My RC hexapod arrived last week, and I assembled it over the weekend.
The servos still need some fine calibration, but it works quite well even as it is. It's a very nice design, and while maybe not worth what I paid for it in any rational sense, I don't regret the purchase. Most importantly, it will serve its intended purpose as a research platform nicely. Because, see, I want to build another one.
I want to build one I can ride on. And I've almost convinced myself that this is feasible. Stay tuned...
The servos still need some fine calibration, but it works quite well even as it is. It's a very nice design, and while maybe not worth what I paid for it in any rational sense, I don't regret the purchase. Most importantly, it will serve its intended purpose as a research platform nicely. Because, see, I want to build another one.
I want to build one I can ride on. And I've almost convinced myself that this is feasible. Stay tuned...
no subject
no subject
As for construction of such a beast I suppose you're imagining a hybrid power design - ICE to charge batteries to run giant servos? Or are you looking at just a limited fully electric design?
You do realize that this will cost a bundle in milled aluminum right? Or did you have another material in mind. You could probably design a star-shaped truss style underbelly and use carbon fiber or bamboo in theory then use aluminum for the tips of the "claws".
I so wish battery technology had been allowed to advance at the same rate as the rest of technology. We're still using a lot of the same basic battery structures we've used for years IMHO.
no subject
While I haven't given up on using industrial servo motors, the actuators will probably be hydraulic, driven from a diesel pump. The solenoid valves for computer control are the main cost worry, but I suspect it's really the only way. I'll have to add position sensors and write my own PID-style control loop to mimic the RC servo functionality, but I'm not too worried about that part.
The legs and frame will be welded tube steel. Cheap, strong, easy, and an excellent excuse to finally get my own mig kit. That part will be fairly forgiving -- a lot of heavy work, but the tolerances really won't be that tight.
no subject
If you do want to reinvent the wheel for some reason I'm thinking homebrew CNC controllers hung off the initial control unit would allow you control of all six axis and actuators. In fact that makes it very good for fly-by-wire in that you have a computer that spits out G-Code to the CNC controllers that control the servos and actuators as if they are XYZ+tool changers.
I still think welded tube steel will be too heavy but you're a math person and I'm not. My choice for an initial build would be to use surplus 80/20 (similar to this: http://hackaday.com/2010/12/14/my-reprap-is-bigger-than-yours/ - search E-Bay for "80/20 garage sale") for as much as possible and then use welded aluminum tubing for the rest.
With something this size weight is one of the biggest limiting factors in regards to both speed of movement and power requirements.
Please take all of this with a huge block of salt as I'm in no way formally trained in physics. Just a long-time tinkerer/thinker/dreamer.
no subject
no subject
no subject
I called up Everett Steel in Ballard, and they would sell me as much as I wanted for 1/2 of onlinemetal's price (or, about $15 for a ten foot stick of 1" OD, 0.0625 wall 4130)
no subject
no subject
no subject
18DoF will also allow it to do thing like turn in place, have adjustable stride lengths, multiple gauge to choose from, body height adjustment (great for watching the Burn over the crowd!), and a dance mode. Cause, well, why not? It's just a button on the console and some code at that point. Imagine pulling up to a rave out of a duststorm on one, and then the whole contraption starts bouncing and twisting in time with the music. Legendary.
no subject
This is amazing work, by the way. I can't wait to see more.