Toddler's BIG Brother
[Deleted User]
Posts: 0
Hi,
·Here's a little project i'm been working on , this was going to be stamp powered but I am·leaning torwards the propeller now,does anybody know where I can find some bigger than normal servo's or have any suggestions for motors (leg's are about 2 feet long)
Thanks
·Brian
·Here's a little project i'm been working on , this was going to be stamp powered but I am·leaning torwards the propeller now,does anybody know where I can find some bigger than normal servo's or have any suggestions for motors (leg's are about 2 feet long)
Thanks
·Brian
Comments
You're going to find out that Toddler doesn't scale upwards too well. Having built over 20 of these exact robots I found that the smaller they get, the easier they turn. In larger sizes they have trouble with the slip turn, especially considering that slop amplifies over a long distance. If your joints are free of play and fairly precise it might turn on certain surfaces, but your Toddler is going to be asking you for a hip joint to rotate the leg. But you need to prove me wrong and make it walk really well!
As for the motors, check Tower Hobbies for big servos. There's also Servo City and MAXX Products - the last being the cheapest for parts you need. Don't forget to check eBay, etiher. There's some great stuff cast off by the R/C crowd, left for guys like you and me!
Ken Gracey
Thank's for the concern (this is really big brother II , big brother one did just what you said ) I am working on a ankle twisting gizmo to help with turning
Thanks
Brian
(brother 1 used wiper motors with limit switches & met with a untimely death after a few beers)
Post Edited (truckwiz) : 10/24/2006 2:45:27 AM GMT
TecFoot is 600mm (2 feet)·high. It has 5 servos - 2 for each leg to control the ankle and stride plus a fifth servo to articulate the hips during turning and for advanced biped walking. TecFoot utilises a BS2-IC (must be purchased separately) based processor board with two of Milford Instruments' µ-servo driver chips and prototyping area. I've attached a small photo of one.
see http://www.milinst.com/·for details (under 'robotics'). Instruction manual can be downloaded.
Being wood, it will be lighter, although the weight of the servos·means the centre of gravity is still high.
Once I've·built it (all plywood, wires and resin glue, like some ancient biplane!), I'll post the results.
John
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Manxstamp,
Isle of Man, British Isles
I have just seen the TecFoot post.
I created TecFoot in 2000 and occasionally the prototype and a production sample are dusted off and demonstrated somewhere such as at the Micromouse competitions. In February this year I attended a Walking With Robots workshop in Bristol, at the @Bristol Science Centre, which was attended by researchers in robotics and Ai from European Universities and I took along the TecFoots. Beforehand I decided to upgrade the TecFoots with Ping sensors to allow them to wander autonomously. They were a great success wandering about on the exhibit floor of @Bristol, the children loved them and Professor Alan Winfield of the Bristol Robotics Lab was photographed along with the TecFoots for the local newspaper. As part of the upgrade to use the Ping sensors I revised the tendon position in the foot servo horns and changed the leg servo horns to better use the full travel of the servo, both of these changes increased the available torque and performance. The software for TecFoot was completely rewritten last year and modified for the horn changes. Written in BS2basic it is a 4G list processing language which allows easy and flexible control of TecFoot.
Video of the prototype TecFoot, with underfoot contact sensors, was taken and is on YouTube. New video of the production TecFoot is yet to be taken.
The YouTube videos are on
www.davidbuckley.net/DB/YouTubeVideos.htm
and the upgrade is detailed in
www.davidbuckley.net/DB/TecFoot.htm
www.davidbuckley.net/DB/TecFootManual/TecFoot.htm
and the new software is in the above manual.
Some of the available commands are
and the autonomous wander program is
TecFoot aims for free areas and continously turns as it walks by altering the splay angle between the legs as it steps. Should the way be blocked then one of the behavioural routines is activated.
The behavioural routines are
Writing the program in a 4G way is much easier, clearer and quicker than writing more BS2basic. Not to mention you can do more with less program.
David Buckley[url=http://][/url]
What a complete presentation of the building of a robot, both for beginners and experienced folks alike. The picture detail makes everything so easy to figure out how and where all the pieces fit together. Job well done!
One quick question though. In the message above you mention "... the new software is in the above manual". If that was intended to imply a "portable document" (PDF, word processing file, or text file) that link is pointing to the wrong place as far as I can see. Here is the link you have indicated which points to HTML pages: www.davidbuckley.net/DB/TecFootManual/TecFoot.htm .
If that was intended to be just the HTML text, no problem. It just would have been easier if the manual was in a portable format. It then becomes accessible from more than one computer, and many of use both a desktop and laptop when using the Parallax PBASIC Editor and other files to support our "habit"
One other quick question. Is there a complete listing of all the Tec-Foot language commands. I didn't notice anything, but that command listing may be in the manual.
Thanks for all your hard work. I'm sure many of the folks here will appreciate it, both now and in the time ahead. New members seem to come on here every day.
Regards,
Bruce Bates
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Programming can't be all that difficult, it's nothing but 1's and 0's
Thank you for the compliment.
The link to the TecFoot program is a link in the manual TecFoot.htm - TECFOOT.BS2
www.davidbuckley.net/DB/TecFootManual/TECFOOT.BS2
along with links to the old manual and thence the original program TECFOOT1.BS2.
The commands are listed in the program.
The original manual was supplied on disk with TecFoot and it is not so long ago I thought 'Oh what the hell' and put the TecFoot documents on the web. I could have zipped them but anybody interested should know how to save them locally. Just do 'File | Save as - web page complete' for IE6 (7?) or something similar for Firefox (whichever add-on you choose, I like 'Title Save'). If I put the program in a pdf you would have to get it out to use it. As it is IE6 will download it to the Editor if installed or just save it locally, Firefox will open it in a new tab, just right click and 'save target as'. All the links are local so will work after you save the manual and program etc.
If you are interestd in the program you may like to read the Hextor Manual - to big for here 4MB, I'll put the zip on my site.
The 4G list processing started off in BigFoot, very simple, was clawed back to conventional 3G in Toddler, and developed for Hextor and occupies most pages of the BS2sx along with the BOS compiled into an PIC16F876.
Loki
www.davidbuckley.net/DB/Loki.htm
has Poses, Acts, Dos (plural of Do), Behavioural Dos, Behavioural commands, selectable default Dos, underfoot sensors and four servos and uses just, and only, a standard BS2.
For the Amblers I have implemented the lists as all text, including parameters.
{BigFoot, TecFoot -> Toddler -> Ambler}
www.davidbuckley.net/DB/Ambler.htm
but the code is not on the web.
I am in two minds as to whether port it onto Penguin since my compass doesn't work. But it would turn Penguin from a very nice executive toy into a good tool to explore the problem of building robot minds.
That actually is what the programs are about - robot minds - and anybody trying to build one in a 3G language, ParallaxBasic, C, Java etc is doomed to failure. Minds don't have the single entry points and single exit points and subroutines galore of 3G languages. As far as I can gather the Japanese are still compiling the list of moves for their RoboOne robots into conventional 3G languages (Basic and probably C). However, at the present rate of development it may not be long before one of them makes the jump to 4G and creates a 4G GUI environment. Once that is done the integration of many sensors will be easy, sensor fusion will no longer be a hard problem, and robot minds will become common place, providing you can understand Japanese.
At the moment while it is possible to write a flexible Mind in a a paged BSx they don't really have enough pins. A Propeller, or a flock (er Wing?) of Propellers, is probably the way to go and if someone would write some string handling for it, that could be a really good start.
This post didn't stop where I thought it might! I would be interested in further discussion.
David
My how you get going Just in passing a little more white space would make your posts more readable.
Thanks for all that information. I will try what you suggest about attempting to download the web page with the save as web page, if I can find it. As far as the PDF is concerned, I NEVER print mine. I always leave them in an area of my hard drive and open them from there. Why kills trees when it isn't necessary
The Basic Stamp BS2p40 has 32 pin ports (with the two serial ports for DEBUG and programming) which can be utilized in two sets of 16 pin ports each. The way they do this is with the commands MAINIO, and AUXIO, and IOTERM which can be found in the PBASIC Help File, or the PBASIC Reference Manual. The Stamp BS2p40 also has banks of memory (aka "paged").
I would LOVE to see string handling in the PBASIC language, but I'm afraid it's not going to come about easily. Some of us have asked before, and the answer wasn't promising. Perhaps in the next incantation of Stamps, the Stamp III or something like that. I don't use the Propeller, so I can't speak about whether it can or can't do proper string handling. Even a command set like list processing languages, similar to SNOBOL, LISP, or any of the others, would be GREAT!
I'll check out the HEXTOR manual as soon as I have a chance. Thanks for pointing it out.
I can't help you with the Penguin - sorry, I never got around to purchasing one. It's on my list though.
Regards,
Bruce Bates
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Programming can't be all that difficult, it's nothing but 1's and 0's
Whatiswhitespace?!
I think I print about one page a year, costs me a fortune in ink cartridges.
I do have a P40, but I think it is dear. All the previous robots were developed with a view to Milford Instruments selling them. Having a P40 on-board wasn't an option.
The Propeller doesn't have string or character variables as such.
Just view the TecFoot Manual then'save as web page'
www.davidbuckley.net/DB/TecFootManual/TecFoot.htm
In its index right click on TECFOOT.BS2 and 'save link as'
Put it in the same folder as TecFoot.htm and TecFoot_files.
Lisp etc - to many brackets - and I have never seen an implementation which would fit on a robot the size of even TecFoot and consume<100mA. Maybe there is something for a GumStik.
Most of these languages require their data to be held elsewhere in files and so are not suited to micrcontrollers.
Data on Flash cards is fine if you are gathering data and going to read it out into some display program. Not good if you want to start with the data in viewable form and then put it on a flash card, needs a custom GUI, which is what the Stamp/Propeller Editor is, OK a TextUI.
The BasicX has strings but the editor doesn't seem to have name substitution for data.
"Help us ChipGracey, you're our only hope"
David
Just added the HextorCD.zip to my Hextor hexapod page
www.davidbuckley.net/DB/Hextor.htm
www.davidbuckley.net/DB/HextorCD.zip
David