pic: ABS-122, shifter-in-tube chassis

[asid61]

Originally Posted by Cash4587

We never had issues with it slipping or jumping or breaking or ever coming off. We just find it very hard to write fast, accurate and repeatable autonomous modes with so much slack in our chain. It gives semi un repeatable results in our vision code when only the front 2 wheels move while the back 4 don't having to make up the distance in chain slack. It would have been much better to just run tensionors from the start.

I see, that's a good reason to run proper tensioning.
How did it cause errors/to what extent? If two wheels are moving, theoretically the whole bot should move anyway. Would direct driving the center wheel help?

[Cash4587]

Originally Posted by asid61

I see, that's a good reason to run proper tensioning.
How did it cause errors/to what extent? If two wheels are moving, theoretically the whole bot should move anyway. Would direct driving the center wheel help?

So this year our robot had about 60% of our weight on the front wheels. Direct driving the center wheel or any wheel might have helped. We ran a single gear reduction 12-66t and then a 12-22 #35 chain reduction to the wheel. (We later switched from 12-22t to 12-28t and is much better) In our case, direct driving any wheel may have helped but we really enjoyed this setup. We would get no movement on our robot for about 1/8th of a turn on the back wheel to catch up in chain slack. The whole robot wouldn't move because we had tons of turning scrub also so I guess a few things would have helped. But in any case, the chain slack and slop in the hex shaft really was a pain when trying to tune our code. I guess that's why 971 goes through so much to get all of the slack out of everything and makes custom hex.

[GeeTwo]

Originally Posted by Cash4587

So this year our robot had about 60% of our weight on the front wheels. Direct driving the center wheel or any wheel might have helped. We ran a single gear reduction 12-66t and then a 12-22 #35 chain reduction to the wheel. (We later switched from 12-22t to 12-28t and is much better) In our case, direct driving any wheel may have helped but we really enjoyed this setup. We would get no movement on our robot for about 1/8th of a turn on the back wheel to catch up in chain slack. The whole robot wouldn't move because we had tons of turning scrub also so I guess a few things would have helped. But in any case, the chain slack and slop in the hex shaft really was a pain when trying to tune our code. I guess that's why 971 goes through so much to get all of the slack out of everything and makes custom hex.

If 60% of your weight was on the front wheels, I would recommend (other things equal) that you drive the front wheels directly; this is ultimately the same reason that West Coast drives the center wheels directly - to make the drive to the most dependable wheels invulnerable to a broken chain/belt.

On the other hand, if 60% of my robot's weight were on the front wheels in a static configuration (and presuming I had at least six wheels), I would be VERY worried about the robot falling on its face in a braking maneuver.

[MichaelBick]

Originally Posted by GeeTwo

On the other hand, if 60% of my robot's weight were on the front wheels in a static configuration (and presuming I had at least six wheels), I would be VERY worried about the robot falling on its face in a braking maneuver.

That is completely dependent on CG height and, given 4587's short robot, probably not a major risk.

[GeeTwo]

Originally Posted by GeeTwo

On the other hand, if 60% of my robot's weight were on the front wheels in a static configuration (and presuming I had at least six wheels), I would be VERY worried about the robot falling on its face in a braking maneuver.

Originally Posted by MichaelBick

That is completely dependent on CG height and, given 4587's short robot, probably not a major risk.

I know it sounds strange, but if this is the case, I would seriously consider not driving the back wheels at all -- save the chain and sprockets and heartache. I might even consider skipping the back wheels and putting some pegs or furniture gliders. Here's why: If 60% of the weight is on the front wheels, the CoG is closer to the front wheels than the middle wheels. If the CoG is low enough that you are "not at major risk" of braking hard enough to tip forward, you are probably "at minimal risk" of ever needing to put weight on the rear wheels, and at "fuggetaboudit" for need to drive those rear wheels.

[Cothron Theiss]

Originally Posted by GeeTwo

I know it sounds strange, but if this is the case, I would seriously consider not driving the back wheels at all -- save the chain and sprockets and heartache. I might even consider skipping the back wheels and putting some pegs or furniture gliders. Here's why: If 60% of the weight is on the front wheels, the CoG is closer to the front wheels than the middle wheels. If the CoG is low enough that you are "not at major risk" of braking hard enough to tip forward, you are probably "at minimal risk" of ever needing to put weight on the rear wheels, and at "fuggetaboudit" for need to drive those rear wheels.

Interesting idea. From what I can tell, unpowered wheels used to be more common on early FRC drivetrains, but have fallen out of style in the past several years. It would probably work well in this case, but I do see one thing that might be a concern. In a six-wheel drop center (which might or might not be what we're talking about, I don't know 4587's robot), doesn't the robot tip backwards onto the back four wheels when driving forwards? Now I know that the position of the center of gravity affects the tilting of the chassis, but this would be on a case by case basis. If the chassis DOES tilt backwards when driving forwards, not powering the back wheels may make acceleration quite slow.

[MichaelBick]

Originally Posted by GeeTwo

I know it sounds strange, but if this is the case, I would seriously consider not driving the back wheels at all -- save the chain and sprockets and heartache. I might even consider skipping the back wheels and putting some pegs or furniture gliders. Here's why: If 60% of the weight is on the front wheels, the CoG is closer to the front wheels than the middle wheels. If the CoG is low enough that you are "not at major risk" of braking hard enough to tip forward, you are probably "at minimal risk" of ever needing to put weight on the rear wheels, and at "fuggetaboudit" for need to drive those rear wheels.

The major issue with this (for my team) is pushing matches: for example, during forward pushing you can easily have 100% weight on the back two wheels.

[Clayton Summerall]

The back wheels are solely used on defenses. Happy we didn't go 4 wheel.

[Cash4587]

Having un-powered wheels in your drive train is generally not a good idea. They create friction and make it really, really hard to turn. Replacing them with a caster or some low friction nylon for example, is also not a great idea because this year our back wheels help us get over defenses, and when you put something in place of a wheel that doesn't roll, and hit defenses as hard as we do, it won't turn out so well. Removing wheels and just replacing them would not be as good as just designing a drive train with 4 wheels only. In cases like this year, we needed 6 wheels to not get stuck on defenses so that's what we went with.

When it comes to drive trains we go with something we are confident with. 6wd and 8wd are types we've worked with before and have given great repeatable results. Adding furniture sliders or caster wheels to our drivetrain is probably something we will never consider doing on our drivetrain.

[Bryce2471]

To OP:

I recognize that this is a mostly theoretical drive train but have you thought about how you would remove a CIM, if it were necessary?

[Cash4587]

Originally Posted by Bryce2471

To OP:

I recognize that this is a mostly theoretical drive train but have you thought about how you would remove a CIM, if it were necessary?

It looks like he oriented the motors in a way that you can access the bolts around the gears. But I am just assuming so I could be wrong.

[asid61]

Originally Posted by Bryce2471

To OP:

I recognize that this is a mostly theoretical drive train but have you thought about how you would remove a CIM, if it were necessary?

I actually added a lightening pattern later such that a ball end allen key can access the bolts to remove the CIMs. As it stands right now, removing the gear in between the two CIMs would suffice to remove them.

[Bryce2471]

Originally Posted by asid61

I actually added a lightening pattern later such that a ball end allen key can access the bolts to remove the CIMs. As it stands right now, removing the gear in between the two CIMs would suffice to remove them.

I guess the question I was trying to get at is: How would you remove the idler gear with the bearing flanges to the inside of the tube?

[asid61]

Originally Posted by Bryce2471

I guess the question I was trying to get at is: How would you remove the idler gear with the bearing flanges to the inside of the tube?

...
Good point. That raises the other question: how would the shaft get in there to begin with?
The lightening holes might let you get away with not removing the gear. But IRL if I made this, all bearing flanges would be on the outside and/or I would use riveted bearing blocks to save space inside the tube. That particular spot is actually easy to put at least one of the bearings on the outside because I have space in there.

[Cash4587]

Originally Posted by GeeTwo

If 60% of your weight was on the front wheels, I would recommend (other things equal) that you drive the front wheels directly; this is ultimately the same reason that West Coast drives the center wheels directly - to make the drive to the most dependable wheels invulnerable to a broken chain/belt.

On the other hand, if 60% of my robot's weight were on the front wheels in a static configuration (and presuming I had at least six wheels), I would be VERY worried about the robot falling on its face in a braking maneuver.

We had #35 chain direct driving the first wheels. Given the strength of 35 chain, and the fact that we put the chains in line, as well as calculated the direct c-c we had no issue other than chain stretch which is to be expected. We had all 6 drivetrain motors above those wheels accounting for the rest of the weight minus the catapult and battery.

With this being said. After 2 regionals, Champs, and an offseason already we never replaced 35 chain, or 25 chain. We were never worried about it breaking either. It is just something that doesn't occur that often if chain is used correctly, which we did, just without tensioners. Which in our case worked very, very, very well, and in many cases would be fine for most every other team in FRC not wanting to do some of the precision complex things things we need to get our code really dialed in.