I don'nt understand half of what you're talking about (all the electronics stuff) but my goodness I find your videos FASCINATING. Also, you know... droids. Who doesn't find those cool? Keep it up!
+Darren Carnall - sixth scale figures, sculpting and painting, I totally agree LOL, I can imagine listening to Einstein would be a similar experience, totally bewildering - but amazingly interesting and fun to follow ;-)
Story Time: The first one of your BB droid videos I watched were the two for your BB-9E, and then I came back to watch all the way through your v1, 2, and 3 videos... Since I knew a v3 was coming eventually, I watched all of the v1 and v2 videos at 1.75 speed and only turned it back down to normal speed for this v3 #1 video. Long story short - I got really used to you talking very fast and now you sound kinda stoned or sleepy. It is funny.
I'm fairly excited to see this happen after watching version 2. You're basically doing some of what I had in mind for my own BB-8 build, but I haven't started it yet. Still saving for a printer so I'm living vicariously through your channel.
You have inspired me man... After watching your videos for one week I purchased a Luzlbot Taz 2 on eBay and will try to build an R2 using your R6 cad drawings but modifying the dome shaped head. I'm planning on becoming one of your monthly supporters and see where these new adventures lead me. Just wanted to say thank you and keep on doing what you are doing.
+Tadd Strange I stand by what I said. I really enjoy his work, but watching the same thing over and over is getting boring. I would understand if later on he wanted to revisit the build and improve, but he just finished two others. You could either have one perfected BB-8 after months or a year. Or you could have eight that have minor changes along the way; but I can also understand why he's doing it. There are a lot of Star Wars fans out there and I'm sure they enjoy these videos. You can't please everyone.
Always impressed and thankful for the way you explain your process and thinking. Watching your vids has completely changed my thinking on what I can do with my printer.
Aw Prime. Lets see this through. It's wonderful to hear your thought process, as well as the directions you took for your problem solving. But from what I see...Your droid's framus may be improperly hooked to his dingus. I'm confident you'll work it out. Great Stuff.
Are you sure that your control issues might be related to the restrictive nature of the Arduino "Wiring" framework? It would be really cool to see you teach people how to learn a more modern software and hardware platform?
hmm... Isn't your flywheel technically illegal? I figured it would be illegal to take money out of circulation... Well, I guess they're still circulating inside the ball sooo... Carry on.
I'm so happy that you started the third version, but, can you paint it normal (clean) this time? Whitout the dirty parts? I think is better. I know that you wanna recreate the one from the movie, but (personally) I prefer the versions without the dirty parts (like Sphero and Hasbro toys). What do you think? Greetings and congrats!!!
SpeedLocked One of the things that occurred to me as he was talking about the single axis stage model is that natural wear would show up the way it rolls on a single axis really clearly (since you'd get a band of wear along the only part of the surface that contacts the ground) and break the illusion that the mechanism was a free-rolling ball.
it just occurred to me that you can use a starter motor from a car and actually make a sphero style bb-8 and not have it drive only on one axis, but truly on all axis. the wiper motor would drive it forwards and backwards with a hard polyurethane wheel, the starter would sit retracted and not drag, then when you wanted to turn it would spin up, extend and turn the ball, the drive being a hard surface would offer limited resistance and it would turn... you just have to replace the rounded cog on the starter with a urethane wheel... job done.. being a DC motor, the starter will spin in either direction.. and they're the same price as wiper motors
+James Bruton (XRobots) place the starter facing backwards do it spins sideways, turning the ball on spot. it's just another wiper motor turned 90 degrees that retracts when not in use.... to do quick tuns back and forth (more maneuverable) you'd need a front facing motor too, make the front turn left and the rear turn left, as starters can't spin opposite directions quickly... they have to spin up to extend against the spring.... to shorten the distance and quicken the reaction time of the starters, place a bushing behind the gear(you would replace with a soft wheel to grip the inside of the ball) so that it's already 99% extended and only needs to travel 2 or 3mm to grip... starters normally retract and extend 20-30cm to clear the engine flywheel.
+James Bruton (XRobots) that is to say, it would spin on the dpit in relation to the drive assembly inside the ball, which would then drive off forwards or backwards along whatever axis it stopped on... (appologies for typos.. i've fat fingers & on a phone atm)
+James Bruton (XRobots) i1159.photobucket.com/albums/p622/Rhet83/bb8/bb8.jpg .. I made a doodle, to work the whole thing would balance on the black drive wheel, and use the head to do it, build a frame off the drive wheel and attack everything to it, if it's not entirely symmetrical a small bit of ballast would fix it... use 4 scales on each corner to build it balanced, then you existing designs to do it.. the hard part... making a ball totally smooth inside and out, to drive on and in... the ball would have to be fairly thick and you pocket screws to join in one big seam
+Phil Verhey Starter motors aren't made for continuous drive, whereas windscreen wiper motors are. Starter motors have a very high torque, but poor heat dissipating ability, so they burn out if engaged for too long.
cant wait to see this one built. version 2 was awesome and I wish I could have one of these zooming around the house at that size :) love your work keep it going.
i think the best way to control it would be finding a way to run 2 threads. one thread would be your pid loop keeping everything stable, and the second would be the wireless communications. by altering your values in your main stability loop would then change its center of gravity, keeping the loop size small and consistent. and the same would be true about your wireless control loop. it could update as fast as it can update while also maintaining a redundancy for possible signal loss or inaccurate command signals. having 2 threads means they can run at different speeds maintaining the proper speed for the special functions they are preforming ie. bluetooth and inertial measurement. i also think you should have an independent head combining v1 and v2 to make an above average v3 with the proper improvements, of course.
jesus I would love to see a tour of your place and all your gadgets. What are you most proud of? This seems the most technologically difficult. You have to be some kind of electrical engineer, and even on that level it is astoundingly impressive.
12:08 It all sounds pretty complicated to me. Don't you think that you should have the reverse bushing stabilizers and the cybernetic discharge manifolds offset at hmm say 12 degrees?
I think that you need to reduce the amount of motors and the amount of calculations to speed up the processing loop. It may prove better to use two accelerometers again but redesign the processing part using more processors- basically distributing the computing done. The droid loop must definitely be master and call for event updates.
congratulations for all your stuff I have an idea for you why not trying to combine #1 and # 2 and motorising the head above the ball to avoid magnetic coppler and allow more moves of the head ??
Have you considered using a large gyroscope in order to balance it and prevent the swinging that you had in the second version and improve overall stability?
James, If you use lead as ballast in the flywheel, instead of steel based coins, then the magnetometer won't be effected. Also it will be low enough that the head magnets shouldn't influence it either, As the surface field of neodymium magnets is relatively small.
+James Burton I am very impressed with your work. I was wondering for your next project you could consider a full scale and working HC-XD from call of duty black ops 3.
you are truly the king of BB8, i bet you start getting offers from real production company's for movies and TV! I am eagerly waiting for an Utron update! James, would you consider making a mk1 ultron ( the broken Iron Legion robot)
How about putting the batteries at the bottom for weight stability then being able to shift the battery weight perhaps and shifting sled platform from side to side to cause the sphere to lean left or right?
not quite what I meant take a look at the GM envy two wheeled electric self driving robot car it has a battery "sled" that slides forward and back for balancing but if such a "sled" were to slide side to side in your bb8 sphere it would cause it lean left or right what I mean by sled is the batteries slide on a platform that hangs under your central axis like a pendulum at the bottom of the sphere but not making any contact with the sphere at all...
I feel like you've demonstrated that the head could have enough torque to roll (bank/tip) the body as an inverted pendulum. I feel that this would greatly simplify the internal mechanics.
Wouldn't having parallel motors in a similar setup to your version 2 along the width of your main sphere provide the banking effect that you're looking for (with tank controls implemented). It would keep the main mass at the plane of your preference, the overall weight would keep the sphere rotating on a an axis parallel to the equivalent of a central axle, and the conflicting forces of the tank controls would allow for the in place rotation you mentioned at the start of the video. If the motors don't stay in parallel to the equivalent to an axle, you could always attach an actual axle thats purpose would be solely to keep the motors in place. (With each motor attached to the axle independently of one another. Although that might require you to split the force of each motor into 2 inputs each in the configuration aesthetically similar to a compass with each input on either corner to provide stability)
Hi thanks for the videos. I find it extremely interesting particularly the flexible filament you used. May I ask what type of filament is that to create the tyre or the roller? thank you
+Anthony Bowman Looks like I am going to have to break up some of the parts into smaller piece and then acetone weld them together. My print bed is a bit smaller than the TAZ printer beds. Mine is only 200mm x 200mm. Shouldn't be a problem, though.
If you have one motor "driving" the bot..( against the inside of the outer surface; presumably with a wheel making contact), could you achieve the goal of the additional wiper motor(for banking) by making the drive wheel able to turn with a smaller motor (at the contact wheel), such that..."imagine the drive motor as a small car that is able to drive side to side within the fixed track that is your main drive "groove"..?? Its seems like a "servo" motor that changes the pitch/angle of the drive wheel could cause the "banking" effect and shift the ballast weight around "auto-magically" as long as the drive motor is fixed like a pendulum at the point of the "drive axis" you have identified as requiring great torque. If that can work you wont need 3 wiper motors and will have more internal space.
With the previous droid(s), why did you go with bluetooth and not wifi? Was it because bluetooth is easier or uses less power? Look forward to seeing more of this one soon.
why not base all of the movement based in the actual top head part of the droid. E.G. moving the head back and forth using actuators and then using the gravitational pull of it to roll it. then you can have the magnetic frame on the inside to hold the head
what about a large gyroscope instead of the axle shaft based in the middle of the robot. then instead of controlling the head section through the magnet on the shaft, have a large magnet infrastructure in the middle. this would allow the droid to move in multiple directions
Well couldn't you use motors like those out of childs toy electric cars to create a gyroscopic system inside of your robot and have multiple gearing on it , I'm sure it would take a bit of inlaying the belts but I'm sure it can't be too hard plus if your willing to use more plastic in your print along with something like HIPS you can do a multi-layered print in order to make the system work more easily , and you can get most of the components from the local thrift shop , the head could be done with a magnetic layering inside the bottom of the head with simple RC mechanics to use small electro- magnets to pull the head from side to side or front to back , and the magnets will hold it in place , and you could link them to the same controller but have it almost like a Nintendo controller to control the small electro-magnets to pull and push the head where you want it , Just a thought , if its any help cool if not Sorry for the annoyingly long post
Is there a formula you used to find out how each panel should be made. If so can you do a video on how to do the modeling. I would like to try it. I have a creality ender3. Print area is 10 in X 10in X 10. Thanks.
I've really been enjoying this series! I'm a do it yourself kind of guy (carpentry mostly), but I have never dabbled with electronics. I'm really looking to step my maker skills, any recommendations on a good starting point? Also, I'm poor.
Would there be a any advantage / disadvantage to running two motors in the bottom and configuring a control system to vary the drive speed between the left and right motors to aid in steering?
hi james, i wanna build this robot, but i wanna ask, I have the Robo 3D R1 Plus printer, it has a heated print bed with measurements of 8" x 8.86" x 10", my question is, Is it possible to build this version of BB8 with this printer?
+James Burton: Thank you very much for your excellent video, Sir:) I have just a small question: Wouldn't it be easier to control your whole BB8 (or at least the internal calculations) with a raspberry π? I would use a Wifi-Stick for the communication and add a i²c-board to the raspberry..... I know, that they can be pricy compared to other mini-controllers, but I think it is worth the money for the calculation performance and resources you get..... I always used it for my self-build RC-projects and I was very happy about it :) Have a good Weekend and greetings from Germany ~Unknown6656
***** : I can fully understand your point, Sir :) I just thought, that it might be easier to manage all servos, LEDs, communication channels etc. -------------------------- And thank you for uploading such wonderful videos - I absolutely love watching them and your devotion and enthusiasm to your robots and projects is fascinating and catchy :)
Can you make it so that this version is somewhat autonomous that should be really cool or a more hidden remote so that it looks like it is autonomous? That would be really cool!
for the problem with the connection between te controler and bb-8, take a look at Oberserver Disign pattern en.wikipedia.org/wiki/Observer_pattern this is a consept of writing code and i think i woud be usefull in this case (srry for poor grammer)
Great work, following all of the BB-8 versions with great interest. Question, why not use interrupts to handle the comms (controller)? That way only when there's data from the controller it gets processed; this is a very simple approach which will unload the main control loop. The comms interrupt will handle the incoming data and place it in a buffer and then when it gets all the data (new sticks pos, switch toggles, etc.) it lets the main control loop know, prompting the main control loop to react to the set points. Because the main loop gets unloaded this way, it can run at full speed without worrying about user input. To control the timing in the main loop (PID control is time sensitive) a similar approach can be taken but using a timer interrupt to set the time interval. This will greatly reduce any timing jitter, smoothing the control loop's output. BTW, Arduino's interrupts documentation might be misleading when it reads "Interrupts can slightly disrupt the timing of code", it lacks proper context.
Hmmm. Seems like using serial for this has some downsides. Low-level wireless stuff like the nRF24L01 is packet-based, so if you make each packet a complete set of control data it doesn't matter if you lose a few, there'll be another copy along in a few milliseconds. Serial doesn't have packet boundaries and none of the modules I've seen implement flow control to stop you overflowing the buffers and losing bytes of data, which obviously desynchronizes everything irrecoverably from them on.
9 ปีที่แล้ว +1
I'm working on a new project with a brushless DC bicycle motor. I ti offers cheap motor controller for this kind of motors they are easy to program and require no knowledge about motor control. search for TI instaSPIN. The electric bicycle motors come with a build in planetary gear so they are perfekt for something like that. I hope I can release in the next month a tutorial for that.
try using a raspberry pi as the main controller in the Droid. it is faster (i think) then any 8 bit arduino (sorry for the butchered name) and i think it can multi task pretty well. and use the pi zero for the controller with a wifi dongle stuck in the one USB port. i work with the pi a bit and its never really let me down. just saying
Can You Recommend an Appropriate motor for Building a Anaheim stage droid Replica. The Bevel Gear Driven By Belt both appear to be 1:1 Rato But this is where im stuck no clue about working out torque needed or frame pannel weight radious ratio Brushless V Brushed & Planatary v harmonic But My Frame and Pannels Might be heavier i suppect stage droid sphere is Acrylic California Quality Plastics 2 Hemispheres ill be using using club panels and cari christe new 8mm Sphere
I have a 3DS Cubix Trio and just tried to process the cheese02a and keep saying is Out of bounds for print head 1. Does this mean the part is too big for my printer, Is there any work around for this?Sorry if sound stupid as new to printing,
+James Bruton (XRobots) Thanks James, I was so excited as got a 3d printer to use for 2 months, so still learning, this may be out of the realm of this discussion, but how do you cut the pieces up? Guess I will sound real stupid now..sorry all
I've watched all your videos, great stuff ... in this video you mentioned you were using Bluetooth as comms between 2 Arduinos Do you find this just as reliable as the slip ring in the Astromech you built? I am building an R2 and am at the control stage now. I've been pondering the idea of an independent dome with no slip ring and using Bluetooth but I have not found any info online about reliability. Again thank you for such an inspirational channel. I think my next 3d printer will be a Lulzbot based just on what I've seen you do with it.
James I am sure you've answered this many times before. but what slicer do you use? also i'd love for you to make a video just on 3d printing tips and tricks, lessons learned etc. thanks keep up the good work!!!!
I don'nt understand half of what you're talking about (all the electronics stuff) but my goodness I find your videos FASCINATING. Also, you know... droids. Who doesn't find those cool? Keep it up!
+Darren Carnall - sixth scale figures, sculpting and painting, I totally agree LOL, I can imagine listening to Einstein would be a similar experience, totally bewildering - but amazingly interesting and fun to follow ;-)
I can follow what he's saying
Maybe one day I can truly understand what he's saying... For now all I can do is guess...
I'm so happy that you've jumped right into this project instead of postponing it. You're on a roll (pun absolutely intended)!
Story Time:
The first one of your BB droid videos I watched were the two for your BB-9E, and then I came back to watch all the way through your v1, 2, and 3 videos...
Since I knew a v3 was coming eventually, I watched all of the v1 and v2 videos at 1.75 speed and only turned it back down to normal speed for this v3 #1 video.
Long story short - I got really used to you talking very fast and now you sound kinda stoned or sleepy.
It is funny.
I'm fairly excited to see this happen after watching version 2. You're basically doing some of what I had in mind for my own BB-8 build, but I haven't started it yet. Still saving for a printer so I'm living vicariously through your channel.
Never been more excited for a new series of videos on TH-cam!
The analysis stuff is super interesting. Thanks for including it.
Can you make this one has the lighter "thumbs up" that comes out of a flap in the body please?
Space restrictions will mean it's highly unlikely
+LiamFS .. you mean middle finger..
+LiamFS I wuz hoping for a GoPro or remote camera in the eye. At the very least to prove my theory it can't possibly see where it's going.
That was such Disney bad comedy crap. EVERYONE in the theatre cringed at it.
+Mad Cow Rebel went to see it twice, both times people LOVED it. Including at the red carpet premiere.
You have inspired me man... After watching your videos for one week I purchased a Luzlbot Taz 2 on eBay and will try to build an R2 using your R6 cad drawings but modifying the dome shaped head. I'm planning on becoming one of your monthly supporters and see where these new adventures lead me. Just wanted to say thank you and keep on doing what you are doing.
Well, this was fun to watch when he started doing a BB-8. Now I could really go for something new. I'll come back when this isn't the BB-8 channel.
+James Bruton (XRobots) is BB-8 Droid VERSION 2 finished
+Prime92 oooh ho ho. *winces* bet you didn't think james was gonna reply /eh/?
+Tadd Strange to tell the truth I thought someone else would answer I did not know he was going to answer I was so happy when he answered
lol XD
+Tadd Strange I stand by what I said. I really enjoy his work, but watching the same thing over and over is getting boring. I would understand if later on he wanted to revisit the build and improve, but he just finished two others. You could either have one perfected BB-8 after months or a year. Or you could have eight that have minor changes along the way; but I can also understand why he's doing it. There are a lot of Star Wars fans out there and I'm sure they enjoy these videos. You can't please everyone.
Always impressed and thankful for the way you explain your process and thinking. Watching your vids has completely changed my thinking on what I can do with my printer.
i hope you will also make the Thumbs up mechanism that BB-8 did in the Movie.
Aw Prime. Lets see this through.
It's wonderful to hear your thought process, as well as the directions you took for your problem solving. But from what I see...Your droid's framus may be improperly hooked to his dingus. I'm confident you'll work it out. Great Stuff.
Super excited to see how version 3 turns out! Love all your projects but BB-8 is my fave currently.
Are you sure that your control issues might be related to the restrictive nature of the Arduino "Wiring" framework? It would be really cool to see you teach people how to learn a more modern software and hardware platform?
What a fascinating explanation of how to make an idea come to life. Subbed!
hmm... Isn't your flywheel technically illegal? I figured it would be illegal to take money out of circulation... Well, I guess they're still circulating inside the ball sooo... Carry on.
He's not melting them down.
When he decides to eventually tear them down the money will be back in circulation.
+bummer6 Nice wordsmithery K) Same could be said for my mate, money goes into his wallet, somehow it's never his round ;)
sean sears I know, I just really wanted to make that joke. I'm sorry
It's okay, I'll forgive you :P
+bummer6 it's a bb-8 piggy bank
I'm so happy that you started the third version, but, can you paint it normal (clean) this time? Whitout the dirty parts? I think is better. I know that you wanna recreate the one from the movie, but (personally) I prefer the versions without the dirty parts (like Sphero and Hasbro toys). What do you think?
Greetings and congrats!!!
I agree, but I wouldn't let it completly clean. But definitly less weathered then the version 2
weathering could be done all natural, take "it" for a walk to the corner shopper, bit of kick round with mates down the Footy ground, rock climbing
SpeedLocked One of the things that occurred to me as he was talking about the single axis stage model is that natural wear would show up the way it rolls on a single axis really clearly (since you'd get a band of wear along the only part of the surface that contacts the ground) and break the illusion that the mechanism was a free-rolling ball.
it just occurred to me that you can use a starter motor from a car and actually make a sphero style bb-8 and not have it drive only on one axis, but truly on all axis.
the wiper motor would drive it forwards and backwards with a hard polyurethane wheel, the starter would sit retracted and not drag, then when you wanted to turn it would spin up, extend and turn the ball, the drive being a hard surface would offer limited resistance and it would turn... you just have to replace the rounded cog on the starter with a urethane wheel... job done.. being a DC motor, the starter will spin in either direction.. and they're the same price as wiper motors
+James Bruton (XRobots) place the starter facing backwards do it spins sideways, turning the ball on spot. it's just another wiper motor turned 90 degrees that retracts when not in use.... to do quick tuns back and forth (more maneuverable) you'd need a front facing motor too, make the front turn left and the rear turn left, as starters can't spin opposite directions quickly... they have to spin up to extend against the spring.... to shorten the distance and quicken the reaction time of the starters, place a bushing behind the gear(you would replace with a soft wheel to grip the inside of the ball) so that it's already 99% extended and only needs to travel 2 or 3mm to grip... starters normally retract and extend 20-30cm to clear the engine flywheel.
+James Bruton (XRobots) that is to say, it would spin on the dpit in relation to the drive assembly inside the ball, which would then drive off forwards or backwards along whatever axis it stopped on... (appologies for typos.. i've fat fingers & on a phone atm)
+James Bruton (XRobots) i1159.photobucket.com/albums/p622/Rhet83/bb8/bb8.jpg .. I made a doodle, to work the whole thing would balance on the black drive wheel, and use the head to do it, build a frame off the drive wheel and attack everything to it, if it's not entirely symmetrical a small bit of ballast would fix it... use 4 scales on each corner to build it balanced, then you existing designs to do it.. the hard part... making a ball totally smooth inside and out, to drive on and in... the ball would have to be fairly thick and you pocket screws to join in one big seam
+Phil Verhey Starter motors aren't made for continuous drive, whereas windscreen wiper motors are. Starter motors have a very high torque, but poor heat dissipating ability, so they burn out if engaged for too long.
Good luck James, this looks really exciting!
cant wait to see this one built. version 2 was awesome and I wish I could have one of these zooming around the house at that size :) love your work keep it going.
4:35 that extremely serious expression as you debunked myth. Amazing! 😂
Keep being awesome, James!
i think the best way to control it would be finding a way to run 2 threads. one thread would be your pid loop keeping everything stable, and the second would be the wireless communications. by altering your values in your main stability loop would then change its center of gravity, keeping the loop size small and consistent. and the same would be true about your wireless control loop. it could update as fast as it can update while also maintaining a redundancy for possible signal loss or inaccurate command signals. having 2 threads means they can run at different speeds maintaining the proper speed for the special functions they are preforming ie. bluetooth and inertial measurement.
i also think you should have an independent head combining v1 and v2 to make an above average v3 with the proper improvements, of course.
jesus I would love to see a tour of your place and all your gadgets. What are you most proud of? This seems the most technologically difficult. You have to be some kind of electrical engineer, and even on that level it is astoundingly impressive.
12:08 It all sounds pretty complicated to me.
Don't you think that you should have the reverse bushing stabilizers and the cybernetic discharge manifolds offset at hmm say 12 degrees?
I'm gonna need you to finish this one by the end of the week, thanks buddy.
I think that you need to reduce the amount of motors and the amount of calculations to speed up the processing loop. It may prove better to use two accelerometers again but redesign the processing part using more processors- basically distributing the computing done. The droid loop must definitely be master and call for event updates.
congratulations for all your stuff
I have an idea for you why not trying to combine #1 and # 2 and motorising the head above the ball to avoid magnetic coppler and allow more moves of the head ??
You should integrate a camera in it!
Everyone asks if BB-8 of ABQ has a camera. I don't want a camera. haha :D
Have you considered using a large gyroscope in order to balance it and prevent the swinging that you had in the second version and improve overall stability?
James, If you use lead as ballast in the flywheel, instead of steel based coins, then the magnetometer won't be effected. Also it will be low enough that the head magnets shouldn't influence it either, As the surface field of neodymium magnets is relatively small.
+James Burton I am very impressed with your work. I was wondering for your next project you could consider a full scale and working HC-XD from call of duty black ops 3.
You could also use shaders like Army Painting quickshade to eventually get it dirt.
I really really hope that you're going to put the flame thumbs up in. That would make me so happy :D
you are truly the king of BB8, i bet you start getting offers from real production company's for movies and TV!
I am eagerly waiting for an Utron update!
James, would you consider making a mk1 ultron ( the broken Iron Legion robot)
James I really want to build this but idk if u put out the files for the 3d priter
I same this or BB-9E
How about putting the batteries at the bottom for weight stability then being able to shift the battery weight perhaps and shifting sled platform from side to side to cause the sphere to lean left or right?
not quite what I meant take a look at the GM envy two wheeled electric self driving robot car it has a battery "sled" that slides forward and back for balancing but if such a "sled" were to slide side to side in your bb8 sphere it would cause it lean left or right what I mean by sled is the batteries slide on a platform that hangs under your central axis like a pendulum at the bottom of the sphere but not making any contact with the sphere at all...
I feel like you've demonstrated that the head could have enough torque to roll (bank/tip) the body as an inverted pendulum. I feel that this would greatly simplify the internal mechanics.
Another brilliant post James, well done :)
Hey James, have you considered using old school ball mice setups for the head instead of those caster style wheels?
That is an enormous build plate on that 3D printer, what model is it? Some kind of custom Prusa?
Wouldn't having parallel motors in a similar setup to your version 2 along the width of your main sphere provide the banking effect that you're looking for (with tank controls implemented). It would keep the main mass at the plane of your preference, the overall weight would keep the sphere rotating on a an axis parallel to the equivalent of a central axle, and the conflicting forces of the tank controls would allow for the in place rotation you mentioned at the start of the video.
If the motors don't stay in parallel to the equivalent to an axle, you could always attach an actual axle thats purpose would be solely to keep the motors in place. (With each motor attached to the axle independently of one another. Although that might require you to split the force of each motor into 2 inputs each in the configuration aesthetically similar to a compass with each input on either corner to provide stability)
the Sphero app has been updated with new BB8 sounds from the movie.
Really, I have to try it :)
Hi thanks for the videos. I find it extremely interesting particularly the flexible filament you used. May I ask what type of filament is that to create the tyre or the roller? thank you
Thank you. :)
That's genius James, maybe use an omni wheel for the side leaning motor.
Ok, but will it have a little lighter?
Unrelated to this, but will you be looking into the new series of Robot Wars announced by the BBC?
I'm going to be starting one of these this year as well!
+Anthony Bowman Looks like I am going to have to break up some of the parts into smaller piece and then acetone weld them together. My print bed is a bit smaller than the TAZ printer beds. Mine is only 200mm x 200mm. Shouldn't be a problem, though.
Anthony Bowman hi i would like to build this too do you have the cads? And could you share them with me
I would like the files you used to 3d print that part dome.
If you have one motor "driving" the bot..( against the inside of the outer surface; presumably with a wheel making contact), could you achieve the goal of the additional wiper motor(for banking) by making the drive wheel able to turn with a smaller motor (at the contact wheel), such that..."imagine the drive motor as a small car that is able to drive side to side within the fixed track that is your main drive "groove"..??
Its seems like a "servo" motor that changes the pitch/angle of the drive wheel could cause the "banking" effect and shift the ballast weight around "auto-magically" as long as the drive motor is fixed like a pendulum at the point of the "drive axis" you have identified as requiring great torque. If that can work you wont need 3 wiper motors and will have more internal space.
nice! glad to see you're keen on getting a perfect BB8. Looking forward to this, again! :-D
+Gareth Crispin (MrHolozip) Um...... Why does your comment say that is was written 4 days ago? The video was out 1 hour ago.
With the previous droid(s), why did you go with bluetooth and not wifi? Was it because bluetooth is easier or uses less power? Look forward to seeing more of this one soon.
Would a self balancing robot like in version one(with those omnidirectional wheels) inside a completely smooth sphere work?
It would be nice if you would bring all 3 BB8s to one con when you have finished version 3
would making the flywheel tilt side to side be less practical?
why not base all of the movement based in the actual top head part of the droid. E.G. moving the head back and forth using actuators and then using the gravitational pull of it to roll it. then you can have the magnetic frame on the inside to hold the head
what about a large gyroscope instead of the axle shaft based in the middle of the robot. then instead of controlling the head section through the magnet on the shaft, have a large magnet infrastructure in the middle. this would allow the droid to move in multiple directions
why not try and advance their design then?
oh, fair point. that is a lot of money. maybe in the future if the prices of motors decreases then :D
you did a awesome job with this droid
Well couldn't you use motors like those out of childs toy electric cars to create a gyroscopic system inside of your robot and have multiple gearing on it , I'm sure it would take a bit of inlaying the belts but I'm sure it can't be too hard plus if your willing to use more plastic in your print along with something like HIPS you can do a multi-layered print in order to make the system work more easily , and you can get most of the components from the local thrift shop , the head could be done with a magnetic layering inside the bottom of the head with simple RC mechanics to use small electro- magnets to pull the head from side to side or front to back , and the magnets will hold it in place , and you could link them to the same controller but have it almost like a Nintendo controller to control the small electro-magnets to pull and push the head where you want it , Just a thought , if its any help cool if not Sorry for the annoyingly long post
Is there a formula you used to find out how each panel should be made. If so can you do a video on how to do the modeling. I would like to try it. I have a creality ender3. Print area is 10 in X 10in X 10. Thanks.
James great build. I’m curious if your offer BB8 print files in any way? Thank you.
James, what kind and size screws did you use to attach the cheese wheel sections to the rim parts? Sheet metal screws or machine screws?
Very nice to see you appear on the hackaday website :) mucho kudos points pour vous :D
thanks for these awesome build. Are you going to use the same casters for the head rollers ?
In the time it's taken you to build 3 driods I've done practically nothing of note. Im such a lazy slob.
I've really been enjoying this series! I'm a do it yourself kind of guy (carpentry mostly), but I have never dabbled with electronics. I'm really looking to step my maker skills, any recommendations on a good starting point? Also, I'm poor.
Why doesnt you buy a Gyroscope for the tilt-system? Wouldnt that help with the PID-loop?
+James Bruton (XRobots) I heard that afterwards xP. sorry, and good luck!
Would there be a any advantage / disadvantage to running two motors in the bottom and configuring a control system to vary the drive speed between the left and right motors to aid in steering?
Nevermind... the center would have to be split for that to work... I should never watch your videos and think before that first cup of coffee... lol
James what printer are you using as the one I have is pants and cost £££ for the filament plus does not print the big cheese bits. regards Jas
Hey James, could you use one drive motor but make the motor steerable?
+James Bruton (XRobots) if the motor has 360 motion and pwm it could work.
All these technical terms and jargon...I don't know what he's saying but I feel smarter just because I'm listening too it :)
Great Stuff James ! Can't wait for the next Video :)
Is it possible to print the 300mm parts on a 300mm print bed? Or does the bed need a bit space
I was thinking about use a drone gimbal to make a bb8 but in the future.
can't wait for your next video!
James bruton can you make a series on building your own 3d printer please.. ps love your work
Hi! BB-8 keeps getting better! Very cool, these wiper motors! From what car?
are 2 pence coins the best bang for your buck for weight? would lead fishing weights for be cheaper? honest question.
hi james, i wanna build this robot, but i wanna ask, I have the Robo 3D R1 Plus printer, it has a heated print bed with measurements of 8" x 8.86" x 10", my question is, Is it possible to build this version of BB8 with this printer?
+James Burton: Thank you very much for your excellent video, Sir:)
I have just a small question: Wouldn't it be easier to control your whole BB8 (or at least the internal calculations) with a raspberry π? I would use a Wifi-Stick for the communication and add a i²c-board to the raspberry.....
I know, that they can be pricy compared to other mini-controllers, but I think it is worth the money for the calculation performance and resources you get..... I always used it for my self-build RC-projects and I was very happy about it :)
Have a good Weekend and greetings from Germany
~Unknown6656
***** : I can fully understand your point, Sir :)
I just thought, that it might be easier to manage all servos, LEDs, communication channels etc.
--------------------------
And thank you for uploading such wonderful videos - I absolutely love watching them and your devotion and enthusiasm to your robots and projects is fascinating and catchy :)
dame...I can't wait for more
great work I love watching every second of your video s and you my friend have a talent , genius man great work mate keep it going
Do you have any plans to have the 3d models available for us to print so we can build our own BB-8?
Yes it's on Github
Can you make it so that this version is somewhat autonomous that should be really cool or a more hidden remote so that it looks like it is autonomous? That would be really cool!
for the problem with the connection between te controler and bb-8, take a look at Oberserver Disign pattern
en.wikipedia.org/wiki/Observer_pattern
this is a consept of writing code and i think i woud be usefull in this case
(srry for poor grammer)
Do you have some .zip with all files? (3d, components, software, etc...) Thx mate, you are awesome.
amazing i love u XD
I know this might not work but can you combine version 1 and 2
in the background are the hulk budster legs right
Great work, following all of the BB-8 versions with great interest.
Question, why not use interrupts to handle the comms (controller)? That way only when there's data from the controller it gets processed; this is a very simple approach which will unload the main control loop. The comms interrupt will handle the incoming data and place it in a buffer and then when it gets all the data (new sticks pos, switch toggles, etc.) it lets the main control loop know, prompting the main control loop to react to the set points. Because the main loop gets unloaded this way, it can run at full speed without worrying about user input.
To control the timing in the main loop (PID control is time sensitive) a similar approach can be taken but using a timer interrupt to set the time interval. This will greatly reduce any timing jitter, smoothing the control loop's output.
BTW, Arduino's interrupts documentation might be misleading when it reads "Interrupts can slightly disrupt the timing of code", it lacks proper context.
Hmmm. Seems like using serial for this has some downsides. Low-level wireless stuff like the nRF24L01 is packet-based, so if you make each packet a complete set of control data it doesn't matter if you lose a few, there'll be another copy along in a few milliseconds. Serial doesn't have packet boundaries and none of the modules I've seen implement flow control to stop you overflowing the buffers and losing bytes of data, which obviously desynchronizes everything irrecoverably from them on.
I'm working on a new project with a brushless DC bicycle motor. I ti offers cheap motor controller for this kind of motors they are easy to program and require no knowledge about motor control. search for TI instaSPIN.
The electric bicycle motors come with a build in planetary gear so they are perfekt for something like that.
I hope I can release in the next month a tutorial for that.
try using a raspberry pi as the main controller in the Droid. it is faster (i think) then any 8 bit arduino (sorry for the butchered name) and i think it can multi task pretty well. and use the pi zero for the controller with a wifi dongle stuck in the one USB port. i work with the pi a bit and its never really let me down. just saying
Tito Adesanya raspberry pi all the way!
Can You Recommend an Appropriate motor for Building a Anaheim stage droid Replica. The Bevel Gear Driven By Belt both appear to be 1:1 Rato But this is where im stuck no clue about working out torque needed or frame pannel weight radious ratio Brushless V Brushed & Planatary v harmonic
But My Frame and Pannels Might be heavier i suppect stage droid sphere is Acrylic California Quality Plastics 2 Hemispheres ill be using using club panels and cari christe new 8mm Sphere
update now can i keep yore VERSION 2?
I have a 3DS Cubix Trio and just tried to process the cheese02a and keep saying is Out of bounds for print head 1. Does this mean the part is too big for my printer, Is there any work around for this?Sorry if sound stupid as new to printing,
+James Bruton (XRobots) Thanks James, I was so excited as got a 3d printer to use for 2 months, so still learning, this may be out of the realm of this discussion, but how do you cut the pieces up? Guess I will sound real stupid now..sorry all
Why does the command update rate need to be the same as the inner loop in the droid.
I've watched all your videos, great stuff ... in this video you mentioned you were using Bluetooth as comms between 2 Arduinos Do you find this just as reliable as the slip ring in the Astromech you built? I am building an R2 and am at the control stage now. I've been pondering the idea of an independent dome with no slip ring and using Bluetooth but I have not found any info online about reliability. Again thank you for such an inspirational channel. I think my next 3d printer will be a Lulzbot based just on what I've seen you do with it.
Can I have the 3d files I want to do also something like this but I am not so good at 3d modeling?
Thanks
I may be an idiot so please forgive me in advance but can someone explain how exact the flywheel works in the version 2 droid
Please show some footage of it rolling around more
woooooo! I'm excited for the v.3
James I am sure you've answered this many times before. but what slicer do you use? also i'd love for you to make a video just on 3d printing tips and tricks, lessons learned etc. thanks keep up the good work!!!!
Cool. I opted for the cheaper ctc dual, just getting started with her but so far so good! Thanks
I wish I could have a Bb8 as my friend...
Also question. Can you make a BB8 that can respond to certain words and sentences?? That would be so cool!!!
What's the song at the end of your videos?