Re: Team 95 2017 Build Thread
 

Good question!

The bearings are flanged, and the flanges are on the outside of the drive pod. On either side of the flange is a Pem nut with a bolt and washer that retain the bearing in the side plate. The axles are all live axles with spacers between the side plates and a bolt+washer on either side of the flanged bearings to keep them located.

It is a little convoluted, but it offers some advantages:
-We can replace bearings with axles in place, and without a press (not that we've needed to...)
-The axles' bolts can be made very tight without side-loading the bearings like many dead-axle setups do
-The bearing is huge, which gives the side plate a lot of bearing area
-Fabrication is easy for us, we make side plates and axles then bolt everything together! The holes are a single setup on a mill, and the axles are cut to length on a chop saw, then drilled and tapped in a lathe (though this could be done free-hand).
-The live hex shaft setup offers great packaging options as to where we locate sprockets, chain, wheels, etc and responds well to wheel/drive/location changes.

We used this style last year and were quite pleased with it. We converted from #25 to #35 chain and swapped around where we located colson and pneumatic wheels a few times; we were able to do so with little pain. It held up wonderfully to 4 events of hard driving, needing no significant maintenance besides occasional chain tension adjustment. That usage included semi-finals play and two finalist runs, one of which we played defense against some rugged teams including 1519 and 319.

Re: Team 95 2017 Build Thread
 

I like this method. I saw it on one of 254's robots and wasn't sure if they had just threaded a hole or used nuts. I hear that chains tend to stretch a little, but we've never witnessed it in one of our robots. How would you tension (or re-tension) a chain with this configuration?

We were considering using this method with our winch setup instead of our usual t-slot, but couldn't figure out how to tension it without adding complexity.

Re: Team 95 2017 Build Thread
 

Tensioning will always add *some* complexity to the system, but it doesn't take too much.

Last year I designed a little part we called the 'ninja ghost' to do out chain tensioning. See below. It's a special cam profile that rotates to tension the chain while passing the chain's tension force through the center bolt, resulting in near-zero net torque on the tensioner. The recess in the ghosts' faces fits a 10-32 nut, making the tensioner a single-tool adjustment.



This year we're using a delrin rod running over a bolt to simply push the center of the chain. It's crude, but quite effective.

Climbers respond well to very basic tensioning methods because only one side of the chain is ever under load, generally speaking. Thus off-the-shelf methods designed to always work on a chain's slack side are quite effective. These methods do NOT work well on drive systems that are reversed. I would think a simple plastic guide bolted through a slot or series of holes pushing on the slack side of your chain would work just great.

I feel obligated to point out that chains don't stretch in the technical sense that the chain is undergoing plastic deformation in tension. Each pin and roller pair wears just a tiny bit, and this added slop makes the chain effectively longer. To that end: good lubrication and light loading will help keep chains from elongating over the course of a season. We use PTFE-based dry film lubricants, they are perfect for FRC applications like chain and open gearboxes.

Re: Team 95 2017 Build Thread
 

Seems like you're keeping some pretty big keep-outs above each drive rail - what's the thinking there?

Re: Team 95 2017 Build Thread
 

Another good question.

-I haven't shown the battery in a rendering yet, so I'll fix that...
-The drive pod standoffs are great pick-up points to lift the robot by, so I like to keep the corners of the drive pods accessible
-I don't model 100% of the components, so I like to leave a reasonable margin for things like air cylinder fittings, which can stick out an inch or so
-I try to keep mechanisms to be as big as they need to be, but not as big as they can be



To expand on that last point - in prior years we've made some bad design compromises thinking 'our intake should be as big as possible' or something along those lines. It's been painful to work around such dramatic compromises later in the season, so I try to build in some wiggle room.

This policy has also kept us under-weight from 2010 on... which has been really nice.

Re: Team 95 2017 Build Thread
 

Per request, I've uploaded the ninja ghost model to CD media: https://www.chiefdelphi.com/media/papers/3342

The principle of operation is that the helix surface is always normal to the center pivot bolt axis, so the reaction force of the chain doesn't exert a torque around it. The anchoring bolt can go through a curved slot, as it did in our 2016 robot Kovaka seen below. Alternatively, the anchoring bolt could go through a match-drilled hole once tension is drawn, or through a series of holes. Implementation can be varied to your hearts content! In a slack-side-only implementation it would be feasible to only use the center bolt, especially if there is a good frictional material between the plastic cam and whatever it is anchored against.



Good luck! If you do decide to use/copy/modify/whatever the design, just cite your source! (and definitely continue to call it a ninja ghost)

Re: Team 95 2017 Build Thread
 

Shooter prototype is pretty well sorted at this point. 1:1 775 Pro motor with a 2 1/2in colson wheel, approximately 1/4in of compression.



Assembly and fabrication is progressing nicely. Starting with a little bit of welding.




Practice robot's gear receiver assembly is together.



And then mounted!



It works pretty okay so far!

Re: Team 95 2017 Build Thread
 

Darn 5-image limit...

The practice robot is starting to get its guts put in.



Since we've elected to build a practice bot this year we have also decided to dress up our competition robot. Time for paint!



We also had, for the first time, a bad hex bearing. Hex was broached off-center. Very weird...

Re: Team 95 2017 Build Thread
 

Looking pretty cool so far!

Can I ask where you got those CAN wire connectors and how you secured the wires in them?

Thanks.

Re: Team 95 2017 Build Thread
 

Female Housing
Male Pin
Male Housing
Female Pin (Socket)
Crimper
Extractor


There is probably a cheaper place to get them, but McMaster works for us! These connectors worked very well for us last year.

Re: Team 95 2017 Build Thread
 

Have you indicated a shaft in the bearing? How far out of center is it?

Re: Team 95 2017 Build Thread
 

We did not indicate it. One could put a small section of hex shaft in the bearing and watch it wobble while turning it by hand. I would guess around 0.03in out of concentricity at least.

Re: Team 95 2017 Build Thread
 

Hey James, glad to see the thread coming back again.

Out of curiosity - do you plan on holding a gear to start? I'm not seeing how you will be able to since the release mechanism goes outside the frame perimeter.

Re: Team 95 2017 Build Thread
 

The release mechanism is within the perimeter at the start of the match. It only goes beyond (if at all) once the match has started as they release the gear on the peg. I don't see how this is an issue?

Re: Team 95 2017 Build Thread
 

Ah, upon examining it further the whole mechanism slides in and out.