Issues with Drivebase

[rainbowbear998]

After competing at the Australia regional, my team realised how difficult it is to get over the scoring platforms with the kit drivebase in the long configuration. As a result, we have now replaced our six 4" wheels with four 6" wheels on the ends. However, our robot now struggles to turn on the spot. At this point, we're pretty sure that the fault isn't in the code or electronics, but we've seen plenty of other teams use this setup with no problems. Has anyone else had the same issue? How did you solve it?

[Ty Tremblay]

Your wheels are creating too much friction when you try to turn. Replacing two or four wheels with omni wheels will help greatly.

[rainbowbear998]

Originally Posted by Ty Tremblay

Your wheels are creating too much friction when you try to turn. Replacing two or four wheels with omni wheels will help greatly.

That's what I was afraid of. I've seen teams make this modification with no problems at all though. Would we get fishtailing if we only replaced one set with omni wheels?

[GeeTwo]

If you only replace one pair, replace the one with less weight (away from the center of gravity). For most robots this year with a convex rectangular base, this would be the rear wheels. We used this configuration for Ultimate Ascent, and we had the two solid wheels in the front and two (idle) omni wheels in the rear. Our CoG was quite near the center of volume. We did have a bit of fishtailing at high speed going forward, but at low to moderate speed, we were fine; I don't imagine you'll be doing much sprinting in Recycle Rush. Interestingly, we had less fishtailing in reverse. This was presumably because more of the weight was over our solid driven wheels.

A better solution is to use six solid 6" wheels as the chassis was designed to use. The center wheels are dropped by 1/8", so the wheelbase is only one half or the other for purposes of "tank mode" turns, but full length for purposes of not falling over forwards or backwards.

[rainbowbear998]

Originally Posted by GeeTwo

If you only replace one pair, replace the one with less weight (away from the center of gravity). For most robots this year with a convex rectangular base, this would be the rear wheels. We used this configuration for Ultimate Ascent, and we had the two solid wheels in the front and two (idle) omni wheels in the rear. Our CoG was quite near the center of volume. We did have a bit of fishtailing at high speed going forward, but at low to moderate speed, we were fine; I don't imagine you'll be doing much sprinting in Recycle Rush. Interestingly, we had less fishtailing in reverse. This was presumably because more of the weight was over our solid driven wheels.

A better solution is to use six solid 6" wheels as the chassis was designed to use. The center wheels are dropped by 1/8", so the wheelbase is only one half or the other for purposes of "tank mode" turns, but full length for purposes of not falling over forwards or backwards.

That's our other option, but the other reason that we only used four wheels was to eliminate rocking in our drivebase, which is caused by the drop centre. I think omni wheels are our best choice at the moment.

[GeeTwo]

Opening with some quotes from another thread...

Originally Posted by NathanCJohnson

But we're using 6" traction wheels in the front and 6" omniwheels in the back (the middle wheels are the 4" wheels that came with the KOP drivetrain).

Originally Posted by Ether

You said the middle wheels are 4" and the front/back are 6". Have you geared them in the proper 6:4 ratio so they are going the same tangential speed?

Originally Posted by GeeTwo

The center axle on the AM14U2 is only dropped 1/8". Since the corner wheels are 2" larger diameter, the 4" center wheels float 7/8" off the carpet. In effect, the 4" wheels are just a spacer for the belt sheaves.

Originally Posted by Ether

Except when going over the platform.

Originally Posted by GeeTwo

Yes, that would be a literal speed bump.

There was a robot at Bayou with a KOP chassis, 6" omnis on the corners, and 4" KOP wheels in the direct drive positions. I asked the driver how it handled, and he said that the 4" wheel actually helped getting onto and over the platforms, though he wasn't very articulate as to how or why. I'm guessing that the lower gear ratio and higher CoF turned the robot so that it was attacking the slope more directly, and gave a bit of a boost on the lift. The undersized middle wheel would produce less rocking as the weight shifted to a different pair of axles, reducing the likelihood of spilling a stack.

[runneals]

Originally Posted by GeeTwo

If you only replace one pair, replace the one with less weight (away from the center of gravity). For most robots this year with a convex rectangular base, this would be the rear wheels. We used this configuration for Ultimate Ascent, and we had the two solid wheels in the front and two (idle) omni wheels in the rear.

We started out with all omnis, but it totally fish tailed. We ended up doing the same thing in 2013 too, although it was 2 omnis on front and 2 rubber wheels on the back (with our butterfly drive). I would definitely recommend using a pair of omnis and a pair of rubber.

[String]

Originally Posted by rainbowbear998

That's what I was afraid of. I've seen teams make this modification with no problems at all though. Would we get fishtailing if we only replaced one set with omni wheels?

If you don't want to risk fishtailing and can't have any rock, consider using 4 omni wheels at the cost of some precision in driving. My advice is to reconsider using drop center, and if your team decides that the rock would negatively affect performance, use 4 omni wheels. In all honesty, Recycle Rush does not require a ton of precision unless you're running out of space on the SPs. I would avoid running only 2 omni wheels because it becomes difficult to drive.

[daliberator]

Originally Posted by String

In all honesty, Recycle Rush does not require a ton of precision unless you're running out of space on the SPs.

I have to disagree. At my regional, it seemed when teams rushed things and were less precise with placement, things tended to go wrong, especially when moving stacks to the SC. However, I agree with your 2 omni advice.

[Dunngeon]

Originally Posted by daliberator

I have to disagree. At my regional, it seemed when teams rushed things and were less precise with placement, things tended to go wrong, especially when moving stacks to the SC.

If precision is required to that extent, you're building robots wrong.

[EricH]

If I might make a suggestion:

You're a little worried about fishtailing, so why not NOT put omnis on only one end, or traction wheels only on one end? Why not put one omni on one end, and one omni on the opposite corner, and do the same with the traction wheels?

See: FRC494 in 2006.

This gives less fishtailing than two omnis on one end, similar stability to a 4WD with all traction wheels, and easier turning than said 4WD all-traction.

[MrForbes]

We might all be overthinking this? Our team used the kit chassis, long configuration, with 6" wheels that are just like the 4" kit wheels. We notched the ends of the chassis for wheel clearance, so we could use the normal length belts. We competed in two regionals, and were semifinalists in one, and finalists in the other. No issues with driving. The rocking is a feature, not a problem, it allows the robot to turn easily on carpet. The 6" wheels let us go over the scoring platforms ok.

[GeeTwo]

Originally Posted by MrForbes

We might all be overthinking this? Our team used the kit chassis, long configuration, with 6" wheels that are just like the 4" kit wheels. We notched the ends of the chassis for wheel clearance, so we could use the normal length belts. We competed in two regionals, and were semifinalists in one, and finalists in the other. No issues with driving. The rocking is a feature, not a problem, it allows the robot to turn easily on carpet. The 6" wheels let us go over the scoring platforms ok.

If you're holding the stack firmly enough, rocking is a feature that helps you drive. If the stack is being supported from the bottom, rocking is a problem that may well be worse than the problem it solves. The ability to turn while maintaining full traction is nice, but not if it means that you regularly drop a bunch of totes on the carpet in essentially random orientations.

[GeeTwo]

Originally Posted by EricH

If I might make a suggestion:

You're a little worried about fishtailing, so why not NOT put omnis on only one end, or traction wheels only on one end? Why not put one omni on one end, and one omni on the opposite corner, and do the same with the traction wheels?

See: FRC494 in 2006.

This gives less fishtailing than two omnis on one end, similar stability to a 4WD with all traction wheels, and easier turning than said 4WD all-traction.

To verify my understanding before I do an analysis - are you advising four wheels, with traction wheels on one diagonal and omnis on the other diagonal?

And oh, yes - what was the aspect ratio (track to wheelbase ratio) of that robot? Where was the CoG? Well, now I've got a nice problem to work on in the evenings as I travel this week, If I actually have any spare time.

[EricH]

Originally Posted by GeeTwo

To verify my understanding before I do an analysis - are you advising four wheels, with traction wheels on one diagonal and omnis on the other diagonal?

Yep. It's been used before, it worked as intended, and it picked up at least one award (can't remember which one offhand, but it was in the 2006 Behind the Design book).