1541 Prototype Drivetrain

[bombadier337]

We worked Saturday to get the omni-directional drivetrain setup and it works great! It's just mounted on some plywood, but does a good job. We used Dr. Joe's instructions for using Dewalt transmissiona nd AndyMark omni-wheels. This was before we wrote the code for driving it (it's just using the stock two channel mixing), but it now drives using two joysticks. The control setup is very similar to Halo with the left joystick controlling forward, back, strafe left, and strafe right, and the right joystick controls rotation. This is just a video of us seeing how fast we can get the bot spinning when we first got it running.

http://www.sartanyac1541.com/videos/RobotSpin.mpg

[SURVIVORfan44]

Wow I'm impressed. I was impressed also last year because 1541 was a rookie team that did better than most of the teams last year at VCU. I can't wait to see if the crushed cans will make another appearance on their robot this year.

[Billfred]

Awesome stuff!

Are you planning to shift those DeWalts? The audio in the movie made me wonder.

Maybe I'll have to find a way to get to VCU after all...

[sanddrag]

Can it get up the ramp with those wheels or do you need an assitance mechanism?

[Collin Fultz]

It looks good.

I would like to say...WEAR SAFETY GLASSES!

The robot almost got you. Seriously. Safety Glasses.

Jen didn't want to wear safety glasses, now she doesn't have to.

[Dan Richardson]

Heh that was an entertaining video. I was a mentor on 1083 (which no longer exists unfortunately) last year and we used omni bot ( most interestingly characterized by the This is How we Roll omni pictures at nationals ). Spinning is one thing but just a little for warning the andy mark wheels are great but you may find when programming the robot in all its axis that they cause quite a bit of friction. We had to modify them twice to get them working more effeciently. The rubber sides would tend to want to slide over the aluminum plate and bind.

We solved this relatively simple, we dissamebled them and milled out the gap a bit larger. We then inserted washers inbetween the metal plate and the roller. The washers lessened the friction between the rubber rollers and the plate, it also sured up the integrity of the roller by not allowing it to bend and make contact with the plates. You may not find it nescesary but we did, and this was a simple solution that worked wonders for us.

Omni is a great drivetrain, and I think will be especially interesting and fun for this years game.

[bombadier337]

As to getting up the ramp, we can get it, but we have can't go up the middle. It has to lean against on the plywood sides. As for shifting, we haven't decided. We played around with shifting, and the spinning was done is second gear. We already have the servo mounting plates, so it's just a matter of assembling them. As for the safety goggles, I definately agree. We have them on almost all the time, but I hadn't even realized I wasn't wearing them until you pointed it out. Thanks for the feedback guys!

Edit: The crushed cans will be back.

[kmcclary]

Wow... VERY cool...

Originally Posted by bombadier337

As to getting up the ramp, we can get it, but we have can't go up the middle. It has to lean against on the plywood sides.

OOC, a couple of questions:

1) Are these the 8" Andymark trick wheels, or the 6"?

2) Whose gearboxes are those? That's a sweet package! How are they
switched, and what gear ratios do they provide in high and low?

3) What sprocket pair ratio are you using between the gearbox and wheel?

We been looking at some AM trick wheels ourselves (and ordered a
few for tests... No ship yet, so we're looking at other wheels too).

We were hoping to get away with only a single speed transmission
this year. Four AM or COTS 2-speed transmissions would be very
expensive (worse, EIGHT so we'd have a practice bot too).

But my REAL fear has been that omnis simply wouldn't get us up the
ramp at ALL with a one-speed transmission, without a gear ratio set
that would "crawl" on the ground game.

Now, from what I'm reading, I'm even *more* concerned that
people are saying that they may have to be "tweaked" with some
machining, just to keep them from binding.

How long did it take you to "fix" the wheels?

Based on your experience, any comment on what ratio we'd have
to have to get 8" AM omniwheels to climb it with a full weight
bot, using one regular CIM per wheel? FYI: We ordered the white
rollers (0.9 COF) for our "testing bot", because they wouldn't
mark up the school's and sponsor's floors as much as the black
rubber ones would for demos...

I dunno... Though VERY cool, from what I'm hearing mixing this year's
ramp and AM trickwheels doesn't sound like a promising solution.
Even if they DO arrive, we may want to stick with our "other plan".

Is there someone here with more experience with them know of
a decent gear ratio to make them work with this year's ramp?

- Keith

[bombadier337]

The wheels are 8" stock trick wheels with the epdm (?) rollers. I don't know if we are going to keep them though, because they compress and we don't want to damage the floor. The acetals roll amazingly but don't have enough friction to make it up the ramp, so that's kind of up in the air as to what were going to do. The gearboxes are from Dewalt XRP Drills, interfaced to small CIM motors. We followed the whitepaper precisely, which can be found here.

http://www.chiefdelphi.com/forums/pa...le&paperid=398

We have a 9-tooth sprocket on the gearboxes and an 18 tooth sprocket on the wheel, though we may end up increasing the size of the sprocket on the wheel. Hope that answers any questions.

[JamesBrown]

I realize that this is still an early prototype but mounting the batery closer to the middle will help you rotate(so you rotate around the center rather than in a larger circle. It looks really good. DO you have any plans to integrate a gyro so you can rotate as you translate?

[Arkorobotics]

Doesn't look safe in a way, but can you drive stright accurately with that?

[Alekat]

So this obviously gets rid of most all turning friction, but doesn't it simply cause more friction to drive straight?

[bombadier337]

Takes a little more to drive straight, but yes, we can drive it straight with relative ease.