pic: Swerve! (Module)

[NoodleKnight]

It's not so much tension that causes chain de-rail/breaking, its alignment. #25 is more finicky because the chain is a lot smaller, correctly tensioning the chain helps some but if your alignment between sprockets is out of whack, the chain will hop off. #35 is a lot more forgiving, even if you had a noticable amout of slack in the chain, as long as your alignment is held true (and since it's housed in a box frame) it's not going to cause any real problems, unless you really hate backlash.

+1 on the keyway, pins and setscrews will only fail at this stage of the reduction, if you're having a hard time getting the machining done I can connect you to a few people =P.

As for that miter gear, I'd suggest just machining sleeves that sit on the shaft to prevent the gear from moving in any direction, if you make the gear adjustable in position then under the right circumstances it'll adjust itself (I'd actually just machine sleeves for the entire shaft, then you'll never have to worry about alignment during testing/competition). Then again, if you weld it in that works too, make sure you keep the keyway in there.

[sanddrag]

I think your programmers would be more happy if those were gears in there (less backlash).

[NoodleKnight]

Isn't simplicity one of the key elements of this design?
It doesn't look like you can directly swap the chain/sprocket element with two gears, so you're going to at least need 3, which means you'll need to add in another shaft and its bearings to support it, which means more machining, on top of the gear/gear stock you have to buy and machine as well.

Anyways, shouldn't the encoders be measuring data off the wheel/the shaft that it's connected to? If so I don't see backlash being a problem, unless you're doing dead reckoning, which I'm sure our programming mentor has something to say against =P.
I'm not advocating to keep the backlash, though, if you can get rid of it then that's good.

[DonRotolo]

Originally Posted by NoodleKnight

Isn't simplicity one of the key elements of this design?
It doesn't look like you can directly swap the chain/sprocket element with two gears, so you're going to at least need 3, which means you'll need to add in another shaft and its bearings to support it, which means more machining, on top of the gear/gear stock you have to buy and machine as well.

Not sure why you think you'd need a third gear in there. Can't you just reverse the motor (in programming)?

Don

[James114]

Is there any interference where the chain is going to be? I think from this angle it looks like the chain is going through the side plates... if it were there.

[Nuttyman54]

Originally Posted by Don Rotolo

Not sure why you think you'd need a third gear in there. Can't you just reverse the motor (in programming)?

Don

I think the third gear isn't so much for the reversal as it is for the distance. Looking at the distance between the two sprockets, you'd need the third gear as an idler just to make it reach.


oh and one more vote for keyway. DO NOT USE SET SCREWS!!! We had them on our transmission this year, and they were the biggest headache ever. We ended up getting the sprockets welded on at SVR.

[Greg Needel]

Originally Posted by Nuttyman54

I think the third gear isn't so much for the reversal as it is for the distance. Looking at the distance between the two sprockets, you'd need the third gear as an idler just to make it reach.

The reason i said no chain was not about breaking it was about miss alignment and the side loads that these modules might take. As far as gears go, an easy way to do this would just to make the top (smaller gear) larger, which would decrease the ratio and then change the ratio from the kit gearbox to each crab module. You can most likely match the ratios exactly with minimal effort.

Encoders, yes measuring them directly on the wheel is optimal but you will still have some backlash issues. But in actuality you can put your encoder anywhere in your drive system and just use a simple function in your software to adjust for actual wheel speed.

As for connecting the top I wasn't really worried about the alignment of the miter gears (as others have said spacers would be fine) I was just looking at the serviceability factor, especially if you decide to still use chain.

[=Martin=Taylor=]

Thanks for the feedback.

One more question for you:

The vertical shaft is welded to the top plate. The top plate is 1/2" thick. Is this excessive? Would 1/4" plate work just as well?

[negfrequency]

its a great design, and if you go about it right you can get complete direction control with only effecting your weight (i hate that constraint)

[NoodleKnight]

Nuttyman54 got it right, its a matter of distance rather than motor orientation.

[Madison]

Originally Posted by NoodleKnight

Nuttyman54 got it right, its a matter of distance rather than motor orientation.

The center-center distance on the sprockets is adjustable within the module -- particularly because you've thus far planned to use chain. I'm sure that you'd have very little difficulty finding a set of gearing that approximates your existing center-center distance and you'd save yourself a lot of headaches for doing so.

#35 chain does not require any more or less tension to operate properly. As mentioned previously, it's a bit more forgiving about alignment, but if you're already planning on using a set of miter gears in your drive, you ought to be able to align even #25 chain with sufficient precision to make it work.

You don't want to use chain in this application without an additional tensioning mechanism, however. Chain stretchs with time and use, so even if you've properly calculated the center-center distance to be a whole or half number of links, it won't stay that way for long. As the chain stretches, gravity is going to make sure that all of the slack ends up hanging off the bottom -- not engaging the sprocket and making the chain jump teeth. It won't be pretty.