Hey everyone, I thought I would share this project I have been working on, I’m new to this forum, and am looking for some feedback.
I made this biped with the intention of building a full body. I merged MicroRax parts with the lynxmotion parts, a little bit of hacking and cheating was necessary to get everything to fit together nicely. At this point my goal is to shift weight to one leg and try to lift the other. If anyone see’s any obvious design flaws let me know, I’m work as an Animator for a living, my background is not in engineering, so my approach has favored aesthetics because I don’t know any better.
I’m controlling it with an Arbotix micro controller. I’m wondering if anyone has made an Inverse Kinematics setup using this controller, or something that uses the Arduino environment.
I noticed the control rod bows a little bit. Have you thought of using a thicker rod with short pieces of threaded rod coming off each end? this would handle more load.
i saw this on another site a few days ago. great work. i have a solution to the bending rods.
when working on a new leg design i had the same problem. i used tubing to stop this. also i have problems with the rods popping out from the ball joint. i used lock tight glue for that. http://i531.photobucket.com/albums/dd355/innerbreed/100_2336.jpg
That’s exactly what I had in mind. Using rod or tubing like in your image. I noticed the video has only one rod but the still photos have two. Having two is definitely better than one, but I think going one step further by using a design like in your example image is an even better solution. I also like the look of not seeing threads the entire length of the rod.
You can get un-threaded rod, and thread just a short section of the ends.
If you are comfortable with metal work, you could probably just solder in a short threaded piece to each end of the tube (un-threaded where you solder). you’d have the stiffness, but not the weight. Cut the heads off small bolts to get your end pieces.
this is not the case will them really. i just had a dodgy out of spec pack. jim offered to replace them for me but i had loads left from a previous order.
Since the video I doubled up the rods in the knee, they don’t bow as much anymore, it wouldn’t hurt to reinforce them. Where can I find thicker threaded rods? Or the tubing to put over them (like in the example image)
The shock are there to take some of the strain off the servos. It also couldn’t stand up while turned off, but with the shocks it holds itself firmly in place. I’d like to find a more efficient way to do this, It feels like the setup I have now may work against me, creating more resistance for the servos, (though doubling up the servos may help too)
What is the best way for one to go about threading their own rod, assuming I don’t have a metal shop?
thanks for the feedback everyone, its most appreciated
Check out your local hardware store. You should be able to by 4-40, 6-32 or whatever dies. Dies are 1" or so diameter steel disks, with “teeth” on the inside that can cut threads. Then one usually buys a die holder (like a ‘T’ handle) to hold them. You hold the rod in a vise, if you have one, and then place the die squarely against the end of the appropriately sized rod (probably steel). With some pressure applied, you rotate the die, and it’s “entry” side will start to cut threads on the rod. Cutting oil helps. Plenty of instruction out there on cutting threads.
MSC, EMCO any many, many other stores sell tools over the internet.
Joining a metal working forum can be quite useful for robot builders.
Just using springs will save weight over the shock setup. If you don’t want the leg rods to flex, making a thin 90 deg brace for the rods from aluminum flashing might be possible (make or use a tool like below to bend them). On the other hand, with large bots some flexing might be of benefit. Small bots slip and slide to accomidate less precise control. In big bots relying on slipping and sliding might not be desired and can result in binding and component failure when they can’t. Some flexing can reduce peak component strain to where parts don’t break.
