I’d like to know how you will make the shell, because I would love to do that on some of my robot, but I have no idea how… 
Sam
Not looking to hijack this thread but vacuum forming stuff is fun and this might provide some do-it-yourself ideas for folks not already tooled up to do it (although I get the impression evolution himself might already be set-up to do it.)
Here is a pic of a small box and bracket designed to pull up to about 0.050" thick 9"x12" sheets over a plug
http://users.adelphia.net/~eballou/images/IM000632.JPG
and here is an example of a engine cowling made with the fixture
http://users.adelphia.net/~eballou/images/IM000638.JPG
The box is made from 3/4" MDF and the top is 1/4" pegboard. I pull the vaccum using a 2hp shop vac, and the kitchen oven for heat. I have a larger box that uses electronic perfboard for the top for doing larger and higher detail parts. The box is easy to make. The plugs however can be very time consuming since their finish drives the texture of the pulled surface, and when pulling clear canopies also how “clear” they will be.
If you google vacuum forming box project you can find several box design examples, and rcgroups.com has posts about this as well as finishing information for plugs.
EB
:mrgreen:
Although it may be possible for him to make the legs wider, it may not be wise for him to do so.
Most walkers use a “Zero Moment Point” gait. In short, it’s a gait that keeps the center of mass within the area above the footprint that the robot has on the ground at all times. That’s why walking robots usually have a funny “sway” to their step - they’re adjusting the center of balance so that it is fully supported during the step. Stop a walker during its step, and issues of inertia aside, it will most likely be able to stand right there on its foot.
People, on the other hand, don’t sway nearly as much. That’s because the act of walking isn’t so much a series of static steps like a robot makes, but a “controlled fall”, which is really a matter of overbalancing yourself forwards, and then successively catching yourself with the “stepping” foot, balancing on it while the other foot moves forward, and then falling forward some more, catching yourself on that foot, and so on. Stop a human in mid-stride, and he’ll fall on his face.
A number of recent robots have demonstrated a true overbalanced (human-style) gait, but doing so involves a whole lot more math, real-time computing, and sensor integration than a zero-moment gait, which can often be implemented through a simple sequencer.
In other words: The wider he makes the hips of his walker, the more it will have to sway in order to support itself during each step, and the more strain will be placed on the hip motors, due simply to having to support the robot’s body at the end of a longer lever arm.
Eddie you have a nice setup there. Vacuforming for Cowls and canopies is alot of fun. I just made my 36" X 18" Vac table last year. I started out carving hardwood plugs to vac over. now for more detailed pieces, I carve it out of foam, fiberglass the foam to make a female mold, then fill the mold with Alumilite to make a durable and rigid plug that can take the heat and release agent. Plus, carving a custom design from foam is very easy and quick.
Sam, there is alot of good guides online for DIY Vacuum tables and the process. It’s very popular among R/C Airplane hobbyist for making cowls, bombs, cockpits, Canopies and bodies for smaller helis.
where do you find the 0.05’’ plastic?
Sam
You can find all different kinds of plastic. Styrene plastic sheets are the most comon and cheapest. One of the better plastics is PETG. It is more rigid and durable.
You can pick these up at most Local Hobby Shops , especially ones that sell Model train parts since these sheets are used alot for crafts and molding minatures.
You can buy them online too.
I got my Lynx Scout from and a pile of goodies today. I’m trying out 12 X HS475’s overvolted to 7.2 Volts. And I’m using an SSC-32 and maybe the MiniAtom botboard and Atom Pro chip I got.
The package contains loads of SES brackets. I should have bought 2 just for the brackets. Everything was packed great.
Here’s some pics:
http://i138.photobucket.com/albums/q253/gfisher2002/robots/Scoutparts001.jpg
http://i138.photobucket.com/albums/q253/gfisher2002/robots/Scoutparts002.jpg
http://i138.photobucket.com/albums/q253/gfisher2002/robots/Scoutparts003.jpg
http://i138.photobucket.com/albums/q253/gfisher2002/robots/Scoutparts004.jpg
http://i138.photobucket.com/albums/q253/gfisher2002/robots/Scoutparts005.jpg
http://i138.photobucket.com/albums/q253/gfisher2002/robots/Scoutparts006.jpg
I also got this battery?? I didn’t order it so maybe it came with something?
It’s a 5 cell AA 2600mAh pack. It made my onlybatterypacks.com but it has Tenergy cells in it. In my experience, a AA battery with more than 2000mAh has to be a real good cell to actually live up to that capacity. And, in my experience, I’ve found Tenergy cells to be slightly over-rated as far as capacity and resistance. They have alot higher resistance than similiar High end brands. They are priced considerably lower though so I guess it’s to be expected.
If anyone wants it, let me know and I can send it out in a bubble envelope to you. I build custom NiMh and Lipo packs but I don’t use Tenergy cells so I have no use for it.
http://i138.photobucket.com/albums/q253/gfisher2002/robots/Scoutparts007.jpg
I’ll be running 2 X 2500mAh 7.2V lipo packs for 5000mAh. They are just slightly wider and twice as thick together, than the 5 cell AA pack. But they weight 1/2 as much and provide 5000mAh of capacity in parallel.
http://i138.photobucket.com/albums/q253/gfisher2002/robots/Scoutlipo.jpg
Well, enough chit-chat, I’m going to build my scout. I hope I can “force” those HS475’s to hold that beast up. I’ll be selling it in a couple months if anyone is looking for a scout. I only wanted it to test some gaits I thought up and familiarize myself with the SEQ software better to prepare me for programming my Giant Biped. Just drop me a PM to get dibs on it.
would you have any web adresses you could give me for the plastic?
thx, (nice pics!)
Id love to take that battery off your hands. Sadly one 7.2v lost its life when I was experimenting with chargers. 
Now I know which charger to use 
I use polystyrene mostly, anything from 0.015" up to about 0.050" (~3/64"). I might be able to pull 0.062" sheets but I’ve never tried. You can get clear sheets from local hobby shops and I bought a big sheet of 0.031" white from McMaster, which is what the actual cowls for that project wound up being made of. I made a clamping bracket to hold the sheets although I’ve read where some folks just staple the plastic to their frame. I couldn’t tell you which works better.
Making plugs is time consuming. I found something called plastic balsa at an art supply house. Nasty dust to breathe but very easy to sculpt. I use thinned epoxy to seal it and then several coats of polyurethane which are progressively sanded up to about 1600 grit wet paper. For a big canopy though you can make it 0.015" small and then pull a sacraficial sheet and leave it on the plug. that will absorb lots of pin hole type imperfections and when you pull your real canopy it will be much more clear. got to remember to use plenty of talc as a release and you can’t get too crazy with the heat gun if something doesn’t pull just right or things will stick together. 
Sure, I’d be glad to help. And I hate to see too many things just laying arund so I might as well get rid of it. PM me your address and I’ll drop it in the mail tomorrow night.
Sam, most of my sheets are bought in bulk from manufacturers. One of my better sources is: professionalplastics.com/cgi … r/3/cmid/4
They have every imaginable type of plastic. And are very helpful.
If you are just looking for a few sheets, I’ve bought alot of last minute stuff or odd sheets from these guys:
cchobbies.com/catalog/sheetsevergreen.htm
Decent prices and great service.
Or, I have used these guys in the past as well:
usplastic.com/catalog/produc … e%5Ftest=1
They have a variety of stuff.
Well I finished up the Scout. I haven’t routed the servos yet but I’ll finish up tomorrow after work. It’s alot bigger than I thought. It really has to be seen in person to appreciate it. It’s a lot cooler in real life than in the pictures. Some great design work and fabricating Jim. Everything fit like a glove. I just hope the HS475’s will power it.
I’m going to trace out the plates for the chassis and cut some from 1/16" carbon fiber sheet. It will cut the weight of the chassis by 60%. And I’ll probably ditch the BotBoard and Atom Pro unti I know it can walk with the weight of the SSC-32 alone.
Here are some pics of it standing tall, and proud , and crouching without power or support.
http://i138.photobucket.com/albums/q253/gfisher2002/robots/Scoutassemb001.jpg
http://i138.photobucket.com/albums/q253/gfisher2002/robots/Scoutassemb002.jpg
http://i138.photobucket.com/albums/q253/gfisher2002/robots/Scoutassemb003.jpg
http://i138.photobucket.com/albums/q253/gfisher2002/robots/Scoutassemb004.jpg
http://i138.photobucket.com/albums/q253/gfisher2002/robots/Scoutassemb005.jpg
http://i138.photobucket.com/albums/q253/gfisher2002/robots/Scoutassemb007.jpg
I’m still waiting on my form bending machine to start on the Giant brackets. I called them today and I guess they haven’t shipped it yet. They are backed up from Christmas holiday orders. They said it would go out tomorrow. So I guess I won’t have it till next week. I’ll have to play with my new scout till then. Darn.
Ok, I got my wiring almost finished and my lipos built. The original 5 cell 2600mAh NiMh pack weights 5.6 oz. The 2 lipo packs together weight 4.8 oz and are 7.4 Volt 6000mAh. So I shaved a hair of weight there. They are 3.7 Volt packs in series for 7.4V.
I mounted them on foam weatherstripping pieces to absorb shock. You can see them in the pictures.
So far I have only tested using the SSC-32 and SEQ to make it stand tall and crouch. It bobs up and down with seemingly no effort. the servos aren’t working to hard and after 10 cycles of crouch, hold for 10 second, stand, the servos aren’t even warm. I hope they can handle walking.
Here’s some pics.
http://i138.photobucket.com/albums/q253/gfisher2002/robots/Scoutbatteriesandwiring001.jpg
http://i138.photobucket.com/albums/q253/gfisher2002/robots/Scoutbatteriesandwiring002.jpg
http://i138.photobucket.com/albums/q253/gfisher2002/robots/Scoutbatteriesandwiring003.jpg
http://i138.photobucket.com/albums/q253/gfisher2002/robots/Scoutbatteriesandwiring004.jpg
http://i138.photobucket.com/albums/q253/gfisher2002/robots/Scoutbatteriesandwiring005.jpg
http://i138.photobucket.com/albums/q253/gfisher2002/robots/Scoutbatteriesandwiring007.jpg
http://i138.photobucket.com/albums/q253/gfisher2002/robots/Scoutbatteriesandwiring008.jpg
http://i138.photobucket.com/albums/q253/gfisher2002/robots/Scoutbatteriesandwiring009.jpg
http://i138.photobucket.com/albums/q253/gfisher2002/robots/Scoutbatteriesandwiring010.jpg
http://i138.photobucket.com/albums/q253/gfisher2002/robots/Scoutbatteriesandwiring012.jpg
http://i138.photobucket.com/albums/q253/gfisher2002/robots/Scoutbatteriesandwiring013.jpg
That’s nice where you put the batteries! good thinking
I kind of got the idea from another member here, Tillin9. He was inquiring about some packs for his biped and he had mentioned them fitting somewhere similiar on the legs. I don’t know if this is the exact spot or not, but I got the idea from him.
I’m making a walking sequence now. Funny, I kind of like the little guy. I have the urge to name him for some reason. lol
I think its cool! Nice place to put the batteries and evenly distribute the wieght on the hipps. pretty cool dude!
Thanks guys. Man, this walking thing is tougher than I thought. It’s easy to envision it in your heads but getting all the servos to do what you want it time consuming. I used the SEQ to make…well…something that resembles walking. lol I get him to crouch and lean onto the left foot, balance, then extend out the right foot (dragging the heal) while leaning more to the left to compensate for the shift in CG. Then he sets the foot down, lifts off the back foot and shifts his weight over the front foot. The lifts the back foot and brings it forward while leaning to offset the weight of lifting the back foot.
That’s as far as I got. It took me some time to set up my min/max and centers and a long time to sequence that movement. But when he does it, it looks pretty cool. I have to adjust some of the leaning when he’s compesating for lifting a foot but he can take a full step on his own. Man this is awesome. I programmed him to nod “yes” and shake “no” as well. LOL just for the fun of watching him do it. lol
This is great.
The HS475’s are doing ok I guess. Once he’s shifting weight on one foot, the knee joint doesn’t have enough power to raise him at the same speed as the other joints. It’s definitely struggling on one leg. So I wouldn’t recommend using all HS475’s on a scout. Although they handle most of the other joints very well. I’m going to pull a couple HS5995tg’s from something else and toss them into the knee joints. I might even put a pair in the ankles and hip Y joint as well. I’m just afraid it will work so much better, I’ll want all HS5995’s and I can’t afford that right now. lol
I’ll try to get a looping gait working and shoot some video of it. With just a small amount of time spent with the scout, I can already see where I will have to modify some areas of my Giant Biped design.
Nice work!
I had thought about putting packs on the hips, though I had thought using the reverse shin would be even better. Honestly, I wasn’t able to do that much testing on either idea (my work has been stalled lately), but am glad you have been having success.
I was wondering if you also tried putting the packs on the torso and did some kind of comparison? Was walking esier/ harder? I had used an ammeter to measure current consumption while walking with various gaits (determined more or less by trial and error) and tried to extrapolate what worked best. While my ballpark readings were very good, i.e. I could tell standing was a few hundred mA and walking was 1-2 Amps with a balancing gait and 2-3 Amps with a non-balancing gait, more precision was really needed for a meaningful conclusion. I ran my pack tests here and on average putting the pack on the hips incurred about 500 mA increase while a shin placement incurred about 300 mA. Both for a non-balancing gait. However, this was with one 6 oz pack on one side (measurements only using data from the channel the leg was attached to) and might have been below my deviation with the original meter. I ran into a lot of trouble getting good readings due to the fact my multimeter wasn’t really designed for dynamic circuits (sampling rate was simply too low, no integration function, etc.), and my oscilloscope didn’t have a built-in ammeter function.
To make a long story short, when I finally got a circuit which could measure a few mA to 10 Amps with good time resolution and logging I realized my servos were horribly calibrated since I hadn’t really done much calibration during my initial build. So I’m waiting on a servo programmer and some extra components to help fix this. I know my new meter should be sensitive enough to read the deviation amoung the servos but simple bench weight tests gave less deviation than the gait tests (and got better when I changed my software from symmetric to asymmetric) so I’m fairly sure there are considerable calibration issues.
I’m also not sure the 475s can do a non-balancing gait. My 5955s can hold the bot up with only one leg having power, meaning I can move the other leg more or less freely and move that way. Once I get my servos calibrated and fix the error in my movement code, I can post a copy so other people with Scouts can try the gait.
One final thing which interested me, is my bot can not really stand without power. I was wondering if you 475s might have an advantage here, or if you had found one of the few poses where the bot can stand without power. Its difficult, but every now an again if I try to position her, I can find a stance which is relatively stable. Servos with position feedback would be great for this, since I could then apply power and read out a minimal energy configuration directly. However, I know of no standard or digital servos with this ability. OpenServo and other I2C robotics servos can do this, but I don’t have any of those.
I haven’t tried other positions yet. That would be assuming I had a program for it to walk.
The 475’s can hold it up in most positions other than if the torso is leaning forward/backwards too much. I added helper springs to the knee joints and it has made a huge improvement. It can stand on one leg now and move the other freely.
I made up a single step sequence just now. Needs alot of work but I almost looks like it’s walking. Don’t forget, I’m still getting the hang of using the SEQ program and my programming abilities are that of a beginner.
Here’s a video of my stumbling little robot. (It’s only 3 megs/ 20 sec)
Go easy on the laughing. This is my first attemp at making something walk. lol
s138.photobucket.com/albums/q253 … CF0002.flv
And here are the helper springs on the knees and the scout standing without power.
http://i138.photobucket.com/albums/q253/gfisher2002/robots/kneesprings2001.jpg
http://i138.photobucket.com/albums/q253/gfisher2002/robots/kneesprings2002.jpg
http://i138.photobucket.com/albums/q253/gfisher2002/robots/kneesprings2003.jpg