Hello forum,
This is my first post here - I wonder if someone can help.
Myself and a friend have decided to build a hexa-bot of our own design. I am working on the mechanical engineering side, my friend will be developing the electronics, and we will share software build.
Currently, we are planning for the bot to be a 6 leg design, likely with a jointed middle body, and we will be adding collosion detection - currently planned to be ultra-sonics or laser based.
Electronics will be Arduino based, either with a single central Mega, or a Nano per leg and then an Uno as a central processor. The second route is a little more expensive, but makes for a much more adaptable platform, and can handle a much more complex set of instructions with a LOT less programming complexity - which is good for me, as I'm the weaker programmer!!
We both have good backgrounds in engineering and code development, but essentially none in robotics itself.
Lastly, we are planning an iterative development process - start simple, for example get something walking in a straight line first, and then build complexity in over time.
So - thats what we're looking to do, and our level of experience.
What I am currently looking for at the moment is papers, plans, articles etc about the leg design itself. I understand the concept, but I am looking for specifics. This could be general theory - the dynamics of legs, through to actual built designs.
There seems to be a lot available for robot arms, and their grippers, but I am drawing a blank on the legs, apart from smallish images of kits, such as the Phoenix Hexapod from Lynxmotion, and some YouTube videos.
I was in a similar position regarding "robot arm grippers" until I read some of the tutorials on this site, and found out the correct term is "End effector" at which point I found lots of information on Google. I'm wondering if I just don't know the correct search terms?
It looks like there are options all the way from 2DOF to 4DOF, and I would like to understand the mechanics of these options a bit better.
Also - I note that none of the hexabot designs I have seen seem to have encoders built into the joints. I checked a few different servos I have knocking around, it it looks like each of them have between 0.5 degrees and 1.5 degrees of backlash in them - which will build up during a walking process I would guess.
When the servo controlled legs are built, is the compensation built into the software, or is there a mechanical method of compensation. The reason I ask is that if there are no clever mechanical methods, it means that we're going to be having to measure back EMF from 22 servos, and thats going to need a lot of complexity to build the measurement circuts for.
Is it the fact that the legs are in tension, and this takes up the backlash - i.e. therefore it can be ignored?
Many thanks in advance for any assistance you can give
Steve
