Custom "Car Style" Drive System

[Mountain Man]

As a part of the Mountaineer Area RoboticS (MARS) 2614 team's internal "pre-season" competition, the Deimos subteam has decided to make an unconventional yet innovative "car style" drive system. It incorporates a differential to distribute power to the rear wheels while the front wheels are steered with a tie rod steering system.

Here are pictures of the differential. More pictures to come as more things get assembled.

[rahilm]

Where is the input on the differential?

[AdamHeard]

It seems like you may want to actually remove the bearing on the idler bevel, right now you have a fully open differential. If you replaced it with a bushing of higher friction (or even better, some sort of clutch setup), you could easily integrate a limited slip differential for better performance.

Originally Posted by rahilm

Where is the input on the differential?

I would assume a sprocket or gear will be fastened to the face of the housing.

[sanddrag]

I too am curious where the power will be input to this device. Typically it is driven through the case/housing/carrier.

The design also has a problem, in that each gear is supported by only one single-row bearing. You need another bearing on there, close by. Then use a coupling or universal joint to get the power out and away from the unit. Basically, make the differential it's own robust unit. Shorten those shafts, and have separate ones that go out to the wheels. The alignment of the gears should not rely on a bearing that is many inches away down the shaft, located on a completely separate mounting piece. As shown, the gears will wear quickly, and possibly bind.

As Adam said, some sort of friction clutch (perhaps even adjustable) on the center gear could be beneficial, to help prevent single wheel burnouts. Watch the movie "My Cousin Vinny" It is explained well.

Remember, design with the idea that you may make it to finals, and you wouldn't want it to be on the verge of failing when the stakes are high.

[rahilm]

Edit: Never mind, wasn't thinking straight

[Mountain Man]

Originally Posted by AdamHeard

I would assume a sprocket or gear will be fastened to the face of the housing.

Correct, there will be a sprocket for chain attached to the side of the housing. It will be driven by an AndyMark toughbox situated above the differential.

Originally Posted by sanddrag

The design also has a problem, in that each gear is supported by only one single-row bearing. You need another bearing on there, close by. Then use a coupling or universal joint to get the power out and away from the unit. Basically, make the differential it's own robust unit. Shorten those shafts, and have separate ones that go out to the wheels. The alignment of the gears should not rely on a bearing that is many inches away down the shaft, located on a completely separate mounting piece. As shown, the gears will wear quickly, and possibly bind.

There will be "mounting brackets" that contain bearings that the shafts will go through to support the weight of the robot. There will be one on each side very close to the differential and one just inside each wheel out at the edge of the frame.

We have tossed around the idea of a "clutching" mechanism on the idler shaft to limit the slip. This is the first prototype, however, and if we do decide to use something like this for competition we will more than likely have to improve the robustness of the design.

[JamesCH95]

Originally Posted by Mountain Man

As a part of the Mountaineer Area RoboticS (MARS) 2614 team's internal "pre-season" competition, the Deimos subteam has decided to make an unconventional yet innovative "car style" drive system. It incorporates a differential to distribute power to the rear wheels while the front wheels are steered with a tie rod steering system.

It certainly is unconventional in FRC, but is there some trick that makes this innovative in the grand scheme of things? Still unique either way.

For bonus points, drive your front wheels with hub-mounted CIMs on planetary gearboxes or even through more beveled gears.

[Al Skierkiewicz]

I am just an electrical guy but in this configuration don't the two output shafts rotate in opposite directions? Is this only the inner mechanism and you intend to add an additional mechanism that it is rotated by the drive motors? If that is the case, wouldn't a second idler, opposite the one shown, be a good idea to distribute gear loading?

[MrForbes]

It's a differential, as mentioned the square tubing housing is the "input", and the two shafts are the "outputs". If either wheel loses traction, then the other wheel won't be driven, as the slipping wheel will just turn at twice the "input" speed.

Differentials are great in cars, but I don't really see much application in the robots we build. Might be because we don't intentionally push our cars up against other cars, or barriers....which is natural FRC robot behavior.

[Aren_Hill]

right there is a beautiful example of a differential setup from 2008, team 34 i believe.
(roboticWanderor's picasa gallery of awesome)

adjustable limited slip

[Al Skierkiewicz]

That's just darn pretty!

[EricVanWyk]

Originally Posted by Aren_Hill

right there is a beautiful example of a differential setup from 2008, team 34 i believe.
(roboticWanderor's picasa gallery of awesome)

adjustable limited slip

Thanks Aren, that makes a whole lot more sense now!

[Al Skierkiewicz]

Thanks, Jim. A Michigan Team (GM sponsored) had a four wheel drive robot with differentials many years ago. It was surprisingly compact. It had some issues, but was interesting to look at.

[MrForbes]

If you want to really show off your machining skilz, make a Torsen differential. It acts like a limited slip, but requires no clutch.

[-=Pz=-]

That does look sweet piece of work