Previous post in the sequence:
It has been a crazy 24 hours. I spent quite a bit of time yesterday afternoon/evening agonizing about my next steps in getting WalkyII to improve his steps. I had been watching WalkyII walking in slow motion, looking for any unusual hitches in his stride. A couple of things that I noticed had to do with the stability of his spindly legs. As he starts leaning into his tilt his feet almost touched. This seemed to worsen his balance as his center of gravity moved outside of his weight bearing foot. Being the 'rocket scientist' that I am, I thought that I need to prevent this tendency for his weight bearing leg to flex inward. That, and as I reworked his lower control arms, I noticed that being more rigid than before, they were twisting the servo control horns (due to the fact that the lower control arms were not parallel to the ground). So, I quickly designed a couple of fixes for WalkyII. The first fix was a slider/bumper that attached on the bottom lip of the body, that would help to keep WalkyII's legs from pitching inward. The other was a bracket for the leg to lower control arm attachment point that would lower the leg end of the control arm ball joint.
Here are some views of the resulting designs (from left to right - top of bumper, bottom of bumper, control arm pivot extension):
So, time to fire up the 3-D printer and build WalkyII some new parts. Here are some pictures of WalkyII after a bit of surgery:
So, now WalkyII legs are straight and the lower control arms a parallel to the ground. Do you notice anything else about WalkyII? How about those oversized cardboard feet? "My, what big feet you have?" "All the better to balance on". More on those feet in just a second. I installed the slider/bumpers and WalkyII's legs were straighter, but now his gait had gotten even crazier. Even with the 'tilt' parameter cranked all the way up, WalkyII was not able to fully lift his opposite foot. And with the 'tilt' up full, WalkyII only got wobblier. So bad in fact that he toppled over in one of his test and he broke is rear-view I/R sensor. Bang! Snap! Broken connector. Turns out those Sharp I/R distance sensors use this soft, fiber-like PCB instead if FR4, and the fall of 10 inches was enough to snap off the small PCB tab that the connector is soldered to. Darn, unlike my normal tendencies to 'over order', I only purchased the two sensors. So far, I have the PCB glued, but I still need to repair 6-8 traces that ripped off the PCB as it snapped. (something to add to my completion list).
Anyway, this led me to the real issue. As WalkyII tilted and waddled, his center of gravity continued to move out side of his weight bearing footprint. Oh, he needs bigger feet! I was explaining this to my wife and describing how I need to design and print some new feet, when she suggested using cardboard to test idea, prior to running off and spending more time in the office. Hey, that is a good idea. To that end, here are WalkyII's temporary feet.
I will follow up with another update, but i really wanted to share WalkyII's new video. In this clip, his 'tilt' is still at the maximum setting (full swing of the servo). He pitches wildly to the side, but is considerably more stable.
More to come!