FRC4322's first custom drivebase

[Nxmq99]

http://imgur.com/CZgpNvB

This was built as a test platform for new programmers.

[GeeTwo]

I only see two wheels. Is there a caster under the battery, or perhaps two flanking it?

This appears to be similar to our Ultimate Ascent drive train, which featured two driven wheels at the front corners, and two idle omnis at the rear corners. Interestingly, we found that it drove across the field far more effectively in reverse than in what we had designated as forward. I suspect that this is because the effective CoG of a roughly cubical robot shifts away from the direction of acceleration. With your currently low CoG, you may not see this effect.

[Nxmq99]

There is a caster wheel under the battery. The front wheels are connected directly to CIMs. We may experiment with driving it in reverse for better control. We built this for fun in a couple of hours, and we were impressed when it actually moved! (Everything is held on with zip ties!)

[Chak]

Originally Posted by Nxmq99

The front wheels are connected directly to CIMs.

There isn't a gearbox between the CIM and the wheel?
So... how fast does it go?

[GeeTwo]

Originally Posted by Chak

There isn't a gearbox between the CIM and the wheel?
So... how fast does it go?

Are the CIMs tie-wrapped directly to the peg board? There doesn't seem to be room for anything else, given the size of the wheels and how much they stick above the board.

With no gearbox, I get a free speed of 90+ fps for 4 inch wheels, but a likely torque-limited acceleration of about 5-6 fps/s (less than one-fifth gee) at 40A draw per CIM. Bursts of greater acceleration are possible without tripping breakers.

[Chak]

Originally Posted by GeeTwo

With no gearbox, I get a free speed of 90+ fps for 4 inch wheels, but a likely torque-limited acceleration of about 5-6 fps/s (less than one-fifth gee) at 40A draw per CIM. Bursts of greater acceleration are possible without tripping breakers.

90+ fps means 60+ mph.
Just off the top of my head, I'm sure there's a reason robots don't go that fast. Something about tripping fuses and breakers?
Maybe you can use some Versaplanetary gearboxes to get reduction with nearly no change to the rest of the structure.

Originally Posted by Nxmq99

This was built as a test platform for new programmers.

No encoders? I'm not a programmer myself, but programming for encoders seems like a good skill to learn.

[Michael Hill]

Originally Posted by Chak

90+ fps means 60+ mph.
Just off the top of my head, I'm sure there's a reason robots don't go that fast. Something about tripping fuses and breakers?
Maybe you can use some Versaplanetary gearboxes to get reduction with nearly no change to the rest of the structure.


No encoders? I'm not a programmer myself, but programming for encoders seems like a good skill to learn.

CIMs (as well as pretty much the rest of our motors) are fairly wimpy when it comes to torque, so while free speed may be high, once you put any load on it at all, the speed drops like a rock. That's why we use gearboxes to trade speed for torque.

[GeeTwo]

Originally Posted by Chak

90+ fps means 60+ mph.
Just off the top of my head, I'm sure there's a reason robots don't go that fast. Something about tripping fuses and breakers?

I started to answer this:

Originally Posted by GeeTwo

With no gearbox, I get a ... likely torque-limited acceleration of about 5-6 fps/s (less than one-fifth gee) at 40A draw per CIM. Bursts of greater acceleration are possible without tripping breakers.

5-6 fps / s is not very exciting. 6fps/s would result in a final speed at the end of a 54 foot run of 12 fps, so that 90+ fps would only mean anything if you had most of a mile of track. That's not the whole story - this was based on a 33 - 40lb total robot weight.

Each CIM produces a torque of 100 oz-in. There are two of them, working at a radius of two inches, so you have a maximum traction force at 40A of 100oz, or 6.25 lb. As described above, the 5-6 fps/s acceleration was based on a 33# to 40# gross weight. By the time you load up to a robot of, say, 130#, that 100 oz can only accelerate the robot (even at perfect efficiency) about 1.5 fps / s. Applying that acceleration from a dead stop at one end of a 54 foot field will yield a speed at the other wall of about 6.4 fps - before accounting for frictional losses.

The AA robot I spoke of above was geared down 3:1 from this - a CIMple box and 6" wheels. Even it was rather logy at getting up to any speed. Most FRC robots are geared down at least 3:1 for a fast robot and some are as high as 20:1 for low gear for a robot that needs plenty of pushing power or high acceleration. More typical for a single speed robot is probably in the range of 7:1 to 12:1.

And of course, you can double or triple the torque-limited acceleration without sacrificing free speed by using four or six CIMs.

Originally Posted by Michael Hill

CIMs (as well as pretty much the rest of our motors) are fairly wimpy when it comes to torque, so while free speed may be high, once you put any load on it at all, the speed drops like a rock. That's why we use gearboxes to trade speed for torque.

CIMs do have great torque - for motors. It's just not enough to accelerate a 100+# robot with any authority. Correct as to speed dropping - that's why I used the term "free speed". Most FRC robots have an actual top speed that's somewhere around 60-90% of free speed, depending on the gearing, alignment, friction, and ultimately air resistance. With less gear-down, expect poorer efficiencies because more stuff will be shaking and drawing air.

[evanperryg]

Originally Posted by Chak

90+ fps means 60+ mph.
Just off the top of my head, I'm sure there's a reason robots don't go that fast. Something about tripping fuses and breakers?
Maybe you can use some Versaplanetary gearboxes to get reduction with nearly no change to the rest of the structure.


No encoders? I'm not a programmer myself, but programming for encoders seems like a good skill to learn.

Exactly. This thing is a little dangerous, and extremely inefficient. OP, I'd suggest you ditch the CIMs, but this thing down to about half its current size, and throw some banebots (with gearboxes) on it, unless you have some planetaries for CIMs. What kind of training are you doing with this? I'm assuming teleoperated control, but I'm curious.

[Indiebot]

The students were learning how to control a motor with code and the lead programmer decided it would be hilarious to attach a wheel directly to the CIM to see if it could actually drive so they came up with this as a joke. We thought it would be something funny to show to other teams.

[asid61]

Originally Posted by Indiebot

The students were learning how to control a motor with code and the lead programmer decided it would be hilarious to attach a wheel directly to the CIM to see if it could actually drive so they came up with this as a joke. We thought it would be something funny to show to other teams.

It's an amazing coincidence that it actually drives though.

[Indiebot]

I saw it when they first started it and it could barely drive. Apparently yesterday they changed the wheels to the old 6in KOP wheels and it was a bit faster, anyways they ended up naming it scout. We probably won't work on this much longer as we have other things we are going to be building this offseason.

[GeeTwo]

Originally Posted by Indiebot

The students were learning how to control a motor with code and the lead programmer decided it would be hilarious to attach a wheel directly to the CIM to see if it could actually drive so they came up with this as a joke. We thought it would be something funny to show to other teams.

As long as you don't have any weight other than the very minimum necessary to drive, yes you should be able to break out of zero. Add bumpers and a few manipulators, and you'll likely be stuck. Our Aerial Assist robot started out like this (just a drive platform, CIMs to 6" mecanum wheels). As soon as I realized what was going on, I blew the whistle. We ended up with TB-minis at, IIRC, 10.7:1.

[Nxmq99]

Originally Posted by GeeTwo

Are the CIMs tie-wrapped directly to the peg board? There doesn't seem to be room for anything else, given the size of the wheels and how much they stick above the board.

The CIMs are tie-wrapped directly to the peg board.

[asid61]

Originally Posted by Nxmq99

The CIMs are tie-wrapped directly to the peg board.

With no reduction to the wheel, that may not be a very realistic drivetrain for the programmers. How fast does it go?