2 truck drive train

[Max Lobovsky]

Originally Posted by Alex1072

Because this acts like a swerve/crab:

it can go like this:

\ \

\ \


or turn in place like this:

_ _

_ _


and what I like most: because the turning is controlled by truck rotation, you don't have to control motor output carefully to get it to go straight. Just servo the steering trucks. It will go straight or turn like a car. The differential will minimize wheel slip.

Now draw that again, but get rid of either the right or left set of wheels. I dont think you understood me. A truck does not need to have two wheels is what i was saying. It would be the exact same thing if it had only one wheel, and it would also solve the problem of inner and outer wheels turning at different radii that Rote AUgen mentioned.

[Alex1072]

If you only have one wheel, the you run into the problem of having to control motor outputs to go straight. I think I might not be understanding what you're saying though.


EDIT: I get it now. There is no inner/outer wheel speed problem because each truck has a differential in the gear box. Having the 2 wheels makes it so that there is no weight on the castors normally. With 2 wheels the castors are only used when you the angles of the trucks get too close to 90 degrees from the robot frame.

BTW I made slight mistake with that ASCII diagram:


the turn in place would look like this:

|
|

|
|


with castors on the left and right for support

[Max Lobovsky]

Ah yes, one wheel per truck would cause castors to be used during all movements. Good point.

[RoteAugen]

now for the real brain-burner:

How to reduce sideways friction on your wheels? (This allows the usage of really fat, high-traction tires with your superior steering system!)

in order to do this, however, the truck idea will not work. why? because if the outer radius is bigger than the inner radius, the wheels have to be at a different curve, no?

there are a few solutions, but I will give you guys the first one 138 used to get you guys thinking in terms of circles


||...........||
..\______/ ................where ||= the wheel assemblies, and
............................. ___ and \, / = aluminum barstock.
...______
../..........\
||...........||

...just pretend the periods don't exist... CD automatically deletes the spaces i had so

now- each of the center bars must move along the same, or initial plane, so they must have a sliding pivot point in the center of each bar. This is what car companies use for the front wheels of their creations.

at each angle, there must be enough slack, or universal ball joints to allow the pivoting of the wheel assemblies.

How to move these simple assemblies I leave to you all to figure out

as I said, if you wish to see a more efficient design, check out Team 138 in the Pits and ask someone about the steering system. They will give you a packet if you are lucky, and it will describe Ackerman steering, and you can see for yourself how we acheive this.... just one thing: DO NOT TAKE PICTURES. DO NOT MAKE SKETCHES. we are very proud and protective of our designs.. but we will not be unkind to you

[Max Lobovsky]

The design posted here calls for a differential. A differential solves exactly the problem you describe. I suggest reading about it on howstuffworks.com or something, but essentially, it will allow one wheel in a truck to spin faster than the other, and therefore drive over a different radius. [edit] heres a relavent page http://auto.howstuffworks.com/differential1.htm[/edit] In fact, this is what car companies use to let cars turn.

advanced traction systems use these, but if im not msitaken, they are even more complicated because they allow forcing more power to one output shaft or the other, not just letting them balance out. [more_edit] howstuffworks.com is great, keep reading in that article, they explain this, too. Check out the viscous coupling. I'd like to see some team implement that [/more_edit]