Subject:
|
Thoughts on Turning
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Mon, 25 Jan 1999 01:08:50 GMT
|
Viewed:
|
954 times
|
| |
|
|
I'm one of those newly returned to lego types, drawn in by the appearance of
the MindStorms kit. Being a software guy by trade, programming my creations
is pretty straightforwards, but the mechanical engineering bits are
stretching my brain more. With that said, I figured I'd toss out my current
ideas on steering systems.
Independantly driven treads:
I used this for my first design because it's simple to implement. I
found it unsatisfactory because of the amount of power lost to friction, the
slow maximum speeds and bad variation on the turn rate with different floor
surfaces.
Rack and pinion steering:
This was the obvious second direction, it addressed most of the problems
above but created an unacceptable set of other ones. Horrible turning
radius and no way I could figure to determine the angle turned. (Short of
the compass solutions being discussed in another thread at the moment. :-)
I'd like to hear methods for working around the shortcomings with either
method above, but I've also come up with a different variation that works
adequately (but not perfectly) for my purposes. I haven't seen a name for
this method before, so please bear with the description:
Independantly driven rear wheels with free spinning front wheels, all four
using the largest spoked wheels from the MindStorms kit. This allows fast,
smooth forward and reverse motion. To turn the robot stops and pivots down
a fifth smaller wheel from the middle of the under-carriage which rotates at
90 degrees to the main wheels. This fifth wheel is mounted forward of the
center of gravity to ensure it tips back onto the powered rear wheels and
it's axle goes directly through a rotation sensor (the improvised sort from
the fiber-optic unit) which allows me to determine the angle of the turn
with a good accuracy.
The obvious con with this solution is the need to stop completely to turn.
Getting this to work mechanically was much harder than the other options as
well, but in the end isn't any more bulky or less strong than the rack and
pinion version.
Well, enough rambling for now. Any one else using a similar sort of design?
Thoughts on improving this one? Better ways to implement the first two
options? Other completely different ways of handling the steering problem?
-Blake (budding lego hacker)
|
|
Message has 4 Replies:
 | | Re: Thoughts on Turning
|
| (...) take a look at (URL) There are two robots there based on an adder/subtractor that uses two differentials. The 2nd design (at the bottom of the page) is much more efficient then the first, so if you build one I'd suggest starting there. There (...) (25 years ago, 25-Jan-99, to lugnet.robotics)
| | | Re: Thoughts on Turning
|
| (...) It sounds liek a very nice way of doing things. Does it require three motors tho? One forward reverse, one left right, and one for lowering/raising the 5th wheel? I could think of a way to do it with one motor for left/right _and_ (...) (25 years ago, 25-Jan-99, to lugnet.robotics)
| | | Re: Thoughts on Turning
|
| (...) Cool! me too! (Actually, my wife has been getting on my case lately because I've been "pilfering" some of the kids' LEGO stuff to experiment, so I'm going to need to get more of my own.) (...) One of the cleverest ideas I saw for a steering (...) (25 years ago, 25-Jan-99, to lugnet.robotics)
| | | Re: Thoughts on Turning
|
| (...) I have built something along those lines sort of. I have two driving wheels in the center and four supports that slide (just for lino). The one support lifts the unit enough for the driving wheels to rotate. So the unit doesn't turn, but the (...) (25 years ago, 26-Jan-99, to lugnet.robotics)
|
25 Messages in This Thread:
  
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
