Hi all,
I'd like to share a video I just took of my full size Lost in Space B9 robot that's controlled by two EZ-B controller boards. Right now they are controlling limited movement and voice response of a few motors, lights and sound files played from a Sparkfun MP3 Trigger board. Although I'm just starting with the animation and have more building on the actual robot the result (mostly thanks to the EZ Robot controller board) is shocking. Please have a look at this (4 minute) You Tube vid and enjoy.
Please excuse some Technical camera lighting and sound issues. This is the first time I'd made and posted a vid online.
EDIT 8/2/13: Just realized I have no good pictures of how my B9 will look when complete. Here's one of the actual TV robots from the 60's TZ show Lost in Space and one recent shot of where I'm at with my build over 1 1/2 year after I started. Enjoy:
Thanks,
Dave Schulpius
Hi all,
I'd like to share a video I just took of my full size Lost in Space B9 robot that's controlled by two EZ-B controller boards. Right now they are controlling limited movement and voice response of a few motors, lights and sound files played from a Sparkfun MP3 Trigger board. Although I'm just starting with the animation and have more building on the actual robot the result (mostly thanks to the EZ Robot controller board) is shocking. Please have a look at this (4 minute) You Tube vid and enjoy.
Please excuse some Technical camera lighting and sound issues. This is the first time I'd made and posted a vid online.
EDIT 8/2/13: Just realized I have no good pictures of how my B9 will look when complete. Here's one of the actual TV robots from the 60's TZ show Lost in Space and one recent shot of where I'm at with my build over 1 1/2 year after I started. Enjoy:
Thanks,
Dave Schulpius
To answer Josh's is question; No he will not be Mobile. No real room in my home for all that. I don't want him scampering around the house. I can always add that later if I want. Right now I want to just complete him so he can stand there and look really cool. Also he is built to exact scale of the original TV robot. The mold that the fiberglass torso was poured into and the rubber arms and legs were pulled from the original TV robot. very, very cool stuff. all other measurements including his height were taken right from the original also. I'm shooting for as close to exact as possible. He is very much man size as the original robot was just a costume with a person inside of it.
Again thanks for the comments that all the help,
Dave Schulpius
Just curious why (2) EZ-B boards?
Did you use up all the ports?
Or is it the upper/lower thing?
Thanks again,
Davis Schulpius
I have a suggestion to eliminate him getting confused by overhearing talk not intended for him. I have been looking at Bluetooth headsets made for CB radios. They have a "push to talk" button. I will probably go that route with mine.
Keep up the good work. Maybe some day we can both bring our B9s to a convention. We can start an EZ-B division of the B9 Club.
That is a Fine job on my Favorite robot in the world!
good work!
Danger, I understand and thank you for the suggestion. It seems like a solution I may end up going with. However I wanted to keep him "stand alone". Meaning that anyone in the room could "chat" with him, including myself, by just speaking to him normally.
Again, thanks for the interest,
Dave Schulpius
Should take to a show,a guy from this forum THOMAS my friend has a life size robot just like yours and has a photo of it and with him and real creator of it BOB may
and he goes to the show alot,i forgot the name of the show,but its where all life size robots show what they can do.
i got that same joke about the robot i guess you got it off the robot movie with robin willams playing in it
my long term project is JOHHNY FIVE from short circuit.
the site that is working on the plans is input-inc.com
this is THOMAS SITE
http://www.robots-and-androids.com/
Anyway,,,
I was showing off for a few friends over the holidays. Here's B9 telling us who he is and the laws he lives under. Right now he's working off the main EZ-B in the torso. You can see the lower EZ-B pulsing in his lower section. It's not active for this vid. Enjoy:
Now I'm working on the radar and then up into the bubble where a 3rd EZ-B will control everything up there. One thing I'm hoping to pull off is a tracking camera up there that will tell ARC someone is walking past and his body and radar unit will turn along with that person. Now that will be spooky.
Thanks, Dave Schulpius
Very cool vid of "under the hood" btw! and your comment about "aging eyes" made me snicker as I think I am the president of the "aging eyes" club round here!(forum)
Thanks for sharing! I hope to do the same when my project gets going as well! |(steampunk head/torso)FRED...how THAT for retro!....Wild wild West...the original BTW..I am sure you watched that one!
Hey, that steampunk thing sounds fantastic. WWWwas a cool show, both the original series and the movie.
I design and made high voltage testers,safety is my biggest item,each item has to be fully tested before it leaves the plant.
Also very high current testers up 1000 amps i design and made
Back to DAVE'S project cant wait to see it finished is very cool,i lke B9 and R2-D2 but JOHHNY FIVE is my favorate
DAVE to need when its done to place in a show or if another lost in space movie comming out
like i hope to do with mine,but so many projects to build dont then i will start on him.
BUT good news after my last trip for work ,going to retire very early to work on robots full time.
We should bug Josh to make an oppinion poll on this forum to post various things...what do you two think?
I have never worked with "hot sticks" before and I can well imagine the difficulty! My challenge was working with 26 gauge "colored" tel wires in the near darkness and freezing wind conditions with bare hands......BUT when its sunny and warm it was the greatest!
@robotmaker...so glad to hear your dad is doing well! Spend as much quality time as you can and record his life experiances for future family!
Years ago the mortality rate among linemen was 50%. That's not to mention the damage done to the linemen's body from the stress of the job from climbing, pushing and pulling in odd angels, lifting and squeezing. These damages don't really show up for a while though. I've had a number of co-workers die on the line and a few others hurt to the point they are really messes up. That's hard to take and can cause nightmares. Thankfully though we haven't seen a serious injury in many years. I admire you and your dad for your service on the line. Thank God your dad didn't die in that accident. I think I actually heard about that incident up in Canada in one of our safety meetings.
Anyway, enough of that! This is a Robot forum. With all the technology and robots coming soon maybe there will be a Robot Lineman that will keep people away from the dangers of line work. I hope not because it's a cool job that gives a lot of satisfaction. At the very least I thought that someday the lineman would have insolated robotic arms that could be worked from a floating or isolated attached platform. Somewhat like we do now with Hot Sticks. However, I truly think that as advanced as robots will get someday you cant replace the dexterity and ability to touch and feel with your own hand. I only use Hot Sticks in certain situations. Working with my own hands and the proper safety procedures is the best way. As cool as robots are I hope they stay out of certain areas of our life.
Robotmaker, Glad your dad is still around. I hope your spending as much time with him as you can and showing and telling him you love him. I wish mine were still around.
Anyway, The area I need to hide the boards in up in the bubble is kinda limited. This area is called the Brain and is a tangle shape and mounted inside the clear bubble. The bubble will be moving up and down by about 4" and back and forth by 90 degrees each way. The less wires running through the support pipe the better. All I really want to run up there are two wires supplying 12v DC.
Inside the triangle brain I'll need to pack in the EZ-B, a light controller board, a LED light board, a relay, a home made transistor switch hoard, wiring and hopefully a camera. Here's a pic with just a couple of the items:
It's going to be tight and I don't want anything sticking up too high.
I love all the other fantasy ships like the Flying Sub from Voyage to the Bottom of the Sea!
You didnt add the link so here it is:
Click To Watch Video
Here's the whole site. Enjoy!:
www.scifiairshow.com/index.html
and to hold more then 2 people that the harrier jump jet does
Lifter sections.
The main challenge was to build this so everything can be taken apart easily, fit properly together to let the Radar swing back and forth on a Lazy Susan and let the Bubble Lifter post both slid up and down freely and swing back and forth with the horizontally fixed Radar. The Bubble Lifter is lifted from below with a HD Servo. The servo is attached to a a special Wishbone style bracket and a type of thrust bushing wrapped around the post I built. This allows the post to be lifted and also to spin. The main trick was to get everything lined up to slid and spin freely and not bind up. Again, the most difficult part was getting everything to fit in the area I had and get it all lined up. I also wanted badly to have limit switches on each side of the Radar swing just in case of an out of control spin. *eek*
Then the next thing to do was to get it all working with EZ-B and write EZ Scripts so the radar will move at different speeds and stop at any point on a 90 degree swing and then return to center. I've already done this with my waist motors so it wasn't to bad. Also I needed to have the script move the Bubble Lifter as the Radar swings. In the end it all worked great and looks very cool.
I've got it all taken apart now and am in the process of painting the Radar section. I'll post a video of all the action when painting is complete and it's reassembled. In the mean time please see the pics below if interested.
What is not in the pictures yet are the Sonar Dishes that will be sticking out of the ear posts on the sides of the radar. They will be powered with Micro DC motors that will also be turned on and off by EZ-B and animation scripts!
Enjoy!:
Here's the kit I bought that will soon be the Radar Section
Here's the Radar kit all glued together, Very nice
Here's the finished kit sitting on top of the Collar and ready to start mounting and wiring
A test fit of everything in it's soon to be final resting place
He looks like a sad puppy with his ears dropping
The Collar section interior wiring showing one limit switch laying flat by the motor and the other mounted with the leaver sticking into the radar section above. No room for all the gears pipes, motors to lay both flat!
Here's a pic of the bubble Lifter servo with the Wishbone and sliding collar brackets I built using a band saw, Drill Press and a file. I do not have the rings attached to the lifting post yet that the sliding collar pushes on to lift the post
A side shot of the Radar Section with the vertically mounted limit switch and how it works. This method also works as a Stop Block to keep the radar from spinning if all else fails. There is also a stop block on the Horizontally mounted Limit Switch down below in the Collar
Here's a pic of the motor and feedback pot mounted. The Pot will tell EZ-B where the Radar is so I can have custom movement and stop points for rotation anamation
The underside of the finished Radar Section with the rotation gear and 9" Lazy Susan attached
Thanks for looking!
Dave Schulpius
I see in photos you got a burns multiturn pot,
I use to have so many many of them,but i been selling on ebay and they went very fast,only keep a few for my self
Sealing up the working parts to keep out the paint and dust
Taped up and ready for the sand paper
All sanded and ready for paint. The ears are also getting a coat of Dark Gray after I have applied a Self Etching Primer
First coat of Dark Gray. Looks good but I'll need to sand a little and repaint the radar. Like a dumb A** I stuck my finger on it when it was wet.
*tired*
More to come. Thanks for lookin!
I like these pots. They are multi turn and only cost a few $ at the RobotShop. I think I only paid about $7 USD. I haven't compared but I think they may work just as well as the ones that cost about $40 USD.
multitun are high precision but not continuous operation ,witch meens it has a stop
But for head rotate and servo feedback they are perfect
B9 will not be scampering around the house or yard so he can stand by a wall and get his power from there. If I ever want to I can convert him over to car batteries and make him mobile with wheelchair motors. That's a project for another year or then next owner when I pass on to the big robot building shop in the sky.
here is the link to his b9 with THOMAS and BILL MAY
THOMAS real B9 FULL SIZE ROBOT
are you taking your to a show too,or just home use
I am looking sometinme soon to take my johnny five full size there,it will have everything like the movie
SO FAR I BOUGHT MANY PARTS ,TRACKS WAS A BIG COST $500 EACH,just waiting on the plans to be done
Sanded and ready to put the 1st layer on
1st layer complete
Second layer ready to paint
1st and 2nd layer complete and ready for taping of 3rd layer
Tapping the disks around the disks and columns to get the dark gray to lay nicely into the groves like I want. I thought this was going to be a real pain and have dreaded this for months. However it really wasn't too bad. I just took my time and worked myself around the disks.
After 4 hours of tapping!
The last layer and step in the painting process. I'm holding my breath for the next 1/2 hour till the paint sets a little before I remove the tape. Then I'll see how the paint laid into the groves. I'm also afraid the paint will tare or pull as I remove the tape.
Well, all my fears were for not. It really turned out nicely and just as I wanted it to. You can see the dark gray laying deep into the groves and giving a sense of separation and depth. I only had one place where the paint started to tare as I removed the tape. I was excited and going too fast. I just slowed down a little and the rest came off just fine. I'll be able to fix that easily enough.
Nice Lines if I do say do myself!
Since I've take these pics I've gotten everything mounted on the robot and working. I have the Radar swinging back and forth and the Bubble Lifter moving up and down through it. After writing a few basic EZ scripts in ARC they look real good working together. I'll take a vid and post it soon. But first I want to get the ear sensor dishes installed on top of the ears and working.
Thanks for looking!
Dave Schulpius
Keep up the good work on it.
When finished I'll have 3 ez-b's. One the leg section, One in the torso and one up in the bubble. I could get away barely with one but the way he turns, leans and bobs on three levels I thought it would be less wires using three.
I'm having some problems right now with the large Bubble Lifter Servo. I'm getting a nasty jump and it locks on for a few seconds when I power it up. It has a different 12vdc power supply along with the 5v and signal from EZ-B's Digital port. When I apply main 12v power to it at robot power up or when a script starts the servo will jump almost all the way to the end of it's rotation very quickly and lock on for a few seconds. The only way I've been able to quiet it down was to run the main 12vdc power through a transistor switch circuit and have EZ-B turn it on AFTER I send an enable signal in a script. I also have to send servo Release and sleep Commands. Even after all this I still get a quick little bump. I guess I can live with that but I'd like to find a way to get rid if it completely.
Anyone have any ideas? Here is a sample script that I wrote that takes away most of the jump and lock up:
Edit : Please see the next couple posts for the solution and new script that seems to fix this problem.
\>>>>>>>>>>>>>>>>>>>>>>>>>
Take the screws out of the servo arms, power up the system, pop off the servo arms and put them on again as close to center as possible, put the screws back in again.
The servos will always correct themselves to what they consider their default position. No matter how much you move them when there is no battery in. unfortunately the only way to fix this is to (while the batery is plugged in) unscrew the screw and take of the servo horn and install it to correct trim.
<<<<<<<<<<<<<<<<<<<<<<<<<<
Makes sense and sounds like good advice. I'll give it a try tomorrow if I have time.
Anyway, I was able to get my servo jumping problem worked out. What worked best was a combination of the points I mentioned in the above posts:
First : making sure the servo was centered (poison 50) when it was at rest and attaching the horn so it was also positioned at rest.
Second : making sure I powered up the servo after the signal was sent for servo speed and position from EZ-B. I found it was very important to have the script "Sleep()" at least one second after I send the servo speed and position and before I sent the power up command that closed the transistor to let the 12vdc power flow to the servo.
Here's the script I wrote that I found works best to move the servo smoothly up and then back down:
Code:
Now I need to make a video of everything working to post and show off!
DAVE cant wait to see your video,your design is really great,hope soon to start on my B9 robot,but so many other projects first,but i am getting there,i be glad in about a month,more time then i ever will need to work on robots full time
The script looks good will have to try your idea on my designs ,thanks
Actually snow truly sucks. I've lived in it most of my life. It's cold, wet and I have no time to shovel it and keep the sidewalks clear of ice. In fact my mother-in-law just slipped on it last week and broke her hip.
Mentioning time, if you plan to do a full sized B9 you had better put aside about 1 -2 years of your life with no other robot projects. Also as others have said; a B9 robot project is never really done.
And i love my job free travel and work on designs all day,but i need to retire early and still have more money i need at the same time,i keep me from building robots a major passion for me.
I can see snow is not fun when you leave there day after day,mm idea maybe a have a robot do the snow plowing
SORRY to hear about your mother in-law hope she does better,i know how it feels if happen to me while back
Thanks for looking,
Dave Schulpius
I have my servo and horn centered to position 50 in ARC
ARC starts and the servo Panel defaults to 1 when I have the Min/Max set to 1 and 100.
I send my first command for the servo to move to position 40
The servo will try to move to position 1 and then move to position 40.
The real life result is that the servo will slam to the top of it's rotation (position 1), hold for a second and then move down to position 40. Not good.
Here's my work around. I set the Min in the servo Control panel to 50. That's my center and resting position of my servo and horn. This sets the starting default position to 50 in the window of the servo Control Panel at ARC's startup. This takes away any larch or jump when the servo tries to slam up to the 1 position. In short, the min position needs to be set to my center resting position of my servo.
I don't know if I'm missing something here with this servo control. Any words of wisdom from anyone?*confused* I don't know why it should default to position 1 at start up. I would think it would go to center. In my case that would be 50 if min/max was set to 1 and 100. I don't really use it anyway other then to see on screen the position the servo is in. My servo is always controlled by scripts. All seems well now so I'm happy.
To understand your servo situation... Is this what happens?
Action
1) Connect to EZ-B
2) EZ-Script command Servo(d0, 50)
Result
- servo moves to position 1, then position 50
This is probably old news to you but I have managed to get recognition errors down to a minimum by deliberately scripted by project with limited choices when it comes to voice commands while at the same time opening up a whole lot more....let me explain.
The new voice command display in a recent update gave me the idea of using a menu system like our local Telco voice answering phone menu ...
I have the absolute minimum of entries in the speech recognition panel and instead use the WaitforSpeech command. This lets me listens for specific responses and navigate through a series of menus responses rather than all responses I used to have in the speech recognition panel. This has been greatly improving accuracy.
Here's an adaption of one of DJSures examples...
Saywait("System Configuration" )
ControlCommand("Speech Recognition", PhrasesHide)
:Response_Menu
# Do this loop three times
$Name_Response_Loops=$Name_Response_Loops+1
$Response = WaitForSpeech(10, "Update Location", "Power down", "System standby", "Main Menu" )
IF ($Response = "timeout" )
$Response="nothing"
ENDIF
IF ($Name_Response_Loops=3)
goto (No_Response)
ENDIF
# Take necessary action according to response
IF ($Response = "Update Location" )
goto(Option 1)
ELSEIF ($Response = "Power down" )
goto(Option 2)
ELSEIF ($Response = "System standby" )
goto(Option 3)
ELSEIF ($Response = "Main Menu" )
goto(Main Menu)
ELSEIF ($Response = "timeout" or $Response = "nothing" )
goto (Response_Menu)
ElseIF ($Response = "cancelled" )
goto(Cancelled)
ENDIF
Halt()
:Cancelled
saywait("Aborting request" )
Halt()
:No_Response
saywait("I didn't hear you say anything" )
Halt()
:Option 1
Say ("Up-date Location command sequence" )
ControlCommand( "Script Manager", ScriptStart, "Menu - Update" )
halt()
:Option 2
Say ("Starting Power down sequence" )
ControlCommand( "Script Manager", ScriptStart, "Menu - Powerdown" )
halt()
:Option 3
Say ("Entering System stand-by" )
ControlCommand( "Script Manager", ScriptStart, "Menu - Standby" )
goto (Response_Menu)
halt()
:Main Menu
ControlCommand( "Script Manager", ScriptStart, "Menu - Main" )
halt()
To avoid false recognition on key commands I also have a similar script that listens for a confirmation.
$Response = WaitForSpeech(10, "Yes", "No", "Affirmative", "Negative" )
IF ($Response = "timeout" )
$Response="nothing"
ENDIF
IF ($Name_Response_Loops=3)
goto (No_Response)
ENDIF
# Take necessary action according to response
IF ($Response = "Yes" or $Response="Affirmative" )
goto(Positive)
ELSEIF ($Response = "No" or $Response = "Negative" )
goto(Negative)
ELSEIF ($Response = "timeout" or $Response = "nothing" )
goto (Response_Menu)
ElseIF ($Response = "cancelled" )
goto(Abort)
ENDIF
Halt()
Hope this helps....
You are correct about the servo movement. If I have the min max movement set to 1 and 100 the servo control window will show 1 at ARC startup. If I have a command to move that servo to, let's say 40, it will jump to 1 first then move to 40. I have to set the min to where my servo is centered and at rest. In my case 50. This way when ARC starts I see the 50 position in the servo control window and it will not jump.
Tameion, thanks for the suggestion. I need to study and think your script over. I'm still a little slow on script writing. It sounds like a great to solution. When I'm done I plan on having 100's of phrases. Do you think your script is practical for that many files on the MP3 trigger? Thanks for all the kind words!
Erreul,
My next step is to start building the bubble and brain. This is going to be where I have the third EZB sitting and the tracking camera. Hope I can find the room and make it look good at the same time.
Will, if you're reading this, thanks for making my big guy more realistic!
I know i am ROBOHOLIC
You are giving life to one of the best robots to ever be on television or in a movie.
I have a 2 ft tall B-9 that will be getting his own EZ-B when I get time to work on him.
I have a few other robots ahead of him right now.
They both have an EZ-B in them as well.
My android may end up with many EZ-Bs as she will have amazing movement in her face, arms and hands. Hopefully even walk someday if I can afford the servos she will need.
I love the way you are using the robots voice.
I need to get MP3s to play on command for mine.
Keep up the good work team!
If you ever decide to teach a class, sign me up!
I'm close to finishing the upper part of my B9. I'm close enough to the end to share a video tour through my brain, err, B9's brain and how EZB controls it before I screw it shut. Here are a couple vids:
The grand tour. Thankfully for everyone I ran out of battery power before I was through. It's 13 minutes. Grab your pop corn :
Here's a video of a camera view from inside the bubble. I was working out some motion issues I was having with the camera seeing reflections on the surface inside the bubble. Kinda cool. I was told it looks like an old "B" horror movie with the monster walking up on the mad scientist with his back turned:
Enjoy!
Dave Schulpius
You are building such an amazing B9.
When I was a kid I was collecting everything I could find to build my own B9.
I had part of the inside of a washing machine for the upper body and vent hose for the arms, but I had nowhere to put such a large robot or anything to use for a bubble head.
I did make working claws.
Someday I need to take on this project.
He can stand next to my full size R2D2.
i just got mine in,they handle up to 4 amps each and less then half the size of TIP12O
ON the scratches on acrylic i bought a 4 part kit that removes scratches very well,light,med are very big scratches,plus 4000 grit sandpaper.
THE B9 looks great
Light scratches often come out by just buffing with toothpaste (of all things, but gives a minty smell to it)
I used it for a very badly scratch dome and i first used wet dry 400 and then 2000 wet dry grit sandpaper and the polish and it looks brand new
As soon i get a chance i wanted to make a very good tutorial on plastics how to them and polish,since at work thats we we do a lot for out plastic housing for handhelp test equipment we make and sell.
scratch remove kit
The company i worked for got info on this from a company who makes plastics and said it was one the best.
and the bottles can be used for many many projects.
WE tried toothpaste kinda works ok,but does do a great job as product made for plastics,its one of my favorate tools for robot making ,since toys always get scratched ALOT
Very very cool!
Your professional touch to every detail is awesome.
I love looking @ your videos on how each level operates.
There are so many great robot builders using EZB that are encouraging to me.
Thanks
Steve S.
I've been using Novis 1,2 & 3 on pinball play fields for years. Works great on diamond plate and automotive clearcoat. I like to use a random orbital car buffer starting with a mix of 2 & 3. Then move down to 2 then 1. The playfields look like glass afterwards and the balls fly!
I have not had as good of luck on clear acrylic. I always seem to have a haze left over. I am afraid to try this on my bubble and leave it looking foggy. It would be like looking through dirty sun glasses.
I would i know toothpaste does leave a residue on some plastics why we dont use it anymore,but great news they gave a lot out at work when we stop using it.
WE at work test the products always on scrap plastics first using different ideas to find the best one,main reason it goes to customers so it has to be 100% perfect,or they wont buy any more of our products we sell.
I am happy for you. Someone had to make this beautiful creature.
Kudos!
Also one tester i design and made is so strong it would pull you in to the machine
Just think it 10000 amps (yes amps) at AC pure sinewave would do.
Den
I've got his bubble in place now. At least till I decide to take it all apart and do some mods I've thought of. I'm just now finishing up his radar ears. Yahh! No more blue tape to watch go round and round. Had an epic fail with the yellow paint on one of the ears. Had to wipe it off and start over or I'd have them mounted tonight. *tired* Well, tomorrow is another day. Here is the final coat (I hope) before they get glued to the base and clear coated. Then mounted onto the radar. After that I'll do another video for our enjoyment:
I'm sure the total cost was pretty expensive.
Other than that, you should show him to school age kids. It would surely inspire them.
I'm not building for the money. Only fulfilling a childhood dream. Anyway, there are a ton of copyrights owned by several companies and people that cover this robot. Even his voice and words are copyrighted. They are very jealous over people making money off from their image and sounds. However they are very open to people making him and appearances where no money is earned like schools and conventions. So, he'll probably stand in my living room or game room for my enjoyment. I'm making him so I can take him apart in several parts but even so all parts are very heavy. It's a process. He wont be mobile (for now) so I wont have to worry about him running off to gain fame and fortune.
Just ran across this site.
You probably have seen it but thought I'd link it just in case some others might be interested in B-9 sounds.
www.danmonroe.com/themagneticlock/ramblings.html
Irobot58, I don't have any plans for another robot after I finish this one. I will probably work on developing his personality and starting a few projects around the house and in my other hobbies that I have been putting off. With that said the next time I do a robot it will probably be either A full-size Robbie the robot or r2 d2.
Herr Ball, I have that set of CD's from the magnetic lock. They are well done in high quality. Hundreds and hundreds of sound clips are included. I had a problem though getting them to ship it to me after I paid for it. I had to file a Pay Pal disputes to get them to notice that I had sent the money. Others have had this problem also. So if anybody wants to buy this set be prepared for a long wait before you get in a little bit of a fight.
Just as an FYI, I bought that same set of discs. I had no problem getting them. Maybe he got his act together a little after you had to deal with him.
I wanted to chime in with my plans if I ever get my robot completed; some food for though, I guess.
I have been collecting stuff for several years. I will have a lot of spare parts left over. I am thinking of making a female version of the B9. Maybe I'll call it the G9 9 (as in "Boy9" and "Girl9").
It will be a similar design but smaller. The torso will be more shapely, somewhat buxom, and taper in at the waist. The neon will be shaped more like a "V" neckline. Instead of the pants separating into two legs, it will be more like a dress. The arms and claws will be more dainty. Of course, the colors will pinks and pastels.
Who knows? By then, EZ-B might be so advanced that the two of them might actually fall in love.
I have seen this picture before, and can't help but wonder, When they loaded up the "Jupitor 2" to head for Alpha Centuri, who had the inclination to pack a Shirley Temple wig that would fit the robot.
But I digress.
A G9 sounds cool.
I have a .wav file that says "Incoming Email Transmission" using the B9 Robot's voice. Do you have this already? If not, Email me and I will send it to you.
rex(dot)gordon61(at)gmail(dot)com
Then you could just type what you wanted and he could say it.
@Rashad Sounds like an amazing project.
Is your walker going to be humanoid?
I will be printing parts for my humanoid robot. He is under 3 ft tall.
I am building a full scale android, but only internal parts will be 3d printed.
I hope to give her latex skin.
Dave, you said that you have the "Robot Ramblings" CD. Have you thought of editing sounds off of that, with Windows sound recorder? You can copy and past parts of sounds together. If you can find the word "day" and splice it in with the end of the word "save" or even "give" or "have" and you'll have your name. It would be fussy work, and you'd have to find clips of similar quality, but I am sure that you could put together an authentic rendition of your name.
My name being David, I may have a harder time editing his voice to say my name.
I need to get the CD first.
I may watch a few episodes and see how well I can do the robot's voice, then use some software to make it even better.
Not that I am the best at doing voices, but it wont hurt to try.
I'll work on "David."
Dave, does a lot of videos. Maybe he will incorporate it into one of his future ones.
Thanks again everybody, Dave
I was able to find some time this afternoon to take a video of my completed bubble and brain section. Also was able to show a good example of my B9's basic movement and animation abilities with the three EZB boards working togather. Sadly I was not able yet to load the sound files sent to me by Danger & Rex. Later. Cant wait.
He's really looking like the real thing now that his head is in place! I also completed the radar ears and Power Pack. These finishing touches really add to final look. Now I need to get his pants on and start working on his arms. Once all the building is complete I will work on more advanced animation routines and development of a better personality within ARC. What Fun!
Enjoy:
Thanks for your vids and updates..keepem coming as they DO inspire the Forum people to keep on EZboting myself especially!
Yes I struggle with remembering all the commands as well...
Some of the ways I have tried to counter this have been:
1. Command "What can I say?" speaks the growing list of the main verbal commands
2. Command "Visual Commands" toggles a command to display a webpage of the command I have using the script command EXEC to open a Chrome webpage stored in Evernote. I use a dos kill command to terminate the browser.
3. Command "Audible Menus" does the same as the above but speaks out the specific command before waiting for a response.
Very roughly...
if ($verbal_menus="On")
exec .... jump to web browser commands...
Elseif ($audible_menus="On")
Saywait ($Command_Phrase)
Endif
Continue with script....
I do the same with toggling sound effects as well...
- Wayne
Weird cant seem to make a YT link work? anyways....its just a IPOD holographic projector
irobot58, I hadn't had any Jack yet when I filmed this vid. Maybe that was the problem.
Irobot58, your link isn't working.
Thanks for the cool video!
Robots tend to not hear everything we say.
Kids are the same way. lol
On vacation I did take my laptop along and I had some time away from the beach and was able to do a little cleaning up to my project in ARC. Mainly got Script Manager set up and transfused my scripts to it. Looks like it will be a nice addition to clean up the look of the desktop and make it easier to find the script I need. I'll try it all out before I delete all the old copied ones from the desktop.
I made a post in another thread where a fellow builder made a comment about limit switches on DC motors & H-Bridges being a good idea. I replayed and thought that reply would be a good addition to this thread. I may have said this before in this thread but it couldn't hurt to repeat it. So here it is. Enjoy:
>>>>>>>>>>>>>>
I'm a big fan of limit switches. They have saved my butt more then once. As you know it's very easy to send your motor speeding away towards disaster or your fingers while your setting things up. Limit switches to cut power on each side of the radius and keep the tears away from your eyes. By adding a simple blocking diode on each switch you can easily revise direction. Without the diodes installed there is no way to back off the rotation because there is no power flowing. Blocking diodes allow power to flow one way but not the other so with them installed in the proper direction you can simply reverse your h-bridge when your rotation hits the switch and run it the other way. Here's one of my setups. You can see the switches to either side of the blue pot I use for location feedback:
I looked through your post but was unable to find how you worked your lights with the voice.
I have read several ways but (being the noob I am) can't seem to get any to work ...lol.
Mind when you have time to explain?
Herr
The light on the neck of my B9 is neon. It's triggered with a Tech22 Neon power supply transformer and controller. Here is a link to them:
Tech22
I have the whole thing attached to one channel of my CF3 Sound System that all my sound runs through last. When it senses sound from this channel it will blink the neon according on how I have it adjusted.
Sorry, this wont work with LEDs or regular bulbs. Only with neon.
Good luck, Dave Schulpius
It's just for looks and a fun but I wanted to make it look and feel like the real thing. On the TV show it was never removed or used in any of the episodes so no one really knows what the handle and cord looked like so I was able to use some artistic leeway. All that was really seen was the head sticking out. The mic head I ended up making isn't exactly the same as the original but it's close enough for me.
As for building the mic I got the body of it from a fellow B9 builder that had a few of these turned at a shop for him. I was lucky enough to buy one of them. I had to manufacture the screened tip and figure out how to get a coed attached. Please take a look at the following video and pictures for how I built it and how it turned out. Thanks and have fun!
Here's the blank mic body I bought from a fellow B9 builder. He had several made and I was luck enough to be able to buy one of them:
Painted and ready to start making it look like the real thing:
Here's the screen head after applying a layer of foam and weaved screen form a simple kitchen strainer. I placed a Brad Nail in the center to help it look more like the original mic from the TV B9:
All assembled and ready to attach the coard (from a real Radio Shack Mic I canalized):
Finished! :
Inserted in the proper place in the Torso:
For comparison here's a shot of the original Mic from the 1st season:
The man behind the curtain:
How it works:
Here's the engine. It's a Key Retractor I got off Amazon that hangs on your belt and holds your keys!:
Installed and working:
Here's a vid of the whole thing installed and working:
Within a few days I'll have a switch wired in so when I pull the mic out of the torso he'll say something. I have a few sound files where he mentions the mic.
I have a few sound files of the original voice actor, Dick Tufeld, saying things like "You may examine my microphone" and "That microphone is only to be used to contact the Robinson's on the Jupiter 2". They should sound real cool when someone tries to pull out the mic.
That was an ingenious idea for the microphone retracting.
All of your work has been very professional.
Steve S
@MovieMaker, Those are two questions I don't really have an answer for.
I stopped counting money spent a long time ago. So far I've probably spent over 8K or more. I have most everything bought for the main part of the robot. However I still have to make the inside mechs of the arms. I don't really know yet what I'll need or spend to buy the things I'll need to get them to move in and out.
As far as weight, he's very heavy. I would estimate around 350 - 400 lbs. I don't plan to have him rolling around the house or yard and am not using batteries so I didn't try to keep him light. I used aluminum as much as possible but still he's very heavy.
I've built him in several easy to take apart pieces and easy to connect Molex connectors. When I do want to move him it will be easy to dissemble him and move him one peace at a time then reassemble him. Shouldn't take more then a couple hours and two people.
You know you are my friend and have very good wisdom and would not say anything bad about you.
only i just put 2 and 2 together and thought you was talking about me like others where doing
I understand this action of holding a load will eventually wear out or burn up a servo no matter how heavy duty it is. After all, my bubble section weighs over 15 lbs and the HD servo I bought cost about $250 USD. I don't want to burn this sucker out or strip it's gears. After weeks of brain storming and several different "mind designs" I decided on using supports under each side of the lifting fork to support the load. Then I'll have smaller servos controlled by ARC pull the supports out. This will allow the HD lifting servo to let the load "drop" the 1 inch I need to show the negative emotions.
I wanted to accomplish a number of things in this design. First was to keep things simple and the other was keep all the stress off all the servos and the mini drawer slide I'm using as a track to move the blocks back and forth and keep them in line. Also as in the rest of my build, I want this to last forever.
I've got it half built (one side of the lifting fork supports) and it seems to work great. Here are a few pics of my progress so far. I'll make a video of ARC controlling it after I get it all built.
Here's a test fit of the supports I built out of Acetal. A very strong, slippery and workable plastic. It should last and not wear out.
Here are a couple pics of one side of the support completed. I'm just using a standard (EZ Robots kit) servo with standard servo linkage to pull the blocks out from under the lifting fork. The block is cut so it straddles a mini drawer slide (used in doll houses) so it puts no stress or weight on the slide. A simple EZ Script lifts the main lifting servo a fraction and then pulls the blocks out. Then the main lifting servo can lower the load. The silver cover over the standard servo is Dynamat for sound proofing. It does quiet the servo down enough to keep the whine from sounding like my ex-wife but it still can be heard. It's livable now.
LMAO on Ex-wife reference.
Thank you!
Here is a picture of the final result and a video after that of the setup working:
Thanks for looking and enjoy!
Dave Schulpius
What holds the bubble in the highest postion? I am trying to think of a way to add another "hold" position.
This is the reason I love robotics, because of the merging of scripting/EZ programming and mechanics!
Best regards, Glen
B9 has always been one of my favorite robots of all time.
I need to build one once I get moved and have some room.
That ratchet idea sounds interesting. How do you let the load off? I would image you need to pull the pawl back with the servo. When the arm goes up do you hear the click, click, click? With your Steampunk thyme I bet that would fit in.
As far as holding the servo totally "highest point"; I don't see this as a problem. The bubble wont be up there for any length of time and it should be Heavy Duty enough to keep it up there for a few minuets or less and last. I just don't want it to hold for perhaps hours or longer.
Enjoy:
Thanks,
Dave Schulpius
Here is another picture of the original robot costume. It is in black and white, but it shows the details well.
Something I found out early was that the B9 robot evolved in his "look" as the three seasons of the show progressed. The first season was in black & white so color was not so important in that year. The pic you posted Danger was from then (of course you knew that
I had a good feeling when I started this build that I could get him to look reasonably close to the original. I had decided early on that some of the major parts that made up the "look" of B9 were past my skills to recreate with any accuracy. These being the bubble, torso, rubber arms. It was real important to me to get him to look original. I had planned to build the tread section and leg supports out of wood and the leg section could have been constructed using black pipe insulation to recreate the rubber ribbed look. Then I found the B9 Builders Club with many of the founding members were offering many of these parts. I was overjoyed to find out the measurements and molds were pulled from the original TV robot and were of very high quality! They were offered unfinished, raw and unassembled but that was just fine with me. I have skills in that area and now I was confident I could recreate a B9 exactly like the one I remembered from the show. I saved up money, worked overtime and sold stuff I had outgrown and soon I was buying these expensive but worthwhile parts. Now for giving him a personality and movement; I was going to just have him be a "puppet" or static model like you see in museums or even painfully learn programming or the Arduino platform. Then I found EZ Robots and I had a way to easily bring him to life. The perfect storm! Most of the rest of the build is mine or methods I copied from other B9 builders. I know I've overbuilt in places and made it more complex then needed but it's been an evolving event with no clear way to put this guy together and get him to work like I want. Also I was having a blast creating this and have no problem with overbuilding if done in a quality and useful fashion. I don't have to worry about weight because he is not mobile and does not use batteries. He'll just stand there looking real cool and plugged into the wall. I have no wood in this robot. Wood warps, cracks, decays and moves. He's all fiberglass, acetal plastic, polycarbonate plastic, steel and a lot of aluminum.
One thing I'm excited about is the ten chest buttons. I have a few buttons attached to things like activating the soil sampler in his tread section, activating a separate Bluetooth receiver that will catch transmissions from a program on my laptop called I-B9 that I can control different phrases and words he says (great when in a convention, party or show setting and want to really interact with the crowd like a puppet) and a couple buttons going to the CF3 sound system to trigger random voice files when I don't want the EZB active and using the MP3 Trigger. I can turn off the main EZB with a flip of a switch in the programming bay near his waist if needed. I also have about 5 other chest buttons attached to ADC ports of one of the three EZB's. This way I can run scripts right from the front of the robot by pushing one of the buttons assigned to them!
Sorry for the long rambling, I get excited sometimes and cant stop myself! *blush*
Thanks for everything, Dave Schulpius
Again, I am not trying to be nitpicky, just pointing that out.
I agree with you the third season is probably the best one to copy. Sadly by then the crown and radar no longer turned. He had been banged up, repaired, and repainted many times. But, as kids watching the show, we were pretty forgiving. He was the hero and companion that every kid wanted. You're very close to having that dream come true.
I am going to try to get mine as close as I can, but I don't have the time or resources to make it exact. If I focus on every little detail, I'll never finish it. It will be close enough for the average person not to notice.
Keep up the good work on yours. I think that most of us would agree that it is one of the most impressive projects being undertaken here.
Thanks for keeping me honest. Keep the fact checks coming!
Dave Schulpius
have you ever think about host a workshop on building, or something. So many people around the world could learn so much from you!
best regards from Brazil.
Tevans.
For some silly reason this question about what season robot the pictures we were discussing has been bugging me. You'd think I'd have better things to think about.
Anyway, The picture you first posted is from the last season 3. Will's custom and the upgrades to the robot show that (like you said
Just to prove here are a couple pics from each season. Notice the neck line of Will's shirt and compare it to your picture. The first one below is from the 2nd season. You'll see the second picture below from the 3rd season matches your picture.
Season 2:
Season 3:
Sorry if this is boring to some but I have a blast with things like this. It does help to keep history correct.
I had never noticed that the costumes had changed between Season 2 and Season 3. Looking at Billy Mumy, I was sure that it was not early in the show. He was only eleven when the show began. His appearance and Angela Cartwright's changed dramatically over the three years of the filming. I guess that at that age, it should be expected. That is why I posted the second, early, picture of Will and the robot together, for comparison.
I always liked that picture anyway. It shows the human element of the robot; Bob May, the man inside of the costume. He had a very close relationship with the actors. Even though the robot was supposed to be a machine, his acting, and the chemistry between all of them really brought that character to life. Otherwise, I do not think that he would have become icon that he still is today after all of these years.
Do you keep a build journal? Or some type of notebook with all your drawings and schematics? I do for Questor. The notebook is now 4 in thick! And it looks like I will need to start another one
I love the human look of B9's arms when Bob May moved them around. I hope to capture at least some of that look when I figure out how to build my B9's arms. I'm still trying to figure out how to build them and what structure and motors or servos to use. I'm toying on building them on some kind of pivot point so they can go up and down along with in and out. I wish I could come up with a way to have them also bend at the elbow without spending thousands of dollars on motors and joints. They will be very heavy when I get his claws, wrists and rubber arms assembled. Not to mention the supporting structure and motors or servos.
@rgordon , I have kept a journal of sorts. Sadly it's not very complete. Mostly hand drawings and schematics of most of the wiring of power, fuseing, switches and logic. I have miles of wire, tons of boards and control units like EZB and CF3. I do plan on sitting down after the build and cleaning up the drawings and put together an "Owners Manual" showing operation methods, parts and board and wiring schematics. He's very complex (probably more then needed) inside but very easy to operate from the outside once someone knows what to do or push or what voice command to give. So a manual is a must.
j
The fixed frame is inside the B9 and the moving frame extends outside the robot under the arm tubing. The elbow servo could be mounted at the end of the moving frame. The magnet fixed to the moving coupling flags (to the controller) when the moving frame is fully extended or retracted.
Yes, its ok to contact me on the email address given in the other thread.
Tony
Also I'm at the end of doing a few needed upgrades to B9 before I put him back togather in his final resting place out of the shop. When that happens I'll only have the arms to work on and do a little retro fitting after I buy a few of the new mini EZB's.
Thanks again. It's good to have a plan! I'll contact you through that email soon.
Dave Schulpius
http://www.firgelliauto.com/product_info.php?cPath=81&products_id=353
We use some for a job at my place, it's a little slow but not so much. It can take very little space in the arm of your robot and permit 12'' extension.
You impress me a lot and I look forward to the completion of your robot.
Thanks for the link! I took a look and they have many different types and lengths of linear actuators. They even have high speed actuators that can travel up to 9" per second. I had first wanted to use linear actuators but thought them too slow for my taste. However 9" per second is about what I'm looking for. They even offer them with sliding brackets but the fastest they offer is 2" per second. Sadly that's to slow.
I'm really getting some good options here and will probably take a little from each to make this work.
Thanks again everyone!
Here's a few pics of the process:
I'll have to remove the leg support section above the knee plate to do this. I've built this so I can easily do this.
I've removed the leg support section. Here is where then legs will tuck down into. Once I've cut out the dividing rubber where the leg section is glued to the knee section I can tuck the rubber legs down in place. As you can see there is not much room between the feet pipes and the surrounding knee plates. I'll have to trim the glued section out very close to the edge of the leg bellows.
I've made a support that will allow me to have the legs supported so I can trim safely and also have a cutting surface.
Getting ready to do the trimming. I have the support/cutting board inserted inside here.
I've marked the edge with a paint pen so I can get close without cutting through (Hopefully).
Coming along nicely a little at a time. Chunk by chunk.
DAMM! Got a little too close. Now I need to make a repair. Shouldn't be a big problem. A little silicone glue and a rubber path and it will be good as new. Just like a bike tire.
Starting the fit
All installed and ready for the leg support section to be fit in. You can see how close the feet are to the outside edge. It's going to be a tight fit.
The leg support is in place. Looks OK but I don't like the way the side of the legs squash in above the knee plates. I'll have to cut support rings out of acrylic to fit inside the leg bellows. That will keep them formed properly.
Will
I agree that I built the leg support unit a little too short. That's easily fixed by adding spacers.
Also great ideas about installing support rods to keep the shape and form but I like the one about filling the legs with soft foam. I'm not sure what I'll end up doing right now. I did already raise up the leg support unit 2" and the bellows straightened out. I may just keep it that way but I do like the look of the Hero Robot's squished look. However after lifting the unit 2", there is still a slight collapse of the vertical wall. Perhaps a couple interior rings will clean it up.
Thanks again, Dave
I'm going to try to make a few retainer rings to help with the inner support of the rubber bellows and lower the unit back down.
My favorite soundcontroller is the Winbond W523Sxx this small chip can have up to 60 seconds of quality speech/sound effects, it has 4 trigger (input) lines and 5 output lines (for motors, LEDs , relays etc), so the little chip can do quite a bit of stuff. The dev board for the W523 which is very small looks like this
So my thoughts are a W523 stand alone board connected into some of you existing B9 circuits could do some neat effects say if triggered by a nearby human (via a PIR sensor connected to one of the W523 trigger lines). I use these PIR modules on all my robots and have loads here in stock so I can send you over a few.
uk.farnell.com/panasonic-ew/amn11112/sensor-pir-compact-std-5m-white/dp/4160265
The W523 could probably have at least 20 B9 speech phrases or sound effects that could be outputted randomly of via sensor detection. Lastly the 5 outputs of the W523 would need to be parallel wired in to your BP effects units, so that the soundcontroller can at least control 5 things (lights/motors etc) on your robot.
Over the years I have developed and code released hundreds of W533xxx chips for hundreds of products, which probably equates to millions of end products sold with a soundcontroller that I developed, but I no longer need to do code releases and this has been the case for some years now.
Dave I would be prepared to do a custom soundcontroller board as seen in the jpeg above (code and add custom phrases/sound effects) for you (FOC) that could sense humans then output various B9 phrases (of your choice, all I need is the wave files), and it could also control some of your lights and movement, this would all be without the need for a computer/laptop connection. With my current workload, you would have to wait a month or so, but I would be happy to write the code and make up a board, let me know if this is of interest?
Tony
Those chips sound like a great idea.
How much would you charge if I sent you 20 phrases from my humanoid robot to put on the chip?
I am making a few robots so I will need more later also.
I am upgrading a 24" B9 toy also, but I think I haft to buy the sounds first to add to him.
mcsdaver@yahoo.com
@Dave
The hero look may look different when the man is in the robot.
I could be wrong, as I haven't even started building a full scale B9 yet.
Mine will have wheelchair motor drive just to make him a little more real.
Although he may weigh enough without the added drive motors.
I know getting him from place to place will be a lot of work.
My full scale R2D2 is very heavy and not even finished yet.
He might need an EZ-B. Hmmmmmmmmmm.
Your B9 is looking more hero B9 all the time.
He will seem so alive to everyone, they will think a man is in there. lol
Great work!
Will's mold measures 19" so there is a big difference. When I'm done I think mine will be about 16 1/2".
I've been working all evening on the support filler to keep the bellows from caving in along the bottom. It's very low teck. Like Will suggested I got some soft Styrofoam from the local art and hobby store. All I could find were sheets that were 1/2 - 3" thick so I bought a few of each. After some fitting and cutting I ended up using mostly the 1" thick sheets and stacked a few on top of each other in the center of the bottom of a leg. I then sandwiched them between two other sheets turned up on end. This put strain outwards to the sides of the lower legs and keeps the bellows from caving in. I then took them out and glued them all together with 5 minuet epoxy. Any other glue I tried to use just melted the Styrofoam. Tomorrow I'll install these low tech wonders and take a few pics of the end product. I've got a urge to glue some aluminum angle on each just to make it look more manly.
@Tony , I'm honored your thinking of my build and trying to come up with ways to make it better. Thanks, I'm always welcoming this type of help. Lord knows I can use all the help I can get. Your suggestion sounds like a great addition. I wish I had come across it 2 years ago. I've already spent a lot of cash and got a secondary system installed to operate independent of the EZB. I have a CF3 Sound System with a Contact IO 8 module installed. along with the 2 already onboard relays this adds eight additional switch/motion sensor inputs, and eight additional relay outputs to the CFSound III. So I can have a whole slew of sound files and motors triggered without the EZB being on. It's not cheap but it works very well. Here's a link to this product.
www.cfsound.com/index_CFSound.asp
Here it is mounted above my little amp:
And here it is looking at it from the underside when I've got the CSS opened up. It's the big green thing in the center:
Again, thanks for thinking of me. I wish I would had this offer when I started my build as the CF3 was the first thing I bought. It was how I was going to animate B9 before I found EZB. I had spent all that cash on it so I used it anyway as a secondary platform.
Dave
@mcsdaver
I would be happy to do you a custom souncontroller, and I would not want any money for it, just cover the postal charges. I cannot supply too many soundcontroller dev boards as they are not available to the general public and only issued to Winbond code developers and they ration them to us.
I need to know what you want the soundcontroller to do? If its just a simple one phrase after the other (sequential) every time a trigger line is clocked then its just a couple of hours work, this includes processing your wave files so they are compatible with the soundcontroller. If you want the EZ-B to be able to select any phrase then its a bit more complicated, the W523 does have a mode (called cpu mode) where it can become a slave device to an external microcontroller, but it uses an unusual "bit-bang" serial routine and may be difficult to script in ARC, I usually do the W523 comms in assembler for my applications. Let me know more about your requirements? We could start another thread on this, so we do not hijack this one.
Tony
Sorry in the delay getting back to you on this. Things have been very busy in my life. I hate it when life gets in the way of the fun things I want to do like building things and learking on this forum.
How can I refuse such a grand offer! I have to accept and work it into my B9 somehow. Thanks! I'll follow up with private email and find out the details of what you'll need from me. I am in no hurry as I have lots of other things to keep me busy. You may as well hang onto my package of motors and send everything at one time.
As far as your email with shipping details; No I didnt see it. I'll check my pile of email and my spam folder. Sometimes things I want to see end up in there or I just plan miss it. I'll let you know.
Thanks again, Dave
Before the filler:
After the filler
Here's the TV Robot's leg section I want to imatate:
Here's how I did it:
Before reassembling the leg section I had to raise the support unit. I simply cut 4, 1" spacers out of the same size tubing and slipped them over the round tubing, re-drilled and bolted them down.
OK, starting the final stack up!:
Sorry guys, I couldn't help myself. I was assembling my leg section with the aid of rope blocks and clamps and I had this visual image. Thought I'd have a little nurd fun on a Saturday night. Here's a pic I snapped:
j
I've now got the first complete stack up done. I call it "PHASE 1". He looks real good with his new pants on. After I raised his leg supports 1 inch he stands 6'6" at rest and 6'8" with the bubble fully up. Here's the final look:
In the next few days I'm going to install the camera and try to get the tracking working again. Last time I had it mounted inside the bubble had got to many reflections form lights. I kept getting false images and couldn't get the tracking to work properly as the camera was moving with the swing of the radar. I wanted to get my radar section to follow people as they move past. This time I'm going to mount it on the tread section and it will stay stationary. I'll try Relative tracking this time.
After I attempt that and either succeed or fail I'll be moving him out of the workshop and into his final living space (yet to be decided upon). Kinda like being born!
The next steps will be designing and building the arms and developing his personality and animation routines.
Thanks for all the help everyone has given so far!
Dave Schulpius
Looking into the new year I plan to finish up a pinball machine deep cleaning and mini restoration I involved in now (of a Lost In Space pinball game of all things) and get then get back to work on the robot. I plan to get restarted on the robot in just a few weeks. The biggest task is going to be designing and building the arms. This will be a challenge as I've said many times before that I know little about robotics. This will be a challenged because I'll need to figure out the best way to get his arms to move in and out and also bend up and down at an elbow location. I'd also really like his shoulders to move up and down. I also will want the claws to open and close. This last task should be the easiest part. I've had lots of ideas and help from people and read tons of ways other people have done this. The hard part is I've not seen anybody do this all at once. I think the trick will be finding the proper motors, liner slides and positioning feedback sensors.
I'm really looking forward to working on this and seeing what the end result will be. As always I'm open to any ideas as I move foward.
Thanks to all who have helped and inspired me over the past two years of this life changing and fulfilling project!
Dave Schulpius
Magnus for quite some time now. I bit off more than I could chew. So I put him on hold till later and have been working on smaller bots.
I would definitely like to help you with this phase of the build because it will ultimately help me with the arms on Magnus later. So I will be glad to help you research the design and help brainstorm. I am sure others will pitch in. This is the best forum ever! There is so much talent here.
Also Jerry Chevalier's build has a neat way of moving the arm up and down.
@Danger! I think, had an arm design he was working on for his B9. Maybe he can help out.
@Danger, Cant wait to see your pics and what your coming up with. I've been looking at many designs on robot arms that have been designed already for both the B9 and other robots. B9 is unique in that it need so slide in and out, bend at the solder and elbow and have the claws open and close (and maybe even spin). Along with all that it is very heavy and it needs to be built to support load and have motors capable to lift it. I'm estimating 10lbs each if using good strong but light weight materials like aluminum or some type of plastic.
-Inside diameter of rubber arm
-Length when fully retracted
-Length when fully extended
-Distance from torso arm opening to the elbow bending point
-Distance from elbow to the wrist joint
-Mounting method to robot torso
-How much room is there inside the torso for mounting shoulder motor
-Any other info that may be helpful in determining various arm designs
-Do you have dimensional drawings of the arm opening?
-Does elbow need to bend 90 degrees?
-Does shoulder need to rotate more than 90 degrees?
With this info I may be able to build a full size model out of wood to functionally test out ideas. I have some small linear actuators (50 lb pushing force) that may be able to move the elbow but I am not sure how fast you want it to move. These are strong but not really fast. For the elbow I am thinking something like a Backhoe arm....
I have some rubber arms that I had ordered for Magnus to help with the simulation but, I don't know if it is the same as yours until you get me the dimensions.
Let me get these specs together and get back to you. I've been wanting to gather this info anyway so thanks for the push. I like the backhoe idea. I operate one every so often at work so I'm understanding where your going with this.
Thanks for the interest.
Dave
Here are some more questions and items to ponder:
-Will the arms be used just for animation or will they be required to lift more than their own weight?
-How fast do they need to move?
-These arms will be heavy even if we use aluminum, PVC, wood or combination of these materials. Will need to have strong structural support at the shoulder joints.
-Is there enough room inside the torso for wheelchair motors? Gonna need a lot of torque for the shoulder joints.
-Need to think about how to get the wires through the shoulder joint area to the other motors and servos inside the rest of the arm. ServoCity has some great bearing, hollow tubing, and gearbox arrangements to consider. They have a lot of application photos of their products that would be good for you to look at to get ideas.
-How will you remove the torso for working on the central support structure with the arms in place? I had the arms on Magnus so that I could open a small hatch, pull one pin, disconnect a multi-pin connector and the arm would slide right out. However, I don't think you have that much room inside the torso. I will find the the pictures of the arms I had started on just to show you what I mean. I was using some 400in/lb torque motors, chain, sprockets, and pillow block bearings.
-Will all this extra weight be a problem for your torso rotation motor, lazy susan and limit switch setup?
-Worst part is realizing that this is gonna cost quite a chunk-a-change *eek* But I think if we do get it to work well, then, it will be a revelation for all B9 Robot builders everywhere 'cause I have never seen a completely functional B9 arm in all my travels on the web. Most only extend and retract. Some only move up and down. I have not yet seen one with 5 DOF. (Shoulder, extend/retract, elbow, wrist and claw.)
I'll be getting all the info and some pics today or tomorrow and post here. I have a couple new puppies in the house to keep under control and that keeps me jumping. *eek* Also have to take the mother-in-law out for her birthday dinner tonight. *tired*
In the mean time;
I only need them to support their own weight but I'm sure someone down the road will try to hang something on the claws or have him hold something.
I would like to see the speed fast enough to simulate natural relaxed human speed but not frantic.
I should have enough room in the torso to hold wheelchair motors and was looking at the ones Tony is using on his Lexi build. However the elbow motors may not need to be that big. Toymaker sent me a few nice worm gear motors that may work there. What ever the size the whole rear end of the assembly will have to pass out through the 5" (approx size) arm hole when I disassemble. I'm looking forward to seeing your pics.
My plan is to mount these arms is much like yours. I need to remove them by reaching in from the outside and unscrew them before I unassailable the rest of the robot. My rubber arms have a nice flange at the end holding them in place and I just pull them out of the opening to remove and reach in. I have a central support system that is free from the main torso. I plan to attache there. That will keep the weight off the torso and transfer it down through the skeleton. I'm sure my heavy duty lazy Susan and motors will accept the extra load. Routing the wiring shouldn't be a big problem.
Money? Well, I work a lot of overtime at work and this has helped a lot with funding this build. I don't see that changing. I'm in this far and committed to go the distance to realize my childhood dream.
Thanks for the interest and help Rex!
Let me first start by explaining this. My brother and I are working on this together. He currently is a trash collector. I drive for a company that hauls hazardous waste. Until recently, when work was scarce we would often haul tractor trailer loads of demolition from transfer stations to the landfill. We both keep our eyes open for anything that can be salvaged for the robot. My cellar if full of auto parts, small appliances, discarded street signs, sheet metal, too many cordless drills to count, etc.
I say this because my design is partly based on using parts that I have on hand. I kind of look through the piles of junk and figure out how to incorporate them together into something that might work.
The arm will consist of a "shoulder plate" that will be on a sliding mechanism in the torso to slide back and forth. Currently it is made of MDF. I will replace it with ¼" aluminum when I get the bugs worked out. When it is all the way forward it will be flat against the back of the arm socket. The motors to move the elbow joint are mounted on this plate. They kind of fit around the arm sockets inside of the torso. The scribbled sharpie lines that you see on the front are where I traced around the inside of the arm socket. The elbow joint is just far enough out to clear the socket.
The idea is to keep the weight of the motors inside the robot. I am not sure if this will work for you Dave. You already have a lot of space taken up inside your torso. I am going to get the arms in first, and then work everything around them. You may be able to make it streamline enough to fit.
To move the plate, I have a pair of automotive electric window tracks. I will mount them with the motors low and in the back of the torso to offset the weight of the arms. I am not sure if I will use drawer slides, or just put holes on the plate and have it slide on rods of some sort.
The elbow is a universal joint from a steering column. I am using them because I have a matched pair of them. This could actually be replaced by a short, stiff, spring if one wanted to go that route. The cables are thicker than they should be, and they are attached to makeshift spools in a haphazard fashion right now, but it works pretty well. Like I said, it is kind of a proof of concept right now. The windshield wiper motors seem to move it at about the right speed and torque.
The arm itself, as you can see is a collection of PVC parts. It is lightweight but strong enough for what it will be doing. If anyone is interested in the details I can explain what parts I used and how they fit together.
I do plan on putting a wrist joint out on the end of it. It will only pivot up and down. To move them I have push/pull cables that will be powered by motors and levers inside the body, again to keep the weight off of the arms. To open and close the claws, I could also use small cables, but actuators to do that would be fairly light. I haven't decided whether or not they will be in the arms, or in the torso.
Finally, to rotate the claws, I will need small motors in the claw mechanism. I do not see how to avoid that.
It is still a work in progress, but I know that you are trying to get ideas for your arms. Perhaps this will help.
(I am also including an early concept cg diagram that I made before I started).
I also like the cabling. That should help the arms to move an any direction. I had been thinking of this myself but worry about snagging on the inside of the rubber arm and nor having room. Perhaps I can stream line but I'll still to figure how to remove the whole thing through the arm hole. When I lift the torso off over the central support system it just passes by each shelf. I should have plenty of room in side the torso though of I can remove the unit through the 5 3/4" arm hole.
@Rex, Here are the measurements and some more pics of tolerances of what I have to work with:
-Inside diameter of rubber arm: 5"
-Length when fully retracted - 8"
-Length when fully extended - 18"
-Distance from torso arm opening to the elbow bending point - I'm guessing about 5"
-Distance from elbow to the wrist joint - About 13"
-Mounting method to robot torso - Will mount directly to a shelf of the CSS or it's threaded rods.
-How much room is there inside the torso for mounting shoulder motor:
Depth of torso - 11" to 17" depending on where I measure from arm hole to rear radius of torso wall
Width of opening between CSS & Torso wall - 7" at widest point
Depth from bottom of arm hope to bottom of torso - 9" but most of this is unusable.
-Arm opening - Perfect 5 3/4 round hole.
-Does elbow need to bend 90 degrees? - Yes or as close to as possible.
-Does shoulder need to rotate more than 90 degrees? - No, Just up and down about 30 degrees or so.
Inside diameter of rubber arm: 5"
Rubber arm is held on by it's own flange
Arm fully extended at about 18"
A view of the area where the arm will attach
I want to attach the arm unit to the either a shelf or threaded rods of the CSS
Hope this is all clear. It's great getting help from you guys. Three heads are a lot better then my one feeble mind.
Thanks again!
Incidentally, I did some experimenting with the arm as it is now. It does not move as fast as a human arm does, but at about the speed one would expect from a robot (perhaps two seconds for a full swing from top to bottom). It can lift a five pound weight at the end of the existing arm without difficulty. That was running off of a small battery charger. He may do better with a real battery. He won't be shoveling snow, or moving furniture, but he will be able to shake hands or pass out beer at a cookout.
BTW, what's your measurement from the torso arm hole to the elbow?
I was pouring over pictures on the B9 Builders Yahoo site and found a recent addition of an arm build that may be just what I'm looking for. It looks like it moves in all the needed directions except up and down but that my be easily added. The best part is it looks like the builder is close to my location so I may be able to meet with him. Here's a picture:
Here's another version by the same builder. Not sure if it's 1st or 2nd gen:
Seems most of the in and out action is handled by Linear Actuators.
Oh and I do love the B9 robot you built, it's a nice addition to the room
@Troy, I agree the heads are creepy. I have some stands being made for the larger ones so they wont look so disembodied. They will actually have negatives of a sort. But that's okay, because i love butterflies and rainbows.
First, and most importantly, you are certainly welcome to attend any one of my cookouts. If you are in the Boston area any weekend in the summer, stop in. The cold beer I can promise you; as to whether or not the B9 will be serving it, that will depend on my robot building skills. If I can just keep him from leaving the Tiki Bar and rolling straight into the pool, I will consider myself lucky.
As to your question about the measurements. the axis of the elbow is 6" from the shoulder plate which will be right against the inside of the arm socket. The end of the forearm that I have so far is 16" beyond that. This corresponds with my bicep about halfway out of the arm hole, as I would think Bob May's would have been. The pivot for the wrist (an inch or so in from the end of the forearm) will be about the middle of my fingers. I would guess that was where Bob would have gripped the claws.
A couple of other things I wanted to mention. If you decide to go with a universal joint and cable system similar to mine, it is important that the cables attach to the arms directly in line with the X and Y axis of the universal joint. Otherwise when it moves in one direction, the distance for the other two cables changes either making them slack, or binding them up. The other thing was this. Even though the arm will be relatively smooth and round on the outside, I am thinking about wrapping felt around them to protect the outer arm material.
I am not sure how much of my design you can use. The way that I have the motors on the inside, I do not see how you could easily pull the arms out through the front. The design with the actuators looked pretty good. It does look like it has the potential to snag the rubber arms though.
Lastly, I too want to compliment you on your game room. I am sure a lot of effort has gone into it, but it looks incredible.
PS
In message #227 did you mean to say they would have "shirts?" Dr. Smith could be a $---, but. Oh never mind.
here on post #87.
Even though you wont use it I thought you might like to see it.
Looks like they will last a long time.
Watching your build really inspires me Dave. Your work is museum quality. Maybe one day I will resurrect Magnus but, for now I will stick to smaller bots. My problem is I over think everything and try to incorporate to much into my designs. Which means the build never gets to a point where the robot is finished enough to be useful. The only robot I ever got to a point where it was actually moving around and behaving like a real robot is Questor. I learned a lot from that build. Now I have plans for a new robot that will be a just a little bit larger. Only problem is I will need most all of Questor's parts to do it.
Happy New Year to you and your family good sir.
thanks for your kind words....LOL they are kinda beefy aren't they
The motors I found for $65 dollars at a surplus place. They have 400 in/lbs of torque. *eek*
The motors were originally used to unwind and retract the awning covers on RV campers.
This moves nice and fast. Would not be very strong but the effect is nice.
Don't put yourself through to much stress over this design. You've already helped more then you know. I think in the end what I cone up with will be a conglomerate of many designs. Isn't that the way things that are worth wile always end up in the end anyway?
Take care, thanks for all the ideas and have a great New Year!
Your friend,
Dave Schulpius
Hope you are doing fine in the new year. How is the arm design coming along? I have not had time to get back to tinkering on it. To much going on with family matters. *tired* Have you heard anything from the gentlemen you were talking about that used the servo design? That looked really great.
Contact me via email when you get the chance:
rex.gordon61@gmail.com
I'll copy this to your email;
Thanks for checking in. I haven't had a lot of time either. Same as you; there has been lots going on on the home front. Trying to finish up some other non robot projects around here.
I did contact the guy with the design and he's more then willing to share what he has. However he's out of the country right now. When he gets back from China next week he's going to send me all his designs and animations. So things are looking up and I'll probably start gathering hardware soon. I'll keep you posted on how things are coming along when they start moving.
I think the motor drives that I sent you should make great arms for a B9 size robot if the motors can fit in the arms rubber boot material. You would need 2 coupled drive motor for the shoulder. These motor drive makes an amazing high torque, precision movement and very quiet arms, like you will see in the video below.
If the motors work out for your arm application, I could send you some more if you need them.
Tony
I do plan to try to use the motors you very kindly sent me. They seem like they are just what I need and watching your video I see they are close to the speed I would like to see. The best part is that it looks like they are strong and silent. They should also fit nicely into the arms I have.
I just don't have the time these past couple months to do much wrench turning on the robot. What time I have had I've been gathering info and trying to come up with the best way to put this all together. What I have decided is that I really don't need a shoulder motor. From watching the old TV show and looking at pictures, the actor in the robot costume (Bob May) hardly raised the arms at where the shoulder may be considered to be (which would actually be inside the torso). The arms slid straight out and then would bend them up and down and wave back and forth about 6" out from the torso. However he also did bend at the wrist which is about 4" below the claws. If I follow this example I would need the arm assembly to slide straight out on a Linear Slide. Then on the end of the Linear Slide I would have an 4 way elbow joint that would end up about 5" out on full extension then another 4 way wrist joint about another 15" or so out just below the claw. Of course at the end of all this in the claw assembly I would have a micro servo to open and close the claw (and maybe another one to spin it back and forth). The claw assembly should be the easy part. In the end it would look something like from the outside:
Again as I stated before my only concern with the motors you sent is that they not be fast enough to satisfy my taste. Didn't you mention in the past a way you were able to get max speed out of these motors? The other thing is that ARC has support for relative servo positioning and servos seem like they would be easier to get natural arm movement with the 4 way joints. I like the silent strength of the motors you sent but my concern it I may not be able to get the natural movement using only the motors built in pots and EZB's ADC ports.
If you have any comments or ideas I'm all ears.
Thanks again for all your help and the time you've spent with me on this matter, Dave
As you probably know I have produced the pre-production smart servo board for these motor drives.
But it is only a test production run (of 20 boards) so I do not have any spare yet, later when the production boards are made I will send you some, but that will be some months away. I take the shoulder motors up to 18volts (via a buck convertor) and get very good speeds.
In the meantime Rex is going to use the kangaroo X2 board and this should work very well with these motor drives, these drive motors are quite amazing and only have peak currents of 500mA even at very high torque, so you will not need a high amperage motor controller (that the X2 plugs into), in fact because of the low current requirements of these motor drives a L298 H bridge happily powers them without the need of a heatsink as you can see on my controller board.
Both ways these motor drives will act as standard servos as far as the EZ-B is concerned, so it is exactly what you are looking for.
Tony
Just to be clear, it looks like I'll need one Kangaroo x2 for each motor? Is this correct?
EDIT: Da! X2 means it will control up to 2 motors. Looks like it mates up to h bridges that can control 2 motors.
Too bad your servo boards aren't ready yet. They seem to be the answer to all my needs. *tired*
I'll have to take you up on your offer of more motors. I'll need at least 4 for each arm and I only have 4 now. However let's wait a few weeks till I get my design firmed up.
Thanks once more Tony!
Dave
Who knows maybe some fans will want to use yours in some fan made movies.
They will need someone who can play Dr. Smith well!
I need to find time to work on my B9 (large RC toy B9) that I got from ebay.
So much to do !
Funny irobot58 said something about C3PO, because I will be starting up my old C3PO project again soon.
Now that my 3d printer is working well again I will make the needed STL files to print.
My C3PO will be under 1/2 scale , but he will seem alive.
My Working C3PO Link
Your link doesn't work. I'd love to see what your working on.
What you may need to do is what I have done for the shoulder smart servos, by taking the motor voltage to 18volts I get 50% extra torque and a really nice speed. This is how the EZ:1 arm can easily pick up a full can of Coke with its end effector, what is amazing about this though is that lifting the Coke payload to the horizontal position the dual motors peak current (not start current) is less than 900mA! This is remarkable for an arm that is 600mm (2foot) long. Once in position there is no hold (motor) current as the motors are worm driven, only a few mA is required for the smart controller board and the Blue LED. The EZ:1 main battery is 12volt, but I get the 18V from a great little buck convertor that I got off ebay for about £10, it works really well.
Yes, one Kangaroo X2 would let you control 2 motor drives, so it should be quite cost effective.
Tony
I've ordered two Kangaroo X2 boards and hopefully they will be here when I'm ready to start. First thing I'll need to do is figure out how to use them and interface them into the EZB.
I'm just starting on motor control, speed and strength now.
Thanks to Tony (Toymaker) I have a starting place for my arm design; Motors. If you read back in this thread you'll read where Tony (Toymaker) had very kindly sent me some Bosch VMC Actuator Motors. These are very strong and quite little worm driven DC motors with built in pots for position feedback. These seem to be a little hard to find here in the USA but Rex (rgordon) seems to have found them. More later on this. You can read about them here:
www.ipar.nl/assets/files/Leveranciersinformatie/1540_Kat_EM_en_LoRes%20elektromotoren.pdf
Today I did a few tests to see if they will work like I want. I was very happy with the results. I really think these will work nicely. Now that I feel that they will work for me I need to start learning how to control them with EZB. I plan to also use a Kangaroo X2 board to turn these from DC motors to servos that ARC can control and make the arms hopefully look like the real thing.
Here's a vid I took of my testing:
J
With my double motor shoulder, both motors are controlled from one smart servo board (with only one feedback pot used) and it works great. The motors are so precision that they just mechanically sync even thought they are 2 different worm drives. I am sure this will be the same with your Kangaroo x2 setup, just wire the two motors in parallel and wire one feedback pot to only one channel of the dual motor controller, leaving you with a free motor controller for another Bosch motor servo further along in the arm (probably where you dont need so much torque).
My robot shoulders are running happily at 18V with no problems, remember that these drives are unique in that the motors are only drawing current when attaining their new position, once at the target position (unlike regular servos) these motors draw no current at all! Another amazing thing if you put an ammeter in circuit while you are trying to stop the shaft moving (like you did in the video) then I bet the current drawn would probably be less than 300mA, these are brilliant motor drives!
Here are some more detailed views of how I assembled the double motor drive servo, hope this helps.
This is a front view
This is the rear view
Tony
Thanks for sharing information on your build using the Bosch motors. I will be following your progress using the Kangaroo x2 setup, because I am using the Bosch motors in my next larger build. It appears they might work fine for your application.
Do I understand you right, that you will use a motor driver and the Kangaroo x2 setup?
Did you mention before, that you are already using a pot on another part of your robot being controlled by EZB?
Tony has been helpful with his use and experience of the motors.
EZ Robot community is great.
Thanks,
Steve S
@Steve, I haven't yet started working with the Kangaroo, hopefully today. From what I've read it needs to work with an H-Bridge or Motor Driver of some kind. It's designed to work with a SaberTooth but I think any motor driver rated for the needed load and voltage will work. Also if you want position control you need a pot. It also has inputs for limit switches and will find them automatically when trained. If you have no limit switches you can also train it where your end of range of motion is and it will remember and not go past that point! This would be a great feature for robot arms or other parts where there is no room for limit switches. I could have used this up in my Radar section. I has a devil of a time fitting limit switches up there. Here's a link to a PDF on the Kangaroo:
Kangaroo 2X
Also Steve, I am using several feedback pots throughout my B9 for positioning; Radar, Waist and Hip bend. In the scripts I have to use math to find the positions. They aren't too hard to to write once I found the proper equation. I only have to have it stated in one script and use other scripts to call it with the variable value stated where I need that script to stop rotation at. However if the Kangaroo X2 works like I hope, I can do away with that bulky script system and use ARC's native servo controls like the relative and multiple servos.
flex sensor
You could try to run the motors at 24V that should give another good increase in speed. Its not like normal servos, that if they are given over-voltage its usually the electronics that burns out and not the motor, what we have here is just the DC motor being given over-voltage which is the armature windings (the motors electromagnets). These motor drives are for the automotive industry and are high specification so the winding may well run at 24V with no problems. You could just try to run one (at 24V) for a time period (loading it if possible) to confirm it will be ok. If this does burn the motor out, I will send you a replacement. As I mentioned before because of the worm drive in these motors the windings will only be drawing current when they are moving to the next position so most of the time they will be in zero current draw.
Another idea is to have a simple 2 voltage switch where you connect your main 24V for higher speed operation then switch in the 18V or lower (via a buck converter) by a relay to get the slower operation. The relay could be energised by another port line on the EZ-B, this way its only 24V when you want faster arm actions like "warning Will Robinson" and then back to 18V for general arm movements that are more calmer/fluid, this would be much simpler to do this than trying to introduce PWM into the Kangaroo x2 (in servo mode) setup.
The plastic gears are high impact and incredibly strong, I do not think you will see any problems here.
Tony
I've found that trying to mate the Kangaroo to a regular h-bridge or motor controller is just to complicated for what I want to do here. Thanks to Tony (again) there seems to be a way to do it, although it's untested. He stated in a personal e-mail this morning:
Code:
The Kangaroo x2 is designed by the same company that makes the Sabertooth. It's used as a daughter board to add functionality and more ease of use to the Sabertooth. I could try to hack it to work with the pololu VNH5019 Motor Drivers I've been using but honestly I just don't want to go down that road. I have no wish to make this more complicated then needed. So I'm going to buy a few of the much less expensive Sabertooth 2x5 motor controllers so I can natively mate the Kangaroo to get my fluid servo type movements from these DC motors. The 2x5 Sabertooth can handle current up to 5 amps and there is no way these motors will ever pull that much. Also I really want to use the servo controls in ARC. I feel I'll get a lot better and smoother movements using them and be able to create more complicated animated movement routines.
I spent hours this afternoon ordering parts like brackets, hubs and Sabertooth boards for my arm build. It feels good to have a plan and somewhat of a design now. A lot of this will be building and adjusting on the fly.
Sabertooth has boards starting at 2x5, 2x12 and up. The Kangaroo X2 board says it works with 2x12 on up but says nothing about the 2x5. The board I want to use is the 2x5 so fearing that the Kangaroo wont work with the 2x5 I also ordered a couple 2x12's (just in case).
Now I wait on deliveries. No worries, lot of other things to do in the mean time. Fun times indeed!
I look forward to seeing how the flex sensors work for you. I've never seen it in action.
I guess I need to search more for a motor strong enough and small enough to do what I want.
I'm a little discouraged. I had hoped these little guys would do the job. Here's a vid of what it looks like. If anyone has any ideas, I'm all ears.
My robosapian project has such a mechanism using a spring to assist in lifting. I had it apart today in fact replacing about 30 wires that had insulation falling off! One of the forum members warned me of the common issue with the robo's...I really think it might wok for you even using the servo that you have!
Glen
Can it lift it with power hooked straight to the motor?
Glen, I'll have a closer look at any way I can add springs.
My EZ:1 arm has 2 Bosch motors at the shoulder running at 18v and the 60cm (23 inches) long arm can happily lift a drinks can in its claw.
I originally had trouble with shoulder lifting torque at 12v and that is why I went up to 18v which then worked great for me.
For quicker testing take the EZ-B out of the loop and just take the (parallel) motor wires straight to the 24v source, swapping polarity to change direction.
Hope this helps.
I had this setup attached to EZB because if it had been able to lift the load I would have ran another endurance test. I wanted ARC to continually lift and lower the load for several hours it see if the motor would take the abuse. Didn't get past the first lift. *tired*
I'm almost ready to just use the same servo I'm using for my bubble lifter. It's a Seiko PS-050 and a very quick and strong monster servo. I think it would snap your finger without a problem.
Seiko PS-050
However they're very pricey. They cost about 250.00 USD each. I'm having sticker shock over this one because after doing the math it's about 1000 USD's per arm just for the servos! Maybe a trade off would be to use these servo's at the first shoulder joint and something less expensive at the wrist (like your motors you sent me doubled up if they pass today's test).
I'm still looking throughout the internet looking for a small, strong and quick DC motor or servo that would handle this weight. I'm willing to double them up to give me enough power if I have the room. I need to keep the foot print of the assembled joints about 4" (101.6 mm) round. The motor setup you (Tony) sent me would have been 3.5" (88.9 mm) doubled up.
Tony
$perseverance = infinite
$focus = moreinfinite
$frustration = (infinite - 1)
If ($perseverance + $focus > $frustration)
sit back and enjoy your favorite drink
sleep(10 hours)
Else ( Chop knotted fir stumps)
sleep(no sleep)
Endif
Edit: I just noticed in my vid that I only have the PWM set to 55% witch means I'm only getting about 6v to the motors. I'll see what 24v gets me.
Glen, Love your coding!
Granted I'm only lifting 1.5 lbs here and I think I'll be lifting as much as 8 lbs when all is finished. Tomorrow I'll add more weight to see how it handles that. Then I'll remove the second motor to see how it handles that.
I'll take vids and post.
Thanks Tony for pointing this out!
I know you must be excited to finally get going on the arms.
Would you recommend those motors on an 0mni2000?
I cant really make a recommendation as I haven't really put them to any use yet. I've only lifted 1.5 lbs a few times. I'll know more over the next few days as I run them through my performance and endurance tests. Stay tuned. Same time, same channel
sit back and enjoy your favorite drink
sleep(10 hours)"
Looks like more voltage was the key! 8 pounds ! that sounds like, more then a six pack! Looking forward to your next video! Who knows , maybe one servo will be enough which will help out with the budget ...which means more um imbibing/celebrating
@Troy indeed ! with a successful servo test they will sell more for sure!
Alternatively, you may need to use the expensive $250 servo at the shoulder joint and then use the Bosch servos down stream on the arm where less torque is required (because of leverage) this could possibly save you $750 per arm.
Tony
@Tony, I'll see how far I can push these as soon as I have a little time.
Tony, Your right. After extending my leaver to 21 inches (53.34 Centimeters), about the same length that my final arm will be, I weighted it with 3 lbs (1.36078 kilograms). I found that the paired up motors running at 24vdc wont pick up the load. It hoisted it nicely at half the length though. So looks like I need to use larger motors at least at the first joint.
Here's a couple vids of the testing:
Here's the follow-up after the blooper:
I guess the next step will be deciding on the new motors.
This might work, not cheap though.
I use for Questors drive wheels. They are rated at 212 in/lbs of torque but they are kinda heavy.
motors
Wiper motor .....and an excerpt..... " This type of motor is called a"gearhead" or "gear motor" and has the advantage of having lots of torque. My unscientific test (using one wiper motor and a torque wrench) found that at 12 volts, on high speed, the motor has 13.5 pound-feet and on low speed, has 17.5 pound-feet of torque. "
Again, thanks for all the ideas. Please keep them coming. :-)
arm_calculator
Rex, I'll surly use this calculator.
Troy here's the links I promised:
Here is a link to a Google search with many results:
tiny.cc/8peuax
If you want to do a Google of your own, search useing this info:
Heater A/C Air Flap Control Motor - 4B1 820 511 K 98-05 Audi A6 S6 RS6 Allroad
Cheers!
Sorry Troy. I totally missed that you asked for this info. My bad.
Here's what I've found in testing them:
The pot inside only reads for 100 degrees of the rotation. Attach an ADC port's signal pin from EZB to the center pin of the lower set of three pins on the motor and 5v & GR pins to the two outside pins. Install a ADC control in EZ Builber and assign it the the ACD port you have the motor's pot attached to. When the motor rotates, the ADC control in ARC will move from Zero to 5v as the motor rotates within that 100 degree ark. When the motor passes outside of that zone the ADC Control will fall to Zero till the shaft come back around to the start of where the pot starts to read the 100 degree ark again.
If you remove the actuaitor arm from the shaft by carfully cutting it free with a Drummel or hacksaw the shaft will continully rotate.
The motor runs fine on 24vdc although it's only rated up to 12vdc. The two pins side by side on top of the three used for the pot are the power pins. I ran it for 6 hours with no load at this voltage. I never ran it for any leangth of time under load at this voltage. The higher the voltage is the stronger and faster it will run. With the motors doubled up I was able to lift a 3lb load 12" out from the shaft at 24vdc. I could lift nothing at 12vdc. I did not test lifting with only one motor but I'd guess the proformance is cut in half. It will hold a load without creaping down when power is cut.
The motor runs almost silent but does vibrate and hum a little when it moves under load. The noise is louder if you have it attached directly to a surface without useing it's attachment holes. These attachment holes have rubber inserts and if used with #8 machine screws absord moat all noise and vibration.
There is no noticable play in the shaft with the orginal arm attached. When I removed that arm I did notice a little play. This was really noticable when I added a hub and 21" bar to it so I could make my extended load tests. What was just a fraction of movement at the shaft was at least 1/8" or more of play at the end of the 21" extention bar.
The shaft will accept a 6mm hub. I like the clamping hubs found at servo City but a set screw hub will work also. If you use a clamping hub I dont think you will ever have to worry about it coming apart.
Although I havent yet tested this I feel these motors would be great used with the Sabortooth 2x12 and the Kangaroo X2 boards. This combo would turn them into servos and ARC would have easy native control through all the differnt servo controls and commands to build complex and smooth anamations.
Hope this review halps.
Can you privide a link to the specific clamping hub you used? Thanks.
I plan to use them on my next build.
Hope you find hardware for your beautiful B9 robot.
I also appreciate your "blooper" video and give you credit for posting it. Maybe we should start a "Blooper" thread post?
My robot threw one of her foam arms during a demo at the campground, and recently one of her new aluminum arms (hanging by servo wires) that I did not have fully secured, but no video for either time. They would have been priceless!
Steve S
I just got my two Sabertooth 2x12 in the mail today from The Robot Shop today. Cant wait to find a little time this weekend to mate one to a Kangaroo X2 board and see what EZB can do with it. If they perform the way I hope I'll use this combo on all my DC motors on B9 and turn them all into Servos.
As far as the "accident" blooper; my first thought was to reshoot the demo. Then I thought, "why not just be real"? Maybe someone will keep from hurting themselves by seeing what a dope I can be and what happens when I try to rush.
These are most of the parts Tony is pointing to that in his post about these Bosch motors but are cheaper and available from The Robot Shop here in the USA and other countries:
Large 6mm Set Screw Hub
Small 6mm Set Screw Hub
Aluminum Large "C" servo Bracket
Aluminum Long "C" servo Bracket
0.770" Clamping Hubs
0.625" Aluminum Clamping Hubs
It's really pretty simple once I spent a little time playing with it. I was able to build a simple circuit of one of the homemade boards I built. I still have to learn more but have the basics down. I'm not sure yet on how to share the circuit I built directly from that site but I used the Snipping Tool built into Windows to save it to my computer and then upload it here. Once I figure it all out I'll be able to actually design a board just by dragging around parts I dropped onto the screen. The software checks the design and then I can order the board to be built. They even will check the design manually before they build it. Seems pretty cost effective also because they offered to build 3 of these boards for about 10 USD each. I think I'll do this some day soon and replace all the homemade boards I've built and scattered around my B9. It will give me a better end result and probably more reliable. It will look better also.
The below circuit uses a digital signal and ground pin from EZB to switch a small npn transistor to ground. This then uses the third leg of that transistor to draw down a pin on a different light control board I have controlling lights in the brain and belly of my B9. This is a feature built into that board that will change a light pattern when a certain pin is grounded. If interested here's a link to that board He has a few vids of how it looks when working. It's a nice board but a little big for small robots:
Wizzes Workshop
I have a friend in the B9 Builders club that asked me to send him a copy if the circuit so he could build one for his B9. Here's what I put together at 123D Circuits:
Bret
Love to get in on any boards you would have manufactured, at any time in the future. As soon as I launch the Kickstarter, going to work on finishing the B9 build and ARMS! Trying to snag a larger 24 by 24 x 24 inch FDM 3d printer. Then I can offer an animatronic arm kit to the club with printed parts. Also trying to come up with special molds (rotating open molds) to produce thin lighter weight silicone arms for those wanting to animate. I think my current arms are about 3 pounds each.
@Bret, Glad to hear your building a bot again. I'll gladly share any info I have with you. I'm flattered you think my design would work for you. I'm very happy with my first design (the one you are asking to copy). It was very smooth and precise. Part of the frustration I mentioned was that I rewired everything in my leg section to run off a Sabertooth 2x12 and Kangaroo X2. I've decided to rip all that out and go back to my first system of a simple Pololu H-Bridge and a ADC port on the EZB. Anyway, please give me a week to get the drawings converted into a legible schematic for you. Like I said Hearth and Home are eating up most of my time right now. My waist movement system is also a 24v so it should work well for you. To start with the pot is a 10K multiturn pot that I have attached directly to an ADC port of EZB. Simply wired like this:
EZB ADC 5vdc pin to either outside lug of the pot
EZB ADC Ground pin to the other outside lug of the pot
EZB ADC Signal pin to the center lug of the pot.
Here's a couple links to pots that will work:
expensive pot
less expensive pot (the one I used)
Here are a couple pics of the waist motor and pot mounted in line:
Here's a couple pics of the limit switches mounted. They are wired normally closed and in circuit so when triggered the power to the motors will cut off. A simple diode on the micro switch will allow the DC power to reverse and back the motor off the switch.
Hope this helps for now. I'll get the drawing to you as soon as I have a little time.
@Will, I eventually plan to get my boards manufactured. However that may be a long tome from now. That's very low on my priority list. I kinda get a rush out of having my homemade boards mounted in B9 and working so well. There's ugly but sound. When I get to that point I'll gladly include you. It's great news that your planning to design and make a lighter version of your amazing B9 arms. Also the thought of you making available a animated B9 arm structure is mind blowing. I've seen your video of what you can do with the B9 arm and simple servos and want mine to respond just like that. I'd mortgage the farm to buy one of your kits.
Well, back to what's been going on; I'll try to make this short. The Sabertooth 2x12 and Kangaroo X2 combo turned out to be a exercise in futility. I really needed a good speed ramp up and down with variable speed and position control. As much as I tried I couldn't get a smooth movement or a good ramp. As far as I can tell there the Kangaroo has no native ramping in position mode. I had to write a script to pulse a increasing speed command over the serial connection to get ramping. Maybe that's why I cant smooth movement. Anyway, after reading the packet serial command manual there may be a way but I'm getting nothing but static from Dimension Engineering's support to help me confirm or help with a example code. Their manuals are written higher level experience users. It all adds up to scrapping the Sabertooth 2x12 and Kangaroo X2 combo and going back to the simpler, smoother and better responding EZB / Pololu H-Bridge combo. I plan to take a vid of how unacceptable the Sabertooth 2x12 and Kangaroo X2 is before I rip it out. I guess if I was a rocket scientist and knew C++ like the back of my hand I could get a nice ramp and smooth rotation but I don't want to spend over a year learning how to do all this. EZB will let me do it in a weekend.
If interested here are the threads where I worked through a few of the issues I had getting the Sabertooth 2x12 and Kangaroo X2 working:
www.ez-robot.com/Community/Forum/posts.aspx?threadId=5374
www.ez-robot.com/Community/Forum/posts.aspx?threadId=5304
www.ez-robot.com/Community/Forum/posts.aspx?threadId=5510
More to follow..............
The Kangaroo and Sabertooth combo is very nice, I am very impressed how well it works.
I was able to get close to having it work but was still having clunkly results. I tried to write a code with the suggestions you made useing a precentage to start ramping up and down. It worked but just couldnt refine it to act like I wanted to see. Maybe I was useing the wrong values to figure the precentage to get the ramping right. I was mostly relaying on feedback from the pot attached to EZB's ADC to tell an EZ script when to sent the ramp down EZ script serial command. It wasnt pretty and didnt act it.
You say you've got a script written into a PIC. Is it something EZB can transmit over serial to the Kangaroo?
@Bret, The gear set was bought from a vendor off the B9 Builders Club site. His name is Andy and he also supplies these for the R2D2 builders. I'll get the link for you later today when I get home.
Waist gears
I am using encoders for feedback, some people on this forum think that encoders are not very accurate, this is not always the case. Here the encoder give 64000 clicks per revolution, with the setup you see here after leaving it running for over an hour the drive wheels were still perfectly aligned the accuracy is unbelievable. With this type of system and anti-slip wheels a robot can move very accurately and map rooms etc.
I will get a video up of the Kangaroo rig working when I get some spare time.
Tony
Cheers
Richard
@Richard .....you must be on a graveyard shift, posting at 4:07 am!
@irobot58, the hall effect devices are usually low resolution, so probably will not work so well, you really need high resolution 2 channel quadrature optical encoders, I use these.
uk.mouser.com/ProductDetail/Avago-Technologies/HEDS-5500A06/?qs=RuhU64sK2%252bvR8YZlqGE72A==
I've been bugging DE about this issue and finally got back a reply. I was asking if there is any way to get ramping to work in Position Mode and a clue on how to do it. They said no, but.........
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
We do not have a combined position velocity and speed command. However there are ways you can achieve this.
You can send a combined accelerated velocity command until you get up to speed. When you are up to speed you can send a combined position velocity command and it will keep tracking at that velocity.
You could also write a loop to slow down. It would manually step the speed down as you get closer based upon position.
Also you could go into the control tab in the DEScribe software and reduce the system gains so it doesn't accelerate and decelerate. This would be in Control/advanced. Have the position gain double the position reset time and that will make it respond more gently.
We do have plans for these in the next firmware revision but that won't be out until summer/fall.
<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<,
So there ya have it. Looks like the fix for native speed ramping will come directly from DE by the end of the year. In the mean time Tony, the script your offering will be truly needed and welcomed.
@Bret, Here's the schematic you were asking for. This will allow you to use the EZB V3 (or V4) and if you have a problem and end up with a run away or overshoot the limit switches will cut power when triggered. The diodes installed will allow you to back the motor off the limit switch even though the power is switched off. Ya gotta love DC voltage!
EDIT, Added 5+vdc between EZB & H-Bridge,
@Glen... I am just a really early riser...
I also took this opportunity to do a few other upgrades. Biggest of these are replacing the L298n motor controllers that moved my radar and waist sections with my Sabertooth 2x12 motor controllers and a daughter board made by the same company (Dimension Engineering) called a Kangaroo X2. The Kangaroo X2 is able to give the Sabertooth both position and speed control turning my DC motor into a giant servo. I had a lot of challenges learning how to set this all up but in the end it's really pretty easy and it was only my own ignorance that made it hard. Here are a couple links to other threads showing how I over come my problems to make this work:
Turning Off A Sabertooth
Notes On Kangaroo X2
I still have a few tweaks to do on how they operate but for now they're working well. Later, with scripting help from a more knowledgeable member I'll BA able to add position feedback from the Kangaroo to the V4 EZB through the UART port of the V4 EZB.
I also had a few challenges in installing the V4 EZB's. Most notable was I had to get a new router because my older one would loose my EZB's IP addresses when powered down overnight. Here are a few links to other threads where I worked through these issues with the forum members and EZ Robot's support team:
Peaking Under The Hood Of My V4 Ezb
V4 Two Way Serial Communication - Where Do I Connect To Read?
V4 - "my Batteries Are Low, My Batteries Are Low"
V4 - Chimes, Voices And Sounds
Connecting To Two V4 Ezb's
Power In - Power Out On The V4 Ezb
Breaking Out The Sound From The V4 Ezb
Here's a vid I made from the above link "Breaking out the sound:
OK, Here's a few pics of the refit in the neck part of B9 (Lower pics to follow later):
"Before" picture of the old L298n motor controller in B9's neck before the Sabertooth install. I already have the new V4 EZB installed. You can see the buck voltage regulator that adjusts the main robot power down from 12vdc to 6.9vdc that supplies power to the EZB v4. This means that the output voltage of the Digital I/O pins will be the same voltage. After doing some figuring I concluded that the equipment powered directly through EZB will tolerate this voltage.
Here are the upgrade pictures:
Here you can see the Sabertooth with the little Kangaroo X2 sticking out of the right side. I love this little daughter board. It's simple to uses and set up and very powerful, smooth and precise. It actually adjusts the power feed to give a boost or buck when it "feels" a change in resistance!, The little brown homemade board sitting in front is a high side switch that lets EZB turn on and off the Sabertooth when B9 is powered down. See the link above "Turning off a Sabertooth" that explains this. Without it the dump battery keeps the Sabertooth powered up after I turn off the robot.
The little homemade board in front is a TIP120 switch that lets EZB turn on and off the ear sensors powered through a external 12vdc power source. The one in the rear controls the different light and speed patterns for his chest lights.
Here's the v4 EZB. That little brown board on the side is how I'm breaking out the sound into my main Sound system amp and speaker. See the link above on how I did this.
Tricking out the Bot
Enjoy! Thanks for looking! More to follow.
Dave
Again, well done and thanks for sharing it.
I really am glad I've switched over to the Sabertooth 2x12 / Kangaroo x2 combo boards. They are so smooth and strong and add a level of control I really like. With the Kangaroo attached it also adds a fail-safe feature that if it senses a runaway, stall or other voltage problem it tries real hard to shut down. It's actually happened a couple times for me already and may have saved my butt.
B9 deserves the very best!
Well, I've been working at it for hours and there's no way I can fit the V4 into B9's brain. It's just too high. Uncased, sitting naked and with servo cables plugged in, it's a whole 3/8th"higher than the V3. This might as well be a mile as now the lid of the brain sits almost 1/2" above the ledge
Now I'm going to have to run about 5 feet of servo cable through a bobing and rotating section to get to the nearest ezb. Good thing I only need to run 3 cables. I'm considering fitting one of my old V3 ezb's with WiFi because I know that will fit.
Very cool, I wish they had let this guy finish this.
They should pay him to do one episode to see if fans love it.
I think it looks great.
Very cool looking B9 ! Mine is a never ending work in progress. I have made everything from scratch, drives my wife nuts with the amount of time I spend in the basement.
I ordered the Moebius model today. I alternate between the B9 and the R2 Droid as far as working on them.
With the B9 being so large, I'm planning on running him on a remote with very little logic on board. Using the EZB for the R2.
Regards,
Don
Are you a member or the R2Builders club?
Our club makes great droids!
Good luck on animating R2. There are a few members here that have built both a mini and full size version. You'll get lots of help if needed. Like I said in your other post; download robot and control examples into ARC and see how things work.
Have fun!
@Dave Schulpius - did you ever get those arms working? Also: you may want to call that Robert robot "Nigel" or somesuch. The owner of the "Name that Cannot be Mentioned" is very possessive. That said, that robot would be *amazing* controlled by EZ-B! I can envision you programming him to mix you a cool drink and bring it to you on the patio - while B9 fans you with a palm leaf...
Dave - WarPig is right. The owner of the rights to the "ball upon ball stacked vertically robot" is very protective.
Yes I'm a member of the B9 Club. Learning the EZ-Robot software, along with C for the arduino to drive the more routine operations in the R2 dome, light and panel doors. I'll have the EZB trigger the other processor, saving the ports on the EZB for more important motor and servo controls.
Also a member is the Astromech Club. This work thing is getting in th way of my hobby.
D
As far as arms; I'm still in the R&D phase. I've made a few attempts but the motors or servos I've tried haven't been able to handle the load. I am expecting a delivery tomorrow from servo City and The Robot Marketplace with more parts. I hopefully now have a design that will work this time. I'll be using a high power windshield washer motor for the main elbow that lifts most the weight and doubling up servos mounted in gearboxes for the wrists. It will all be connected with aluminum channel and angle plates. It's a slow and expensive process. Keep your fingers crossed.
I find that American Science and Surplus has a nice collection of unusual motors and etc.
I have 2 of these motors. Their DC, but, being worm gear design, when power is off they hold their position.
http://www.sciplus.com/p/CAR-SEAT-312VDC-GEAR-MOTOR_49248
Just offering as a possibility, if you haven't lookd at this style.
D
UPS just delivered the motor I do hope to use for the arm elbow. I'm excited give it the lift test. It seems very heavy duty and small enough for my needs.
AME 218-series 12V 212 in-lb RH gearmotor - stubby shaft
Nominal Voltage::12vdc
RPMs: 116 rpm (no load)
Amps at No load: 1.4
Amps at Stall: 21.3
Torque Nominal: 98.235 in-lbs or 11.1 Nm (yep, that's in-lbs not in-oz!)
Shaft: 10mm dia x 22mm long with flat (Servo City sells 10mm clamping and set screw hubs!)
Weight: 2.55lbs
I just need to decide on the best way how to control it now. I'm leaning towards a Sabertooth with a Kangaroo X2 daughter board. However a Sabertooth/Roo combo takes serial commands and I'm not sure how the scripting would work out when combined with the servos I want to use in the wrist and claws. I was hoping to use ARC's Auto Position control but serial commands may make that impossible. I've never played with the Auto Position control yet so I'm not sure how serial commands would mix with that setup. Well, half the fun is giving it a go and finding out.
The specs on those are impressive!
HB-25
(note: I have seen them cheaper than direct from Parallax, as low as $25 each, making them competitive to a Sabertooth, but figured I would link direct to the source for the best information and let you find the best price if it meets your need).
Alan
From the documentation it doesn't look like it has input for encoder or a pot. I think it is more trial and error on the bench. You figure out your servo "numbers" for the position to move the motor to, then treat it like a servo not a motor. So not as sophisticated as a Kangeroo or Roboclaw, but may be serviceable for your needs.
I hate to have you spend $ on it based on my recommendation when I haven't used one yet though. Maybe someone on the B9 builders group or another robot forum has some experience if no one chimes in here.
Alan
Alan
I'm going to have to use the DC windshield wiper motor for the elbow motor because I cant get my servo gearboxes to run in tandem at the same speed. I tried everything but no go. I'll use these gearbox servos for the lifting joint of the wrist. Since it will be closer to the end of the extended arm my test show it will have no problem lifting the needed load at a good speed. I've made a real cool motor mount that fits great. See pics below.
For speed and position feedback on the DC elbow motor I'll try using a Softpot Rotary Potentiometer. These are very thin variable potentiometers. By pressing down on various parts of the rotary dial, the resistance linearly changes from 100Ohms to 10,000Ohms.
SoftPot
I'll be using Actobotics parts from ServoCity to connect everything and for the arm structure.
Arm guts
Here's a few pics of what I have so far:
Here's a few pics of the motor mount I built to hold the DC elbow motor:
And here's the SoftPot:
I am very interested in your progress with this.
@Rex, Glad you're so interested. Thanks. I also noticed that this motor won't hold it's position when powered down even though it's a worm gear motor. With the Kangaroo x2 / SaberTooth 2x12 that shouldn't be much of an issue. The Roo makes the DC motor hold like a servo till it gets a Release or Power Down command. As far as current; Im sure I'll be able to beef up any part of my power grid if needed.
I built a working robot using an Erector set and some electric motors hooked to relays and photo resisters.
The robot searched for light and then followed the light.
This was before people had micro computers to use in robots.
Yes, actually I've been steady working on the arms when I find some extra time here and there.
I do have a design that I have high hopes for that I actually haven't gotten to "work" the way I want just yet. I've been building prototype designs and finding what works and what doesn't. This has been a painfully slow and somewhat expensive process.
I finally had success with what I thought was my final design as I had my shoulder DC motor working nicely. I want it to lift up to seven pounds at a length at about 2 feet. My setup included a 12 vdc AME wormgear motor powered with a 12v/12 amp power supply controlled through a 2x12 Sabertooth and Kangaroo x2, using a flat SoftPot for feedback and being controlled with EZB V4 through the Uart port.
I was able to get a successful autotune with the Kangaroo/Sabertooth motor driver and was able to lift about 80% of my wanted weight. Then when I added the final rubber outer skin of the arm and a little ballast weight for overhead wiggle room but my sabertooth/Roo started to have power brownouts. Damm!
I reweighed everything and plugged this and my measurements into a neat little Robot Arm Calculator I found on the Society Of Robots web site. I found I was over the torque my 12 amp setup would be able to handle. In fact the motor I wanted to use stalls at 93 Lbs per inch and will pull 30 amps when this happens. The calculator says I need a motor capable of lifting 118 lbs per inch. Damm! Back to the drawing board. I needed to find a motor capable of lifting more than 118 lbs per inch, more power and another motor controller capable of handling all this.
What I came up with is an AME-226-series 12V Utility Gearhead Motor at The Robot Marketplace. It will handle my load and to lift my max overhead wright of 118 lbs per inch. When doing this it will be pulling about 30 amps. In theory to get this to play nice together I also had to get a 12vdc-30 amp power supply and a Sabertooth 2x32 to handle this little monster. The only thing is that this new motor is a little bigger than my old motor so I just now started making a new motor mount for it.
Here's the stuff:
AME-226 Motor
SoftPot
Sabertooth 2x32
That's where I am now. I can't test run this new setup till I get everything rebuilt and mounted. Sadly I don't see that happening for at least two more weeks considering my "real life" commitments. *tired*
One great thing about this upgrade; The Sabertooth 2x32 has an extra set of power and control ports. I can use a power resistor across the power feed and the extra power output ports on each channel as a voltage clamp and I don't need a battery and the extra circuitry to dump the regen power because I'm using a power supply instead of a battery. This is great as I'll be able to simplify my whole setup! Here's how it looks:
This would gear down the motor even more.
This way the motor could handle the weight of the arm and then some.
Amazing how much time is gobbled up with trial and error. Glad to hear its moving forward. I'm using a compound drawer slide, pulley system and cable to extend the arm and retract it. Finally proved a working prototype, now just have to finalize it and build it times 2. Rough part is coming home, it's already dark out, and feels like midnight. Not easy to go down to the workshop.
Regards, Don
@Don, I haven't gotten to designing how to extend and retract the arm in and out of the torso yet. However I have given it a lot of thought though because it's part of the overall design for how large the mounts and motors can be and where to position them. As I'm sure you know there is very little room to work with here. I'm planning on having two joints in my arm; an "Elbow" and a Wrist". I guess the Elbow joint could also be considered a shoulder joint on a regular robot. Very rough measurements that I'm pulling out of my head right now are:
Distance inside of torso (Front to back) for length of motor and linear movement: 15"
Distance from arm hole to where the arm's elbow will be a full extension: 9"
Distance from arm hole to rear edge of claw wrist ring at full extension: 21"
Distance from elbow to rear edge of claw's wrist ring: 11"
Distance from rear edge of claws wrist ring to wrist joint: 5"
Distance from side to side inside rubber arm to fit all this into and pass in and out through: 5"
As you can see it's all very tight but doable if everything is sized just right. As far s the mechanism to move the arm in and out I had considered a pulley system but am leaning to a Gear Rack of some type:
Then for the slides to save room I'm considdering consealed undermount deawer slides. They would allow me to remove the arm like a kitchen drawer and are rated at 75 lbs:
Concealed Drawer Slides
Do yo have any pictures or video of what you've built yet? I'd love to hear and see more! What are you using for motors and feedback?
Here's my 1st attempt at a elbow joint that failed. As I said in an earlier post the motor, controller and power supply are to weak. I'm upgrading now and am very hopeful:
I had to build three different prototypes with three different motor types before I settled on a final design. If you page back you can see the first two. Neither were strong or stable enough.
In a few days I'll post a list of the DC motor, feedback, motor controllers and servos I finally ended up using. However until then I will say I used a different, stronger version of the AME 226 motor shown above, with the Softpot also shown for feedback for the elbow joint. The Softpot is feeding info to a Kangaroo X2/ Sabertooth 32X2 motor controller. Two mini servos for the claw and wrist release. One standard servo for the wrist horizontal and a servo in a servo City 7:1 ratio gearbox moving the wrist up and down joint. I'll do another video in a few days with the skin off showing in more detail everything that powers and moves this 7 pound monster. Until then here's a peek at this EZB controlled B9 arm. Enjoy..............
I had used two of the 212 in/lb torque motors from RobotMarketPlace
for the previous Questor's drive train and they are definitely powerful.
I was talking to someone the other day about extending and retracting the B9 arms and he said something about trying power seat mechanisms used on car seats. New ideas to ponder I guess.
As usual Dave I follow your progress with great interest. Keep up the good work!
Will, yes the linear rail is currently at the top of my list right now. Just need to find a strong and low profile one that will work. For a rail system to work however I think the rails will have to run from the rear of the torso and extend out of the arm holes almost 7 inches or more. The sliding platform would have to attach to the bottom of the elbow motor mount so the fourarm and claw would extend forward over the edge. Here's the aprox weights and mesurments :
Elbow motor and mount: 4 lbs
Fourarm and wrist servos: 1 lb
Wrist donut and claw section: 2 lbs
Rubber skin: 1 lb.
Overall length: 21" from tail of motor to front of wrist donut. (elbow motor mount is 7" long x 3" wide x 4" tall). That only gives me about an inch below the motor mount for a rail and platform and still fit into the rubber arm and torso arm hole.
Like I said above, tight tolerances everywhere . *eyeroll*
May I say again, Absolutely Amazing work!
I had a thought about retracting and extending your arms, I recently ordered a high speed 4-in-stroke high speed linear actuator from the Robot Shop for my Bubble lifter mechanism, I don't have it yet though. They are available in from 4" to 18" strokes and three different speeds and lifting forces. I don't know if the actuator will fit within the limited space you have. I was thinking to mount the arm mechanisms on high quality draw slides and slide everything with the actuator. They have built in limit switches but no feed back for position which means they will be either fully retracted or fully extended, but this may be OK for the extension of the arms. With my bubble lifter 4" actuator I plan to mount external limit switches and a 4" linear slide potentiometer to the side of the actuator so I can control it with a Sabertooth and Kangaroo.
Here is the link to the actuators
http://www.robotshop.com/en/catalogsearch/result/?q=high+speed+Tubular+Actuators&order=stats_sales_order_count&dir=desc
There is also another company making high speed actuators called Progressive Automations and here is a link to a YouTube clip showing one in action
Any how, just food for thought.
Again, Simply Amazing job on your B9 arms, a work of Genius
Steve Neal
www.servocity.com/html/flange_nuts.html
EDIT: I see @steve.neal had posted same video about Progressive
Click To Watch Video
9 inches per second
www.progressiveautomations.com/actuator-tubular-high-speed-linear-actuator-stroke-size-2-force-11-lbs-speed-905sec-p-102.aspx
www.ebay.com/itm/Linear-Actuator-High-Speed-Stroke-10-Force-11-lbs-Progressive-Automations-/280873887546?pt=LH_DefaultDomain_0&hash=item416563673a
How about the rails used on CNC machines as your rail?
Something like these (two per arm): http://www.ebay.com/itm/2-Set-10-inches-long-LINEAR-RAIL-SHAFT-ROD-20mm-with-4-SBR20UU-cnc-router-/400834421180
However I do think I've found the answer. Igus Plastics makes low profile and very strong linear slides and guides in their Drylin line. I'm actually using one of their plastic bearings in the rear pivot joint of my elbow motor mount. It's rated to handle the weight and stress of my elbow and run consistently for over 5400 hours before it wares out. That's like the elbow mount going up and down constantly for over 3/4 of a year. It makes the pivot point of the motor mount glide like ice on ice.
igus® linear guides
I have high hopes for this product! I'll be looking into finding the proper size to support my load to see if it will actually work.