Belladonna Page 2

June 6, 2002
	Ok, just a little side note. Starting with this page I'm gonna try and date things like above so that it will be easier to see how long things take. I make no promises as to updating frequency however.... Anyway, on with the story of the cams. First I had assumed I would make them out of UHMW. I had also assumed that they would also function as the spacers between the legs to keep them from rubbing. Well to begin with the UHMW rod wouldn't cut cleanly on the lathe. I tried different speeds, and making sure the bit was at center height. Nothing. It wasn't "all that bad" so I tried to drill the offset hole to make it a cam. Ouch. Or more correctly Oomph, the sound I make when having the pound and a half steel collet block chuck flung into my chest by the drill press. Apparently even after drilling through first with an 1/8" bit then a 1/4", the 3/8" bit would just bite in and spiral through, whipping the piece and the chuck I was holding on to out of my hands. I would have put it in the vice, but the vice does not hold the block square. When I finally found the correct size step bit to use (another day long story) the holes would not come out straight. I guess the OD is not all that accurate on that rod, and hence the errors. So I decided to make em out of Al-6061.
	Here we go with the Al. This is the 1/2" thick blank. Due to the above fiasco I was furious at the several days I wasted on the UHMW. I needed some 1" round 6061 and I was impatient to get it. I would have got it from online metals but it takes like 5-7 days and I wanted the cams yesterday. McMaster delivered the rod in 2 days but I had to order one with a "hard anodized" coating. They said it was at max .002" thick. How bad could it be I thought? Kinda bad, watch out for this stuff kids. Dulls lathe bits like crazy. But I got these blanks made up despite all the problems.
	Here is the above with the slot milled for the spring pin to sit in and the off-center hole bored for the shaft. I used a ball mill to get a really tight fit on the spring pin. Like I said I had a hell of a time finding the step bit with the 3/8" step. For some reason I was convinced that someone locally would have it (they have all the 1/8" step models.) This was wrong. I had McMaster throw it in with the rod. Mmmmmm Uni-bit.
	Ahhh the completed cam. Now I just have to make the rest....uggggh more cutting through the dreaded coating. As it turns out all of the coating is removed in creating this piece but for various machine limitations, it needs to remain on until the final turning.
	Here we see the mocked up position of the cam. Underneath this assembly is the bearing retainer plate. Another one will be placed on top of this but it has been removed for clarity. I bored the hole with an 1-1/4" forstner bit. Pretty accurate in the lexan. Everything fits well with just a little play. Surrounding it are the 12 ball bearings that bear the load and reduce friction. Had I gotten really crazy I could have even calculated the proper radii to make the balls fit perfectly with no gap, but close enough is just fine for me in this case.
	Just for kicks I decided to throw a bunch of the HBF Sub C's in to see how many I could cram in if I need/have the weight to. That's 32 of them in there with plenty of room (how do I know this remark is going to haunt me later) to fit the radio gear in. I don't plan on using this many but it's good to know I can get em in there if I have to. I took all these apart the other night as I figured they would be good for something, at least to use as test packs.
June 7, 2002
	I got all the cams done today. Didn't think I was gonna get em all done but there they are. I also rough cut the shafts, this will not be the final length. It'll be easier to pin them with a little extra length on them. I also finally tapped the center bar and put it in. The first broken tap of this project. I was doing so good too. At least it happened in the easiest piece to make, so I'm not too pissed. Tomorrow I pin the shafts and get a start on the feet themselves.
June 8, 2002
	Ahhh, right back to step one. But of course I should have remembered that I wind up making half the parts over again anyway. It seems that the bearing retainers I made in step one were not at all accurate. I understand what I did, it was some layout technique error. I knew it was wrong when I did it, and I did it anyway. So I got to make 24 more, only this time the correct way. This time I used the sticker method again. Here is the layout.
	Here are half of em, the half that need to be countersunk in fact. The other half are tapped 4-40, Yay! no broken taps again. While I was redoing them all anyway I figured that I'd move the holes out a little bit (they were kinda close to the bearing hole, like 1/32-1/64). These will work much better.
June 9, 2002
	Not bot related but cool nonetheless. This is a dragon sculpture I made for my parents a couple years ago for Christmas. It's made from flexible copper pipe, 4 and 8 gauge solid copper wire, and a load of copper foil. I had today to put the base on and reattach the head (a cold solder joint failed) Though any readers would enjoy the pic.
	After fixing the dragon I turned my attention to the feet. Here are two. I had cut out three, but there were some inaccuracies in the bearing hole drilling in the third one. It's a little tricky to do, I used a forstner bit which is made for wood but works in the plastic. You have to go real slow, and keep backing out the bit. It gets really hot, and tends to melt a thin layer back onto the inside of the hole, thus necessitating the need for long painful scraping with an x-acto blade and some sandpaper. You can just make out the remainder of the 3/8" shaft that I stuck a pin in and chucked in the drill. I used this to "run in" the bearings. Pretty smooth now, might need a bit more work later.
	A totally gratuitous shot, mainly to illustrate how the invertability works. The feet simply stick out of both the top and the bottom of the bot. The top and bottom will be covered with 1/8" lexan both as armor and to keep the electronics in. The shafts still need to be pinned and I need to make the other 4 feet.
June 13, 2002
	One thing I can say to anyone thinking of attempting a design similar to this one; think twice. This is one of the most frustrating things I have ever attempted. This shot is of a bunch of shims I had cut out of transparency material. Basically the ball size is the same as the thickness of the lexan feet. When you put on the retaining plates they squeeze the balls too tightly, and don't allow them to move freely as they should. I printed the design onto sheets of transparency material and cut em out with an x-acto knife. Now the bearings spin very nicely.
	At long last all six feet completed. I had originally planned to cut them using a router trim bit and a machined template. This did not work, the bit kept kicking back and gouging, it was a huge mess. Instead I used a painstaking process that goes something like this. Cut out to rough size with a jigsaw. Rough sand to 1/16" with a drum sander, then use progressively finer grits in a palm sander to get the edges smooth and straight. Overall I am quite satisfied with the results. Just for fun here are some numbers.....6 feet, 12 cams, 24 bearing retainers, 60 shims, 48 4-40 flat head socket cap screws, and 144 1/4" bearing balls. Everything except for the bearing balls and cap screw was all custom made. That was a lot of work
	At last, assembly of the legs. I didn't get a shot of the drilling to pin the shaft, but I will tomorrow when I complete the other side of the drivetrain. This is where the frustration really came into play. Basically it goes like this. Take the pinned shafts put a pin in the innermost hole of each one, then put the legs onto the shaft with the slot in the cam facing the pins. A little tapping with the hammer to seat the spring pins in the slot. Then cut a very narrow (a little less than 1/8") section of 1/2" OD Al tube, and put it over the leg to serve as a spacer. A little tapping to seat that and then another set of pins. Repeat two more times, with much screaming and cursing as the pieces did not want to listen. They eventually did as you can see in the shot. I also cut the chain and installed it. Dunno if I'm even gonna need tensioners as the chain fits pretty close without them.
	Now for the moment of truth, I grabbed the spare battery I had for Morningstar and a couple of wires, and put the whole thing on the living room floor. YES!!!!!!! IT WORKS!!!! Words cannot express my jubilation at the fact that this thing actually works. It's not super speedy but it moves pretty well for a walker. I was also only testing on 6 cells. I'll probably run with 7 or 8 in the end depending on how I decide to control this thing.
June 14, 2002.....IT WALKS
	This is the setup I used to cross drill the shaft "accurately". To have three feet on a side, each one should have a 120 degree offset from the last. How do I do this? With this collet block chuck. It uses a huge (if you are used to the dremel collets this is huge) 5C collet. You put the work in the collet and tighten it down with a nut on the back. The fence that I clamped down keeps the bit in the center of the shaft. You simply set up the block, then drill the first hole. Since it's a perfect hexagon you rotate two sides (each side of a hexagon is 60 degrees) and presto perfect 120 degree spacing.
	These are the two shafts with the first set of pins installed. The second side of the drive went together with much less cursing so I was able to calm down and take some shots as it went together.
	The first leg is installed, followed by the spacer (you can barely make it out under the pin. Then another pair of pins is installed. Then you just repeat the process for the other two legs. A final set of spacers is then added to the ends to fill the gap between the last leg and the frame. This prevents the bushings from coming out.
	The completed chain and sprocket. I got really really lucky and didn't really have to futz with the chain a whole lot. It seems that I chose the almost perfect spacing for the sprockets so that it was reasonably tight without a tensioner. You can see the coupler on the right. I know that before I said I would pin the coupler to both shafts but this turns out to be really impractical. Besides which there is not too much torque transmitted there. This sounds counterintuitive but it's true. Last night I had run one side without any fastening whatsoever. It's because the legs themselves transmit torque between the shafts as do the chain and the flats on the shafts. I added these setscrews more as insurance against sideways motion. This also avoids the nightmare of having to line all this stuff up as I try to install the pins in the tiny half inch wide space.
	Time for some ghetto wiring. This is a "speed control" Very simple, cheap and fast. They will not be the final control option for this bot. All it is is a Radio Shack 20 amp automotive DPDT momentary switch and a Hitec HS-303. I made some quick lexan brackets to hold the switch and servo in the proper relation to each other, and slapped on this big servo horn which happened to be perfectly suited for this job. I didn't even have to hack at it with an x-acto! Best part of these things is I didn't have to measure or mill anything. I bent the lexan with a propane torch. A little simple wiring and I was set to go.
	Here it is in all it's glory. You can see my ghetto speed controls are way too big to fit inside the frame but they live happily on top for the time being. This thing makes a huge amount of noise stomping around. If you have some annoying downstairs neighbors give me a yell :) It's not super fast but it does move pretty well. I think that ESC's will actually make it move faster. Reason being that moving the legs slower may actually work better and stop the jumping around that goes on when you move the legs so fast.
	A final gratuitous shot of the day's efforts. You can also see the treads on the feet in this shot. They are vacuum cleaner belts that I split in half. I am very pleased with the way this has turned out. I made my own deadline of getting it to walk by the end of this week. There is much more work to do to complete this bot, attaching the armor, and building the weapon. Of course I need to finalize the electronics as well, it would be hard to make the bot invertable with those huge switches on the top.
June 23, 2002
	Didn't get all that much done this week after last weeks building frenzy. The good news is that almost all the parts are on the way for the weapon (I still need one more McMaster order), as well as the real electronics. In this shot you can see the bottom (or it could be the top as it's invertable) lexan plate bolted on. Just like in cheap office furniture the thin plate fastened onto all the frame members will add a whole lot of strength to the frame. It also provides a convenient retainer for the batteries and electrics. I trimmed it to just the right size with a router and ball bearing trim bit.
	Still not yet at the final electronics set up but getting closer. The servos will be gone in favor of a couple of RCE's from Team Delta more on them when they arrive. I bolted on the top plate on and if you look really close you can see there are a couple of miss-drilled holes. Since I have to make it again anyway I ripped off the covers so I can see what's inside and plan any of the holes I'll need for the safety interlocks and possibly a fan or two if I have room (not likely at this point). I also picked up some cool purple 12 gauge wire from the local auto-sound store. $.60 a foot for two conductors, not too shabby. Best thing, it's massive and it fits in with the color scheme. If I get really crazy I may try and find some of that 12v glow wire and put that in there. I threw in the Team Nar logo as well, hiding underneath is the Rx, the soon to be reworked regulator board and the power distribution strip.