Subject:
|
Re: why won't my biped walk?
|
Newsgroups:
|
lugnet.technic
|
Date:
|
Sun, 22 Aug 2004 21:48:48 GMT
|
Viewed:
|
3673 times
|
| |
|
|
In lugnet.technic, Joe Strout wrote:
> In lugnet.technic, Kevin L. Clague wrote:
>
> > The problem is that you are not truly getting the center of gravity inside the
> > footprint of the down foot. The critical part is to get the center of gravity
> > past the inside edge of each foot.
> >
> > This is harder when the inside edge of the foot is not close enough to the
> > center line of the bot. Your feet might be too far apart. Try to go for zero
> > or one studs between feet instead of two.
>
> I'll try that. Though if it works, I may still be baffled as to why. What is
> it that causes the CG to move over the down foot at all? At the start of a walk
> cycle (when we go from both feet flat to one foot lifting), you have one foot
> lifting up -- the bot naturally wants to tilt in that direction. What makes it
> tilt the opposite way, over the supporting foot?
OK, so I think I know why it won't walk.
Right now, the rotation axis of the ankle is on the outer edge of the foot.
This means that when the center post lifts the foot, the inside edge of the foot
lifts, while the outside edge of the foot stays on the ground.
Try rebuilding the foot with the two outer beams on the inside edge of the foot.
I first noticed your very unusual configuration and didn't think much of it.
Now that I think about it, your beam configuration is root cause.
With the inside edge of the foot having the axis of rotation, when you lift the
foot, you force the inside edge of the foot into the ground, and the outside
edge of the opposite foot into the ground. This forces the COG to shift.
Without both feet pushing the COG, you will not transfer from foot to foot.
http://www.brickshelf.com/gallery/kclague/shorter-mm-walkers/pa230090.jpg
Shows three minor variations of walkers, all which have the axis of rotation on
the inside edge of the foot.
Kevin
|
|
Message has 1 Reply:
 | | Re: why won't my biped walk?
|
| (...) I think you're right. For the record, I tried simply shifting the foot plates in one stud each -- and this does sort of work; the bot now wobbles a lot but, most of the time, actually transfers from foot to foot. But I think your analysis (...) (20 years ago, 23-Aug-04, to lugnet.technic)
|
Message is in Reply To:
| | Re: why won't my biped walk?
|
| (...) I'll try that. Though if it works, I may still be baffled as to why. What is it that causes the CG to move over the down foot at all? At the start of a walk cycle (when we go from both feet flat to one foot lifting), you have one foot lifting (...) (20 years ago, 22-Aug-04, to lugnet.technic)
|
12 Messages in This Thread:
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
This Message and its Replies on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
