up from side to side than anything inherent to the NEO. The 4" wheels are a lot more than the Kitbot is slinging, so I’d probably focus my design efforts there on adding a gear/chain/belt stage. (Bonus for chain: you get a free flywheel!).
Edit: Here’s an extra photo or two from the camera roll, showing what we had going on.
(We made a tactical error; we were using the 5/16" tapped plugs for these rollers, but our rails got made with 1.125" OD bearing bores not 7/8" ones. As I recall, we used some bushings and just ran 0.5" ID round bearings.)
I know it is not considered ideal, but we mounted our launcher wheels very similar to the way you have shown in 2020 and 2022 without any failures after multiple events. Our method was actually a little simpler. We drilled a 8mm hole in the end of the shaft, cut about a 1" slit then used a shaft collar to clamp the motor shaft. I know it is not the most ideal setup, but I can’t say I would not do it again,.
When I worked at AndyMark, that was known as “the Churro trick”. Actually saw action on the 2017 airship rotors, which was fun. And I ran it on a couple of my old Fight Night robots.
Keeping that concentric on a high-speed mechanism sounds like a nightmare, though.
There will be plenty of teams running 1:1 off a neo this year with 2.5 or maybe 3" wheels.
Most 4" wheels in frc have a fair bit of inertia. And while the 96fps isn’t going to likely happen it does bring up some worries if you get to around half that. Maybe not insane worries (but concerns for individuals not paying attention or at close range, equipment on the scoring table, etc) 60fps is about 40 miles a hour, not too bad and plenty of time to react unless you are at point blank (warning pit crew!)
More to the point:
Concerns around motor shaft axial loads/vibrations and concentricity are very well founded and detailed in this thread. Lots of FRC knowledge floating around. It is fun to stir the pot an see what you can get away with sometimes. I don’t think this is something you can get away with on a comp bot in the usecase mentioned. Prototyping? Sure. Comp? Probably not for too long.
Regarding the shooter wheel velocity vs. note exit speed, we did some limit testing with a pair of Cims and 4 inch wheels. We found basically no difference in note exit velocity between a roughly 1:1 drive ratio (~5000 rpm, 85 fps wheel speed) and a maxed out 2.5:1 speed ratio (estimated 8000-10,000 rpm after the motors wound up for 10 seconds, 140 fps wheel speed). When calculating the velocity via ballistics equations, the note maxed out at 40-50 fps in that test shooter.
What I believe this shows is that the note is initially engaging the spinning wheels with kinetic / sliding friction, and if the wheels are spinning fast enough the note will not have time to “catch up” to the wheel surface and grip with static friction before it is ejected. The faster wheel speed only results in more slippage between the note and the wheel, and more orange flakes as a result.
This still leaves room for secondary improvements to increase friction with the note, such as wheel selection and note compression, as well as mulitple wheel stages to progressively increase the note speed. How much speed is needed is a sepatate question, of course.
The manufacturer of the Kraken motor has a competition concept that uses direct to motor driven shooter wheels without any kind of extra shaft support.
How well this holds up over the long term is a question I cannot answer.
My own team did it in 2020 with a Falcon motor on our competition robot.
)
