I’m sure someone has tried this but I have not seen it discussed. Is it possible to use two Brats, one in front of the other mounted on a common chassis, and make it walk like a horse? Maybe spread the distance between the right and left legs apart. With the large flat feet it should be able to balance easier than a biped.
One could certainly use the legs of two Brats, just “hang” them on a rectangular frame.
But the feet? I wouldn’t want Brat feet on a quad! They wouldn’t be useful anyway, as the “horse” wouldn’t lean over to balance on one leg as the Brat does. Some shifting, no doubt would be desirable to keep the COG in line. Smaller feet would appear to be dictated.
Alan KM6VV
You are probably right about the big feet. I’m working on my first walker, and it is not complete, so having the first hand knowledge isn’t there yet. I’m thinking that for a beginning project, big feet will keep “Black Beauty” from falling over so quickly.
You might think it’s funny, but start off with a hexapod! Much easier to make 'em walk with the static gaits.
After that, “advance” to a quadrapod. At that point you’ll be more experienced having had time to experiment with gaits and balance. At least that’s my plan.
I don’t have a Brat, but I did design (from pix) my version of Loki, a 4-servo biped.
Alan KM6VV
Did you use an SSC-32 with your 4-servo biped? Was there pre-loaded gaits or did you have to learn to make it walk yourself?[/code]
there is a lot of open code here on the forums and the lynxmotion website as well as programs and code generators (Powerpod program available for free via lynxmotion.com).
No SSC-32 board on Loki. No preloaded gaits or sequences either, I wrote them myself.
To get started, I wrote R/C servo control code for a minimal (up to 8) servos, it doesn’t take much.
I have an SSC32 and an 18F4620 PIC board in my hexapod 'bot Shelob, but Loki just gets the PIC board, and code in C.
My 'bot program has code that parses commands from the com port, which currently has a Blue-tooth module on it. I can send individual commands to one or more servos in a similar if limited imitation of SSC32 servo commands. I initially ran SEQ to send my Loki servo commands for each state of the gait, and then incorporated the developed sequences directly into the program.
I’ve just added Sonar (SRF-08) and IR (GP2D12) sensors, and I’m currently working on autonomous routines.
Alan KM6VV
Alan,
How well does it walk? do you have any photos or video? If you doubled your Loki project like I described above, do you think you could make it walk like a horse or a cow? I would really like to see photos of your project. What level of programming complexity do you add if you throw on another set of legs? And then the same question if you added another DOF per leg?
(And how do I get to the little box where I can copy what you wrote and then respond)?
Hi Wade,
It’s walk is slow, and quite exaggerated. One of the reasons I built Loki. I Posted photos in the biped Loki thread. A video is planned, as soon as I get a video camera.
lynxmotion.net/viewtopic.php?t=3184
I wouldn’t even try to make Loki into a quad! As mentioned, the walk is quite exaggerated; Loki lifts one foot quite high, then swings it over the other foot. the two feet overlap each other! This is, of course, to keep the base area large so that the opposite leg can be lifted and rotated. Not the kind of motion you want in a quad.
As it is, I’m basically “looping” on a short, 6 state sequence to get a gait. More states, more servos, and I could possibly move a quadruped with 8 servos. 12 would be better. With 12 servos, it may be better to use an SSC32.
press the “quote” button in the upper left hand corner of the message you wish to reply to. It will give you an edit menu. Add text to that.
Alan KM6VV
Now I understand why you think I should have small feet on a quad. Trying to get two Loki’s to walk like a horse would be ridiculous. A new species perhaps?
Nice work with the mill. I’ve got a friend with a waterjet machine, but it would be nice to have my own Sherline and be able to get up at 5:00 AM and try a new idea. Thanks for your input.