Non-driven dif

[TasV]

I was checking my Son's latest robot tonight and noticed he'd put a differential on some non-driven wheels and my first thought was... cool! But then I started wondering... would a non-driven differential even work? Would you still get the same beneficial variable rotation of the inside and outside wheels when turning or would it merely serve as a connection between the left and right wheel that still allows them to turn independently (see pic.. forget that he's forgotten to install one of the bevel gears... he's doing pretty awesome for a 12yr old).

[Tubby]

It would work either way, if it's a non-driven steering axle then it would allow the outside wheel to turn faster than the inside and if there is direct drive to a wheel either side of it then it would allow one to turn while the other stays put so it could turn 360 degrees within its own length.

If it's just there to connect the axle together then I don't think it'll do much. I guess just wait until he's finished it to see whether or not it works as he wanted it to. On the face of it though it does seem quite clever.

[TasV]

I'd like him to try and make it a little more robust with a 5x7 frame... but having tested it on a linefollow track tonight I'm still not sure if it makes much difference. It seems to work well... but I'm not sure if it's any better than if it wasn't there and the two wheels were disconnected. Because the two back wheels (it's front-wheel drive) can also be attached so they aren't connected in the middle we compared it and the inside and outside wheels still turn at different rates when the front wheels are turning the robot. It doesn't turn like a conventional car either because it's running a linefollow program and it turns by driving the left and right motors on the front wheels alternately... so it wiggles down the line.

Edited August 8, 2013 by TasV

[Aqualize]

It doesn't do anything good there unless he plan to drive it later or put a fake engine to it.

[Balrog]

If you have two driven wheels in the front (with seperate motors I guess and steering by having different speeds) wouldn't it be better for steering if there is a single rotating wheel in the back?

Because when your robot steers now, it's movement is hindered be the non-steered back wheels. And the wheels are skidding over the ground?

[piterx]

shouldn't it be the same as having a different axle for each wheel?

[TasV]

They are omiwheels (two of these: http://www.rotacaste...ub--2-pack.html) on the back (holonomic). This gives him increased stability without affecting the turning ability.

shouldn't it be the same as having a different axle for each wheel?

Yes... that's what I thought too and is why I am questioning it.

Edited August 8, 2013 by TasV

[piterx]

you could make a brake system out of that :)

[TasV]

Out of the rear differential or the omniwheels??? The breaking in his robot was achieved by programming motor breaking into it. I've got him to remove the diff. and brace it up with a 5x7 frame... we have one last engineering challenge to solve (proving difficult) and he is done and ready to compete!

Edited August 9, 2013 by TasV

[chorlton]

They are omiwheels (two of these: http://www.rotacaste...ub--2-pack.html) on the back (holonomic). This gives him increased stability without affecting the turning ability.

Yes... that's what I thought too and is why I am questioning it.

I think it is functionally the same at the moment, but you are halfway to adding RWD, and can then explain the use of a differential on a real-world car. Educational! :)

[SNIPE]

it's like having an axle per wheel but one wheel can counter rotate the other with this diff however it might not happen since the wheels aren't driven.

[TasV]

Rear wheel drive and 4-wheel steering is something I'd like to try with the robot... but at the moment the programming of this is beyond me. Combining steering with a line follow program to make an autonomous steering robot would be very sweet. How this might happen is starting to take shape in my head... but it's still too hard for me at the moment. The design and engineering side is not so hard... it's the programming that has me conceptually stumped.

Edited August 9, 2013 by TasV

[TasV]

His robot looks like this at the moment, sans differential:

The front tires are Losi Rock Crusher Micro Crawlers and the rear ones the Rotacaster Ominwheels. Pretty much done now. I just want him to adjust the position of the LED lights next to the light sensors and drop the light sensors half a brick.