Re: Latest creation, 12 motor centipede

Hey, this seems to be an awesome robot. I saw the 3 movies, great job done pal. :) I really like it how it works.

Best regards, Refik Hadzialic

The heatsinks are WAY over-kill. The NMIH-0050 is good for 5 amps continuous, I need 2 peak at best. It was a convenient way to mount, that's all.

The problem with the body, is that it is custom fitted to those motors. I have a limited supply of them. I may consider finding a common motor to base a similar design on.

Mike

Hopefully, there will be newer movies up tonight.

Just don't hurry up! Every good work needs time :) I am building a hexapod with many sensors but you will see it when it is done just it needs quite a lot of time and one problem more, where I live it is hard to get the electronic parts. Long time ago we stopped the industry, hope it will recovery soon.

Best wishes brother, Refik Hadzialic

Wow ! great work Mike, your Centipede as really come along since the last time i saw it. Now if you could just get rid of those whegs for some real legs :)...just kidding! great work on the design.

Larry.

As far as I can tell, the whegs are just a spiral wheel right? What's "wrong" with it comparing to legs? Certainly, a wheel-wheg can spin faster?

Now I am inspired to think about smaller versions....

-- // richard

In all honesty,

I really don't like the whegs all that much, since they lead to a lot of vibration. Tonight I am resoldering batteries, and silicone gluing connectors in place.

"wrong" with it comparing to legs? Certainly, a

Great stuff, Mike!

You mentioned in another post that the bodies are custom laser cut to fit the motors. Just doing a little brain storming - you could probably find a standard size of schedule 40 PVC pipe that will fit your motors. Try to find a size whose inside diameter is a little larger than the outside diameter of your motors. Then use some rubber like from a bicycle innertube, and wrap some of that around your motor to take up the extra space inside. Cut a few 1/8" slits in the pipe ends lenghtwise a couple of inches long. Then squeeze in your motors wrapped in the rubber and use hose clamps to tighten down the ends of the pipe squeezing the motors and holding them tightly in place (which you can do after you cut some 1/8" slits using a standard table saw). The rubber will not only take up slack in the fit of the pipe but serve as a natural vibration aborber too.

Thus, each segment could be composed of a single length of PVC with a motor stuffed in each end. Drill a hole in the middle to bring out the wiring and attach your coupling to join to the next segment.

Just a thought ...

Cheers,

-Brian

That's just about how I built the first prototype. ( great minds think alike ) I machined standard "T" joints to fit the motors, and stuffed springs into the joints. Holes let the wires out.

Prob is, the gearboxes are pinned to the motors, and they vibrate loose.

In face, everything on this generation is breaking loose. I have an hour and a half runtime, and I have finally glued in the batts, glued the wires to the battery ends, etc. Wire ties kept breaking, conns kept coming loose.

The laser cut poly is actually really cheap for what you get. I give final, efficient artwork, laid-out in Autocad so lines cut sequentially, and it's really efficient. I did a set of extra whegs, hubs, some body pieces, etc for $140.00. That was polycarb too. Acrylic would have been a lot cheaper too.

With a 0.004" kerf, panels that nest together in assembly can be cut at the same time, to save cutting distance.

Ah, ok. I was just trying to think of how you might be able to accomodate different motors without having to machine new parts, but it looks like you've already been down that path.

Anyway - cool robot! Just don't let it loose around unsuspecting people - that thing might scare the crap out of someone, bounding about the way it does if they weren't expecting it :-)

Cheers,

-Brian

New movies of the CANipede!

The "whegs" were set in speed mode, so it wasn't climbing as well as it could.

Mike

working on a stabilized platform as we speak.

Mike

It climbs shallow stairs. I need to work on the control a little better. Pulling up from the front, as opposed to pushing from the rear. Downstairs, I need the opposite, the rear needs to feed the fron down the stairs.

For now, i am playing with it via R/C, and then i will add more control/sensors.

Hi Mike, what about your vibration problem? Is the springiness in the whegs possibly too great, causing it to pop up and down?

Regards Rhex, I think they timed the spoke-falls to minimize vibrations, and I think the Case-wheggie guys did similarly. I wonder if this would help for your bot?

Very cool device. Do you have a motor to lift the front end to climb obstacles? Looking at the pictures and movies it looks like all the segment joints are simple springs.

If you wanted to add the ability to lift it's front to climb higher obsticles you could run a small cable across the tops of the joints of the first 2 or 3 segments then use a servo to tighten the cable and lift the front. Wind the cable around a pulley on a servo modified for multiple turns. :) Then mount a pair of wire antennas on microswitches to detect higher obsticles and have one of your Isopods run the servo. Saves you from having to bash into something hard enough for the front whegs to get traction.

Also, I presume the surgical tubing is for traction on smooth floors and to quiet it. If you want another solution that would look a bit nicer and cover the entire leg, try some liquid rubber. This can be found at most paint and large hardware stores. It's intended for dipping tools in to give them a rubber coated grip. You could dip each leg segment (but not the hub) and have it turn while drying to keep the distribution even.

For a camera mount, have you tried a small spring with rubber bands to keep it from ocillating?

Keep up the good work. :)

G.........

Just 12 motors. I wanted to see how far I could take it without articulating the joints. If I want to go farther, I can add 2 servos per joint.

This is, for the most part, an excercise for me, it just turned out cooler than I anticipated. V0.3 is fully expected to have all the bells and whistles.

The tubing was basicly something I had laying around. I didn't want to go to the trouble of casting polyurethane tires for the beast. Vinyl plastisol part dip was another option. I know a guy that knows a guy that has a factory for part dipping, and considered that.

As it turns out, these wheg things look better suited if you have 6 motors on a rigid assembly, otherwise, the vibration will kill you.

Mike

I may be wrong, but I think what they did was to synchronize the motors so that the wheg-feet came down in the same manner as would the legs of a normal hexapod ... as in a tripod gait. You would still get some bouncing, but not so much shaking I think.

Yeah, it dawned on me that theirs was a firm platform, with essentially sitck legs. If I motorized my articulation, I could follow their model more easily.

Mike