I've been watching your videos and have been a subscriber for a long time now. I haven't logged in UA-cam for a while now but I really wanted to comment. I've been waiting for an update on your droid and seeing the weeks go by, I was starting to lose hope. But boy have you been busy! Keep up the good work! It's really inspiring. I also really like how you show the failed attempts too, with the working ones afterward. It's good to see and know that engineering is trial and error until you find the best thing that works.
+SniperFox Thanks for sticking with me. I started this project shortly after the first Star Wars trailer many months ago but for a while it went slowly because I was determined to get out weekly videos too. But that kept me from working on this project. So I decided to just focus on this one only until it was done. That was the reason for the delay. I might try that approach for a while and see how it goes. There shouldn't be any more 2 month delays, but there will be a lot of missed weeks while I focus more on whatever project is the current one. I'm glad you hung in there.
Very interesting to watch your thought processes along the way, and also your choices of materials. Thanks for a very entertaining and instructive video!
Amazing build! Imagine all the mass was at the bottom of the ball internals, and the ball weighed nothing. Then there wouldn't be any wobble on start and stop. So the lower you can get your center of gravity, the better. maybe take apart the battery packs to distribute the cells lower? This would also allow the motors to be further down.
+Matthias Wandel Thanks! I remember thinking about taking apart the batteries back when I was first seeing how things could fit together using Blender, but I was afraid to take them apart. Too late for this version, and still other options to try for stability, like accelerating and decelerating and using a gyro board I bought, but I'll put it back on the list. I'm sure I can dig up some sources out there for what to expect when taking them apart.
Thanks! If it works well enough then I plan to take it to the Ottawa Maker Faire this year. I also take it to various robot and maker meet-ups. We also have an annual Arduino day I'll take it to.
you are a genius.. i have seen many making videos.. but they are kinda bookish.. with confirmed results, but you actually showed the problems which we can face and how to overcome it technically
+verdatum Thanks! Given the wait, I made an extra effort to make it worth it. Glad to hear it was. And editing was around two and a half weeks... well, while also doing the animation, pick-up shots, adding more stuff for completeness sake, ...
+goohz has a channel No worries, it's the first time anyone's asked. I use Sony Vega Movie Studio 11 for editing. I use Blender just for the 3D modelling and also 3D animation.
You bb-8 looks great. Great to see you after a long time. Here is an advice for you. Use those wheels whose diameter is equal to the radius of the globe, then adjust the distance between your wheels so that the whole circumference of the wheel touches the globe. I mean not only the bottom of wheels but the whole circumference of the wheels should touch the ball. Then it will work like the real bb-8 works. Now if you will rotate the wheels the globe will rotate and not the whole internal circuit and machinery.
+harsh bawari Thanks. If I understand you right, that's equivalent to the axle approach, where an axle runs from one side to the other side with motors along the axle to turn it - though having wheels like you describe would be a much more fun way of making it. As for how the real BB-8 works, the movie used 7 or 8 different props along with some CGI. The stage version that was built after the movies was made and is used for promotional purposes behaves as if it is an axle drive though there is another design with a track inside that would also behave that way.
It's coming on really well, I'm looking forward to seeing how you fibreglass it. I should imagine creases might be a problem although a bit of sanding will sort it out. 👍
+Green Silver Thanks. I'm looking forward to seeing how I do it too! I do plan to fiberglass the new globe before I take it apart. I'm even considering ways to embed alignment pins into it. I had one alignment pin in my existing one at one time and it made putting it together much easier.
+Sean Bensadon I haven't found a cheap plastic globe. But are you sure they wouldn't deform much? How thick are the globe's walls? How heavy is your drive system. If you're doing an axle drive then I'm guessing the stresses aren't the same. The drive system I have is very heavy and puts a lot of pressure at the wheels. Having said that, I'll be covering mine in fiberglass to get the rigidity, hopefully.
Kool! I'll definitely check in. I've been getting a little more involved again with social media like snapchat, twitter, twitch, instagram, and tumblr. I haven't made a patreon account yet, but likely will, in the future. I know that you and James have one. I've been trying new things and making new friends in the process lately and soon I hope to be able to collaborate more with others on different projects. Thanks again for being a friend! I'll be in touch! Cheers!
I'm sorry if you have already answered this. Are you going to keep the tape to take it apart in order to access the insides and charge the batteries? I think that this hamster ball scheme is "easier" than the other BB8 building methods but the battery access problem isn't trivial. About the wobble, instead of having a constant PWM action whenever you want to move, could you make the action evolve in a ramp-type profile in order to have less acceleration? It may improve the behavior a bit. Balancing systems are hard to control but you could change the structure, if you could control the pitch of the head support structure you could balance the mass on the inside so when it stops or accelerates, it tries to counter act the momentum. Anyway, your work and your way of explaining is amazing.
+Perromocosso Yes, I'm going to keep the tape method for attaching the two hemispheres and for accessing the insides. It's very difficult to tape the two hemisphere's of the plain cardboard globe together though, since the weight of the hamster deforms it a lot. For that reason I took the time get a new cardboard globe and cover it in three layers of fiberglass, to stiffen it. So far, from my video where I show how I did it, you can see it's looking good ua-cam.com/video/se4zfgIx9sA/v-deo.html. There's no visible deforming and I don't need much tape. Once it's painted I can switch to transparent duct tape (Scotch brand works great.) About the wobble, I've been waiting to experiment with methods like you're talking about until I had the fiberglassing done and my new magnets arrive from China. Unfortunately they're late and I think they're not going to make it and I'll have to get a refund and order again from a more local source. So further experimenting is a little delayed. I'm glad you're liking it though, and thanks for the interest.
+RimstarOrg Thank you for answering so fast. I want to start my own BB8 as soon as I finish my exams but I'm too lazy to put and remove the tape. I thought I could make a hole in one of the hemispheres, modify the charging pin to make it harder so I can insert it and then put a cone or something above the charging port so even if I know where it should be, it guides me to know where it exactly is, so I can connect it and then charge the batteries.
It's coming along well! I'm surprised it works that well now, I thought this design would never work too well, congrats on fine tuning it! Making it even better may be difficult without expanding on budget, here are some of my ideas: Something you might consider doing is getting some NiMh C or D size batteries and wiring them up as needed. Or instead a lithium battery. IMR 18650's or the LiPos used for RC models should be perfect (although maybe a little lightweight compared to NiMh). Also why not add a external charging capability? You may also consider replacing the controller with a $20 RC controller, the receiver is a lot smaller and will allow for easier implementation (no need for logic voltage level shifter board) Or what I would try first, is to pull that RC receiver apart and see if you can reverse engineer the RF receiver and maybe modify it to suit your needs. That could reduce the complexity of the design and pontenially more reliable. It might be time to implement a self-stabilizing code to the arduino to reduce the wobble. This would not cost much, requiring probably nothing more than an accelerometer and gyro, but would require some cleverness to figure out how to make the PID control loop. The easiest thing to try first is to add a low pass filter to the controls. I'll make a quick code in khanacademy to share with you. RC motors may also be better suited, their power density is just ridiculous! A brushless DC motor the size of a pop can be upwards to 750W, or about 1 horsepower! AC induction motors of that size are much bigger. But they can be pricy. A stepper motor may also be a good choice for this thing since they are really precise and torquey! A brushless DC motor will probably only need to be 1/2 to 1/4 the size of the brushed motor assemblies.
+Power Max Thanks for all the tips! Most of what you suggest re batteries and RC is in the future plans in some form. I wanted to go el'cheapo first, but I plan on trying out smartphone control with bluetooth since that seems to be the most popular method and the way things are going. I do like the double duty of the batteries for power and their weight. I'm learning that the brushed motors I have, while serving as good mass, are not a great choice (though they're doing the job and one was free.) Your motor summary helps. Thanks.
Here is some example LPF code: www.khanacademy.org/computer-programming/first-order-low-pass-filter/2933914039 I think if you implement this in between the controler and what signal goes to the motors as PWM, you can reduce the wobbles. It is really basic code, and I predict that a really low (not 0) lag factor will do basically nothing, while one too high will make the bot pretty unresponsive and really slow to react to your input.
Great job. What is your background with electronics? Impressive knowledge and troubleshooting skills you have. I'm so not near your level but working on it
+phillips4 Energy Thanks. My background is just learning as I go along. I learned a lot making the boards for this one. Re the skills, the main trick is to do lots of it.
For something like this, fiberglass is veering towards overkill; particularly if you've never done fiberglassing before. I _strongly_ recommend a product called Shell-Shock by Smooth-On. It's a thixotropic polyurethane resin originally designed for hardening the surface of carved polystyrene. It might take some sanding to get the surface spherical, but at least the stuff lets you sand it. Fiberglass doesn't. You have to apply a very thick gel-coat to get a sandable surface.
+verdatum Thanks! I'll look into it. How susceptible is it to cracking? People have been avoiding acrylic balls, for example, for BB-8 droids because it cracks too easily. I wonder if the same would be the case, in your experience, with Shell-Shock.
If it takes a big fall, yes, it might crack. If you want to make it still more durable, I recommend first applying a layer of Smooth-on's "Plasti-Paste II" It is another polyurethane resin, but imbedded with glass fiber. It's got the durability of fiberglasss without having to deal with things like wrinkles in the fabric, or the horrible fumes of polyester resins. But here again, you must be careful not to apply it in excess because it completely wrecks sandpaper. The plasti-paste has the consistency of putty, so you can use it to some extent to smooth out the relief features of a globe. Just practice with a small batch of it on a piece of scrap before moving forward so you understand how it behaves.
+verdatum Thanks for the detailed answer. Even if I do go with fiberglass, I have your warnings to look out for. I'll do some costing. The main reason for my wanting to do this is to make the ball rigid enough to not deform under the considerable weight of the drive system at the wheels. That's why I thought of fiberglass. With strips of fiberglass embedded along the globe's latitudes and maybe some perpendicular to that I figured I get good rigidity. I'll look further into both.
I’m not sure if you had issues with stability since I’m only part way through the video but would 4 Omni wheels in an X with 2 parallel to each other work inside the ball? I know it can drive forwards backwards and side to side etc outside of one bit thought maybe that wheel setup would work better for a bb8 style droid not sure I’m not an engineer (yet) just an idea
Ik I have asked a lot of questions but I’d like to get your thoughts on an issue I’m having with my drive system. Whenever I do a test run, the head of my drive system (the plate with the magnets) seems to be too heavy, causing the system to want to go on it’s side while driving. I added rollers to help stabilize the drive system but it still seems to want to tip over. I’m assuming it’s because I don’t have enough counterweight? For context, my design consists of a hardboard “base” plate with batteries on one side (all AA, didn’t use more heavy-duty batteries like you, although maybe I should have), and motors, rollers, + remaining circuitry on the other side along with four columns of lightweight wood that serve as the “neck” of the system. Do you think I should put more weight on the side of the base plate opposite the neck? Don’t feel obligated to respond to this. Ik it’s a lot and you’re probably really busy. Just figured I’d give it a shot
It does sound like you're too heavy on top and not heavy enough lower down. Of course the minute you do that, your motors may no longer be strong enough and so on. But it's worth a try adding more weight lower down.
RimstarOrg thanks for replying! And yeah, I was kinda worried about putting too much strain on the motors. I think I have an idea for a system where I could add and remove weight easily. Gonna try that and see what happens. Appreciate the feedback!
That's an interesting idea. I don't see any need to vary the intensity though and it would have to be heavier than the equivalent neodymium magnets but it would simplify making the initial adjustments.
Was thinking about using anti-slip spray in my build as well but I’m hesitant since most seem to have crystalline silica in them, which is carcinogenic. Do you think it is safe to use these sprays? I’ve noticed my wheels tend to churn up bits of newspaper from the inside of my paper-mache ball so I feel like the anti-slip coating would get chewed up over time, releasing hazardous silica dust into the inside of my ball.
I don't know about the carcinogenic properties of the spray. Is the concern when spraying or is the concern to do with the coating after it's being sprayed? If the concern is with spraying then do it in a well ventilated area (I did it outdoors). If the concern it with the coating then I've found that the coating does go away somehow with use. It should be fine as long as it's in the ball but just open the ball in a well ventilated area whenever you do open it. Since the coating doesn't last, I'm not sure the anti-slip coating is a good solution after all -- maybe do some testing with a latex or plastic coating instead? I haven't done any further testing along those lines myself though.
RimstarOrg ok, thank you for the super thorough response! It is good to know that the anti-slip layer does wear off over time. I was thinking about using spray-on rubber instead and now that i know that, that might be the deal breaker. Thanks again for taking the time to get back to me!
+Apfel - Chan Thanks! I don't know what strength the magnets are. I just tried with whatever ones I had and used what worked. They are all neodymium magnets.
+RimstarOrg Are you going to add gyroscope/accelerometer inside? That would be easiest way to minimise wobble by software. You'd have to add few lines to code and one tiny board.
+Piotr Bomba I bought the board a while ago for that purpose (adafruit.com's BN0055 breakout board) but first I want to add fiberglass to the ball to make it more rigid i.e. get it working as well as I can as is before figuring out the best way to enhance it. I don't seem to have any noticeable sideways wobble so one board may in fact do it.
RimstarOrg I think that after you make the ball better you should consider adding BN0055. It may not be used to stablize robot while you dont send any commands. Sensor should calculate if angle is stable. If it isn't software should modify PWM values. In my opinion this would make driving much easier. And the other thing. Even if you had sideways wobble you wouldn't be able to reduce it with your configuration of motors.
+Piotr Bomba Actually, now that you bring it up, my plan was to improve the ball, then try acceleration and deceleration and then if that doesn't work well enough (which I don't think it will) then I'll add the BNO055. I'm looking forward to finally trying the board out. And yeah, it's a good thing I don't have sideways wobble since there's no room near enough to the bottom for more motors anyway. That's the curse of making a more portable sized BB-8. I'm jealous of all those making full sized ones. :)
Thanks! I guess if I were to put a shaft through the middle of the ball instead of the hamster/tank drive then I could put arms on either side. That would actually look pretty cool.
Thanks for your interesting videos! My kids and I thank you! There is an interesting revolution occurring in magnetics right now that might improve your design. Have you heard of polymagnets? They are magnets that are manufactured with multiple poles and in varying arrays. I'm not providing a link for fear that your settings might reject my comment but you can google it. I learned about them on the UA-cam channel Smarter Every Day. I think the concept of engineered magnets is mind blowing! I'd love to see you do a video about them.
I have the dimensions that i worked out for the BB-8 that you see in stage shows on my webpage here rimstar.org/science_electronics_projects/bb-8_dimensions.htm The head for mine is around 20cm in diameter.
This is a very nice project and all the information was so detailed. For me , a complete beginner , where do I have to start to have all the knowledge to do these kind of things. I mean what do i have learn in the first place. please answer.... SUBBED btw :)
+Akshay Soma The real secret is to just start making stuff. Start with something simple. Every project adds new skills and the more projects you do the more complex things you can make. Another key if you're making something complex is to take it one step at a time. Also, show your stuff wherever you can. You get feedback that way as well as tips and you get to see what others have done and learn from them, maybe even work with them. And don't be afraid to learn. Every project is made up of individual processes, materials, components. Research each online and by talking to others so that you understand them and can use them well. Don't be afraid to fail, or even work on something you're not sure you can do. You can treat failure as version 1, and a stepping stone to version 2. And if it just doesn't work out, you've learn something that will help later in making something else. Make things that excite you. When I saw the first trailer for the new Star Wars movie I just had to make a ball robot with a head that stayed on. It was just too awesome an idea and too interesting a challenge. But mostly, just do. And welcome. :)
+RimstarOrg Thank you very much for the quick reply.I will soon start working on my first project.... mostly a rc hovercraft because it is relatively simple and I can find parts readily for low price. Already started searching online about it. :)
+RimstarOrg well said. Adafruit shields address many problems, electronic or mechanical gyros in both head and body, tilt sensors, steppers, and conditional logic, programmable magnets with custom behaviors, just buy a decent polycarbonate sphere or you'll never get past the wobbles. I hope you've already purchased the $150, I this it's sphero something similar, small quality toy, can be controlled via "force band" or app. Tiny little thing, but movement is good, they have sound effects and behaviors, gyro/sensor done well, you could just rip it all out, scale some components up. Or at least have something to work with, you're no where close with this. Good luck
INGLÊS Hi I am Brazilian your problem with this design is the low weight exerted by the center of mass that results in desbalanciamento robo ... tip switch engines with more powerful engines brushless
+Alfath Daril Alhajir I've seen such software of course but I'm not familiar with any of them. Maybe someone else here have some recommendations? Anyone?
Been there, tried that. That was my version 1 ua-cam.com/video/tugoaguh0aY/v-deo.html. It sounds easy, but physics is against you. Keeping the head on while having the RC car still be able to turn the ball is hard.
They're just drill motors take from random 12V cordless drills I found at yard sales, thrift stores and online classifieds like www.craigslist.org/about/sites and www.freecycle.org/.
I use 14.4 volt drill batteries. The motors are from 12 volt drills, so maybe I could have used 12 volt drill batteries instead. In case you were looking at the parts list here rimstar.org/science_electronics_projects/bb-8_v2_parts_list_bill_of_materials.htm, I just added the battery voltage. It was missing before.
+nick thijs Done. You can find a parts list for the dual H-bridge board here rimstar.org/science_electronics_projects/bb-8_v2_drive_electronics.htm and for the RC receiver to Arduino converter board here rimstar.org/science_electronics_projects/rc_controller_to_arduino_hack.htm. Note that if I had known what was involved ahead of time and I didn't value the learning experience in designing and making my own boards I would have bought off-the-shelf ones instead. You may want to consider that too.
pls dont tell me your pwm is really at 1khz... If it really is then the brushes of the motors will die fast and you have soooo much heat loss in the core of the motor that they are more like grills than motors
+Gotin Fuklio I wasn't aware of this issue. Thanks for bringing it up. A little reading shows that due to the Arduino UNO pins I'm using, one of my motors is running at around 1kHz and the other at around 500Hz, though I'll confirm with my scope. I'm guessing 500Hz would be easier on the motors. I can change the pins so that both motors are at 500Hz. Regarding directly driving the motors through relays connected to the radio receiver, I'd considered that but I'll be trying accelerating and decelerating the motors for stability and if that doesn't help enough, I'll be trying an inertial management unit board, and other things too.
You can find it on my website at rimstar.org/science_electronics_projects/bb-8_star_wars_droid_v2.htm under the subheading "3D model for BB-8 (v2)". I had the older version which you see in this part 1 video but I just uploaded the newer version (Sept 20, 2017) with the modifications for the newer magnets and head support structure which you see in my part 3 video. Enjoy!
I don't have a component list or detail measurements, and there's nothing I can do now as I on the road and won't be back for another week You can find a 3D model for Blender on rimstar.org/science_electronics_projects/bb-8_star_wars_droid_v2.htm.
At 7:48 in the video, I show how the batteries are just drill batteries and are held in with velcro. So to charge it, I open up the ball by removing the tape and then undo the velcro to remove the batteries and charge them.
I have no idea what's available in India. Often you can get drills for free because people throw them out when the batteries no longer work. Alternatively you can look for motors online, such as here www.mfacomodrills.com/, but I don't have any suggestions for which ones will work.
I'm on the road right now and don't have access to the info, but I'm guessing it's at around $200 by now. I'll try to remember to put up a budget on my website in the next few weeks, somewhere under rimstar.org/science_electronics_projects/bb-8_star_wars_droid_v2.htm. If I don't remember, feel free to remind me.
+Joe B Actually I was thinking of going the smart phone and bluetooth approach at some point. That seems to be the trend these days, otherwise I would do what you're suggesting.
Thank you for doing this. I would like to belive that you are a big inspiration for many young people, and that is very important.
Your builds are always providing innovative ideas how to replicate complex mechanical devices using common ordinary parts. Keep up the good work!
+USWaterRockets I'm surprised myself what I came up with. The construction's much more elegant than I anticipated - though still hard to do.
I've been watching your videos and have been a subscriber for a long time now. I haven't logged in UA-cam for a while now but I really wanted to comment.
I've been waiting for an update on your droid and seeing the weeks go by, I was starting to lose hope. But boy have you been busy! Keep up the good work! It's really inspiring.
I also really like how you show the failed attempts too, with the working ones afterward. It's good to see and know that engineering is trial and error until you find the best thing that works.
+SniperFox Thanks for sticking with me. I started this project shortly after the first Star Wars trailer many months ago but for a while it went slowly because I was determined to get out weekly videos too. But that kept me from working on this project. So I decided to just focus on this one only until it was done. That was the reason for the delay. I might try that approach for a while and see how it goes. There shouldn't be any more 2 month delays, but there will be a lot of missed weeks while I focus more on whatever project is the current one. I'm glad you hung in there.
Fantastic! I honestly wasn't expecting it to be this good! You did an amazing job! Keep up the good work!
You must have the patience of a saint. Very cool build & great problem solving.
Very interesting to watch your thought processes along the way, and also your choices of materials. Thanks for a very entertaining and instructive video!
Amazing build!
Imagine all the mass was at the bottom of the ball internals, and the ball weighed nothing. Then there wouldn't be any wobble on start and stop. So the lower you can get your center of gravity, the better. maybe take apart the battery packs to distribute the cells lower? This would also allow the motors to be further down.
+Matthias Wandel Thanks!
I remember thinking about taking apart the batteries back when I was first seeing how things could fit together using Blender, but I was afraid to take them apart. Too late for this version, and still other options to try for stability, like accelerating and decelerating and using a gyro board I bought, but I'll put it back on the list. I'm sure I can dig up some sources out there for what to expect when taking them apart.
+Matthias Wandel Oh man. I can only pray that I someday make a video that gets a comment like this from Matthias Wandel. I'd probably die.
+RimstarOrg great BB-8 build you are very clever you should take this to conventions and shows!
Thanks! If it works well enough then I plan to take it to the Ottawa Maker Faire this year. I also take it to various robot and maker meet-ups. We also have an annual Arduino day I'll take it to.
RimstarOrg ok great.
Your explanation is very great and it makes me to love electronics more. I love you Sir, you are my role model.
Wow! This is a very cool build! Nice work :) -Ritchie
+SonicDadDotCom Thanks, Ritchie! -Steve
Very nice work for a homemade DIY build! I like how you show your whole troubleshooting process. Great ideas!
you are a genius.. i have seen many making videos.. but they are kinda bookish.. with confirmed results, but you actually showed the problems which we can face and how to overcome it technically
So glad I found your channel!!! Awesome tinkering videos man!
This video must have taken a TON of editing. Great job, and worth the wait!
+verdatum Thanks! Given the wait, I made an extra effort to make it worth it. Glad to hear it was. And editing was around two and a half weeks... well, while also doing the animation, pick-up shots, adding more stuff for completeness sake, ...
+goohz has a channel No worries, it's the first time anyone's asked. I use Sony Vega Movie Studio 11 for editing. I use Blender just for the 3D modelling and also 3D animation.
Brilliant very enjoyable , and observing the amount of work involved looking forward to part two
Awesome build! Can't wait for part 2!
Keep it up!
This video deserves more views!
I don't watch star wars myself but that looks awesome! you are very smart
You bb-8 looks great. Great to see you after a long time. Here is an advice for you. Use those wheels whose diameter is equal to the radius of the globe, then adjust the distance between your wheels so that the whole circumference of the wheel touches the globe. I mean not only the bottom of wheels but the whole circumference of the wheels should touch the ball. Then it will work like the real bb-8 works. Now if you will rotate the wheels the globe will rotate and not the whole internal circuit and machinery.
+harsh bawari Thanks. If I understand you right, that's equivalent to the axle approach, where an axle runs from one side to the other side with motors along the axle to turn it - though having wheels like you describe would be a much more fun way of making it.
As for how the real BB-8 works, the movie used 7 or 8 different props along with some CGI. The stage version that was built after the movies was made and is used for promotional purposes behaves as if it is an axle drive though there is another design with a track inside that would also behave that way.
It's coming on really well, I'm looking forward to seeing how you fibreglass it. I should imagine creases might be a problem although a bit of sanding will sort it out. 👍
+Green Silver Thanks. I'm looking forward to seeing how I do it too! I do plan to fiberglass the new globe before I take it apart. I'm even considering ways to embed alignment pins into it. I had one alignment pin in my existing one at one time and it made putting it together much easier.
This was very interesting and easy to watch :) you described it perfectly and have a nice voice.
Electronic -mechanic and toys great work !
In currently making one and may I suggest using a plastic globe for the main ball. They are thicker and smoother and do not deform much
+Sean Bensadon I haven't found a cheap plastic globe. But are you sure they wouldn't deform much? How thick are the globe's walls? How heavy is your drive system. If you're doing an axle drive then I'm guessing the stresses aren't the same. The drive system I have is very heavy and puts a lot of pressure at the wheels. Having said that, I'll be covering mine in fiberglass to get the rigidity, hopefully.
I love the design. I can design but not so great with building. I love how similar my design is even though yours is much better
It's been a while! But, I knew you'd be busy workin' on this. Great work! That is a lot of progress you've made so far! Keep it up! :)
+TonyOnekaNobY Thanks! I've been posting weekly updates on my social media in case you ever want to check in in between videos.
Kool! I'll definitely check in. I've been getting a little more involved again with social media like snapchat, twitter, twitch, instagram, and tumblr. I haven't made a patreon account yet, but likely will, in the future. I know that you and James have one. I've been trying new things and making new friends in the process lately and soon I hope to be able to collaborate more with others on different projects. Thanks again for being a friend! I'll be in touch! Cheers!
I'm sorry if you have already answered this.
Are you going to keep the tape to take it apart in order to access the insides and charge the batteries? I think that this hamster ball scheme is "easier" than the other BB8 building methods but the battery access problem isn't trivial.
About the wobble, instead of having a constant PWM action whenever you want to move, could you make the action evolve in a ramp-type profile in order to have less acceleration? It may improve the behavior a bit.
Balancing systems are hard to control but you could change the structure, if you could control the pitch of the head support structure you could balance the mass on the inside so when it stops or accelerates, it tries to counter act the momentum.
Anyway, your work and your way of explaining is amazing.
+Perromocosso Yes, I'm going to keep the tape method for attaching the two hemispheres and for accessing the insides. It's very difficult to tape the two hemisphere's of the plain cardboard globe together though, since the weight of the hamster deforms it a lot. For that reason I took the time get a new cardboard globe and cover it in three layers of fiberglass, to stiffen it. So far, from my video where I show how I did it, you can see it's looking good ua-cam.com/video/se4zfgIx9sA/v-deo.html. There's no visible deforming and I don't need much tape. Once it's painted I can switch to transparent duct tape (Scotch brand works great.)
About the wobble, I've been waiting to experiment with methods like you're talking about until I had the fiberglassing done and my new magnets arrive from China. Unfortunately they're late and I think they're not going to make it and I'll have to get a refund and order again from a more local source. So further experimenting is a little delayed.
I'm glad you're liking it though, and thanks for the interest.
+RimstarOrg Thank you for answering so fast. I want to start my own BB8 as soon as I finish my exams but I'm too lazy to put and remove the tape. I thought I could make a hole in one of the hemispheres, modify the charging pin to make it harder so I can insert it and then put a cone or something above the charging port so even if I know where it should be, it guides me to know where it exactly is, so I can connect it and then charge the batteries.
I really enjoy your video thanks!
Awesome build!
It's coming along well! I'm surprised it works that well now, I thought this design would never work too well, congrats on fine tuning it! Making it even better may be difficult without expanding on budget, here are some of my ideas:
Something you might consider doing is getting some NiMh C or D size batteries and wiring them up as needed. Or instead a lithium battery. IMR 18650's or the LiPos used for RC models should be perfect (although maybe a little lightweight compared to NiMh). Also why not add a external charging capability?
You may also consider replacing the controller with a $20 RC controller, the receiver is a lot smaller and will allow for easier implementation (no need for logic voltage level shifter board) Or what I would try first, is to pull that RC receiver apart and see if you can reverse engineer the RF receiver and maybe modify it to suit your needs. That could reduce the complexity of the design and pontenially more reliable.
It might be time to implement a self-stabilizing code to the arduino to reduce the wobble. This would not cost much, requiring probably nothing more than an accelerometer and gyro, but would require some cleverness to figure out how to make the PID control loop. The easiest thing to try first is to add a low pass filter to the controls. I'll make a quick code in khanacademy to share with you.
RC motors may also be better suited, their power density is just ridiculous! A brushless DC motor the size of a pop can be upwards to 750W, or about 1 horsepower! AC induction motors of that size are much bigger. But they can be pricy. A stepper motor may also be a good choice for this thing since they are really precise and torquey! A brushless DC motor will probably only need to be 1/2 to 1/4 the size of the brushed motor assemblies.
+Power Max Thanks for all the tips!
Most of what you suggest re batteries and RC is in the future plans in some form. I wanted to go el'cheapo first, but I plan on trying out smartphone control with bluetooth since that seems to be the most popular method and the way things are going. I do like the double duty of the batteries for power and their weight. I'm learning that the brushed motors I have, while serving as good mass, are not a great choice (though they're doing the job and one was free.) Your motor summary helps. Thanks.
Here is some example LPF code: www.khanacademy.org/computer-programming/first-order-low-pass-filter/2933914039 I think if you implement this in between the controler and what signal goes to the motors as PWM, you can reduce the wobbles. It is really basic code, and I predict that a really low (not 0) lag factor will do basically nothing, while one too high will make the bot pretty unresponsive and really slow to react to your input.
+Power Max Awesome. And super simple. Thanks!
great stuff! glad to see you posting.
Wow ! you put in a lot of work
Good job
That explains the two month delay, this looks like a complicated task to do.
perfect work bro..mind blowing
Great job. What is your background with electronics? Impressive knowledge and troubleshooting skills you have. I'm so not near your level but working on it
+phillips4 Energy Thanks. My background is just learning as I go along. I learned a lot making the boards for this one. Re the skills, the main trick is to do lots of it.
Wonderful video! Thanks!
if who create this i really appreciate and i also follow your steps thank you so much
For something like this, fiberglass is veering towards overkill; particularly if you've never done fiberglassing before. I _strongly_ recommend a product called Shell-Shock by Smooth-On. It's a thixotropic polyurethane resin originally designed for hardening the surface of carved polystyrene. It might take some sanding to get the surface spherical, but at least the stuff lets you sand it. Fiberglass doesn't. You have to apply a very thick gel-coat to get a sandable surface.
+verdatum Thanks! I'll look into it. How susceptible is it to cracking? People have been avoiding acrylic balls, for example, for BB-8 droids because it cracks too easily. I wonder if the same would be the case, in your experience, with Shell-Shock.
If it takes a big fall, yes, it might crack. If you want to make it still more durable, I recommend first applying a layer of Smooth-on's "Plasti-Paste II" It is another polyurethane resin, but imbedded with glass fiber. It's got the durability of fiberglasss without having to deal with things like wrinkles in the fabric, or the horrible fumes of polyester resins. But here again, you must be careful not to apply it in excess because it completely wrecks sandpaper. The plasti-paste has the consistency of putty, so you can use it to some extent to smooth out the relief features of a globe. Just practice with a small batch of it on a piece of scrap before moving forward so you understand how it behaves.
+verdatum Thanks for the detailed answer. Even if I do go with fiberglass, I have your warnings to look out for. I'll do some costing. The main reason for my wanting to do this is to make the ball rigid enough to not deform under the considerable weight of the drive system at the wheels. That's why I thought of fiberglass. With strips of fiberglass embedded along the globe's latitudes and maybe some perpendicular to that I figured I get good rigidity. I'll look further into both.
I’m not sure if you had issues with stability since I’m only part way through the video but would 4 Omni wheels in an X with 2 parallel to each other work inside the ball? I know it can drive forwards backwards and side to side etc outside of one bit thought maybe that wheel setup would work better for a bb8 style droid not sure I’m not an engineer (yet) just an idea
4 Omni wheels would seem like a good approach. Oddly enough, I've only ever seen one person do it, but they had good results.
Ik I have asked a lot of questions but I’d like to get your thoughts on an issue I’m having with my drive system. Whenever I do a test run, the head of my drive system (the plate with the magnets) seems to be too heavy, causing the system to want to go on it’s side while driving.
I added rollers to help stabilize the drive system but it still seems to want to tip over. I’m assuming it’s because I don’t have enough counterweight? For context, my design consists of a hardboard “base” plate with batteries on one side (all AA, didn’t use more heavy-duty batteries like you, although maybe I should have), and motors, rollers, + remaining circuitry on the other side along with four columns of lightweight wood that serve as the “neck” of the system. Do you think I should put more weight on the side of the base plate opposite the neck?
Don’t feel obligated to respond to this. Ik it’s a lot and you’re probably really busy. Just figured I’d give it a shot
It does sound like you're too heavy on top and not heavy enough lower down. Of course the minute you do that, your motors may no longer be strong enough and so on. But it's worth a try adding more weight lower down.
RimstarOrg thanks for replying! And yeah, I was kinda worried about putting too much strain on the motors. I think I have an idea for a system where I could add and remove weight easily. Gonna try that and see what happens. Appreciate the feedback!
marvellous, indeed. Can't wait to see what would happen in the future ^_^
Great work!!
I love it!
Very nice!
What about an electro magnet for the head?
That way you can vary the intensity of the magnet depending upon what you want to do :)
That's an interesting idea. I don't see any need to vary the intensity though and it would have to be heavier than the equivalent neodymium magnets but it would simplify making the initial adjustments.
Was thinking about using anti-slip spray in my build as well but I’m hesitant since most seem to have crystalline silica in them, which is carcinogenic. Do you think it is safe to use these sprays? I’ve noticed my wheels tend to churn up bits of newspaper from the inside of my paper-mache ball so I feel like the anti-slip coating would get chewed up over time, releasing hazardous silica dust into the inside of my ball.
I don't know about the carcinogenic properties of the spray. Is the concern when spraying or is the concern to do with the coating after it's being sprayed? If the concern is with spraying then do it in a well ventilated area (I did it outdoors). If the concern it with the coating then I've found that the coating does go away somehow with use. It should be fine as long as it's in the ball but just open the ball in a well ventilated area whenever you do open it. Since the coating doesn't last, I'm not sure the anti-slip coating is a good solution after all -- maybe do some testing with a latex or plastic coating instead? I haven't done any further testing along those lines myself though.
RimstarOrg ok, thank you for the super thorough response! It is good to know that the anti-slip layer does wear off over time. I was thinking about using spray-on rubber instead and now that i know that, that might be the deal breaker. Thanks again for taking the time to get back to me!
Great job! But I have got a question: Which strength of magnets did you use? I haven't got any technical knowledge so I'm asking you :)
+Apfel - Chan Thanks! I don't know what strength the magnets are. I just tried with whatever ones I had and used what worked. They are all neodymium magnets.
+RimstarOrg thanks :) I will try to build a bb 8 in summer this year :)
I will buy the strongest I will find on the internet :D
+RimstarOrg Are you going to add gyroscope/accelerometer inside? That would be easiest way to minimise wobble by software. You'd have to add few lines to code and one tiny board.
+Piotr Bomba I bought the board a while ago for that purpose (adafruit.com's BN0055 breakout board) but first I want to add fiberglass to the ball to make it more rigid i.e. get it working as well as I can as is before figuring out the best way to enhance it. I don't seem to have any noticeable sideways wobble so one board may in fact do it.
RimstarOrg I think that after you make the ball better you should consider adding BN0055. It may not be used to stablize robot while you dont send any commands. Sensor should calculate if angle is stable. If it isn't software should modify PWM values. In my opinion this would make driving much easier.
And the other thing. Even if you had sideways wobble you wouldn't be able to reduce it with your configuration of motors.
+Piotr Bomba Actually, now that you bring it up, my plan was to improve the ball, then try acceleration and deceleration and then if that doesn't work well enough (which I don't think it will) then I'll add the BNO055. I'm looking forward to finally trying the board out.
And yeah, it's a good thing I don't have sideways wobble since there's no room near enough to the bottom for more motors anyway. That's the curse of making a more portable sized BB-8. I'm jealous of all those making full sized ones. :)
Its amazing!🤩, Do you think you can evolve it someday into an Android robot?
Thanks! I guess if I were to put a shaft through the middle of the ball instead of the hamster/tank drive then I could put arms on either side. That would actually look pretty cool.
Amazing man
nice!
hi, i'm your subscriber, lol)
+William Jakespeare Thanks!
BTW +Лев Бабинцев, I'm subbed to Jake too. Awesome maker!
+RimstarOrg Yes, very cool
+RimstarOrg Thanks for watching!
ok thanks, but how much the rpm or kg-cm. should be ?
eg. 3kg cm. , or 500 rpm
Awesome!
+KmanMeteorProductions Thanks!
Thanks for your interesting videos! My kids and I thank you! There is an interesting revolution occurring in magnetics right now that might improve your design. Have you heard of polymagnets? They are magnets that are manufactured with multiple poles and in varying arrays. I'm not providing a link for fear that your settings might reject my comment but you can google it. I learned about them on the UA-cam channel Smarter Every Day. I think the concept of engineered magnets is mind blowing! I'd love to see you do a video about them.
What size is the head? I’m thinking about making my own design for BB8
I have the dimensions that i worked out for the BB-8 that you see in stage shows on my webpage here rimstar.org/science_electronics_projects/bb-8_dimensions.htm The head for mine is around 20cm in diameter.
This is a very nice project and all the information was so detailed. For me , a complete beginner , where do I have to start to have all the knowledge to do these kind of things. I mean what do i have learn in the first place. please answer.... SUBBED btw :)
+Akshay Soma The real secret is to just start making stuff. Start with something simple. Every project adds new skills and the more projects you do the more complex things you can make.
Another key if you're making something complex is to take it one step at a time.
Also, show your stuff wherever you can. You get feedback that way as well as tips and you get to see what others have done and learn from them, maybe even work with them.
And don't be afraid to learn. Every project is made up of individual processes, materials, components. Research each online and by talking to others so that you understand them and can use them well.
Don't be afraid to fail, or even work on something you're not sure you can do. You can treat failure as version 1, and a stepping stone to version 2. And if it just doesn't work out, you've learn something that will help later in making something else.
Make things that excite you. When I saw the first trailer for the new Star Wars movie I just had to make a ball robot with a head that stayed on. It was just too awesome an idea and too interesting a challenge.
But mostly, just do.
And welcome. :)
+RimstarOrg Thank you very much for the quick reply.I will soon start working on my first project.... mostly a rc hovercraft because it is relatively simple and I can find parts readily for low price. Already started searching online about it. :)
+RimstarOrg well said. Adafruit shields address many problems, electronic or mechanical gyros in both head and body, tilt sensors, steppers, and conditional logic, programmable magnets with custom behaviors, just buy a decent polycarbonate sphere or you'll never get past the wobbles.
I hope you've already purchased the $150, I this it's sphero something similar, small quality toy, can be controlled via "force band" or app.
Tiny little thing, but movement is good, they have sound effects and behaviors, gyro/sensor done well, you could just rip it all out, scale some components up. Or at least have something to work with, you're no where close with this.
Good luck
“So I added more magnets...”
great work!!!!
+Rulof SecondoCanale Grazie!
can I use any other motor , but cheap.
please suggest me, please
INGLÊS
Hi I am Brazilian your problem with this design is the low weight exerted by the center of mass that results in desbalanciamento robo ... tip switch engines with more powerful engines brushless
did you once make a video about how to make iced tea??
Nope. Not me.
awesome!
Nice
O shit, your in Ottawa? cool... I'm here for school
U of O or Carleton? I'm a U of O grad myself, comp sci.
Carleton, studying computer systems engineering
lamcalvin123
Cool. They look like they have a good program.
Always nice to see a local
what is your qualification?
I kinda wish to have the possibility to find the materials.
thanks sir 😊😊😊
YAY BLENDERR!!!
+Beakercraft I'm with you there!
Whats the best software to simulate or design electric circuit ?
+Alfath Daril Alhajir I've seen such software of course but I'm not familiar with any of them. Maybe someone else here have some recommendations? Anyone?
Omg just put a rc car in the middle with a magnet twice as easy 4x as fast yes im lazy but it works nice hes so much easier than r2d2
Been there, tried that. That was my version 1 ua-cam.com/video/tugoaguh0aY/v-deo.html. It sounds easy, but physics is against you. Keeping the head on while having the RC car still be able to turn the ball is hard.
RimstarOrg keep it up i enjoy wating your videos
So a basic RV droid, no gyroscope to stablization....
which motor you have used in your project ,can you give me the link for it
They're just drill motors take from random 12V cordless drills I found at yard sales, thrift stores and online classifieds like www.craigslist.org/about/sites and www.freecycle.org/.
nice
cool!
it took to me 3 months to make bb8
I want to know what are the specs of the power supply you used.Please help
Which power supply do you mean? I used batteries for everything.
I Mean battery specs
I use 14.4 volt drill batteries. The motors are from 12 volt drills, so maybe I could have used 12 volt drill batteries instead. In case you were looking at the parts list here rimstar.org/science_electronics_projects/bb-8_v2_parts_list_bill_of_materials.htm, I just added the battery voltage. It was missing before.
Thanks
Get a rc car and put it in the middle
Can you make a parts list for te electronic components
+nick thijs Done. You can find a parts list for the dual H-bridge board here rimstar.org/science_electronics_projects/bb-8_v2_drive_electronics.htm and for the RC receiver to Arduino converter board here rimstar.org/science_electronics_projects/rc_controller_to_arduino_hack.htm. Note that if I had known what was involved ahead of time and I didn't value the learning experience in designing and making my own boards I would have bought off-the-shelf ones instead. You may want to consider that too.
thanks
Almost all of his video "Hey gang"
pls dont tell me your pwm is really at 1khz... If it really is then the brushes of the motors will die fast and you have soooo much heat loss in the core of the motor that they are more like grills than motors
+Gotin Fuklio and you could directly drive the motors trough a relays connected to the radio reciever
+Gotin Fuklio I wasn't aware of this issue. Thanks for bringing it up. A little reading shows that due to the Arduino UNO pins I'm using, one of my motors is running at around 1kHz and the other at around 500Hz, though I'll confirm with my scope. I'm guessing 500Hz would be easier on the motors. I can change the pins so that both motors are at 500Hz.
Regarding directly driving the motors through relays connected to the radio receiver, I'd considered that but I'll be trying accelerating and decelerating the motors for stability and if that doesn't help enough, I'll be trying an inertial management unit board, and other things too.
RimstarOrg i recommend 20 hz base frequency
can i get link to the 3d model of inside part
You can find it on my website at rimstar.org/science_electronics_projects/bb-8_star_wars_droid_v2.htm under the subheading "3D model for BB-8 (v2)". I had the older version which you see in this part 1 video but I just uploaded the newer version (Sept 20, 2017) with the modifications for the newer magnets and head support structure which you see in my part 3 video. Enjoy!
cool to the bone
hey gang!
Kip kay is that you
omaigod
sir need all components list and detail measurement plz quick
I don't have a component list or detail measurements, and there's nothing I can do now as I on the road and won't be back for another week You can find a 3D model for Blender on rimstar.org/science_electronics_projects/bb-8_star_wars_droid_v2.htm.
1:36 Trust fall
How you charge this
At 7:48 in the video, I show how the batteries are just drill batteries and are held in with velcro. So to charge it, I open up the ball by removing the tape and then undo the velcro to remove the batteries and charge them.
i dont have them
which is available in India 🇮🇳🇮🇳
I have no idea what's available in India. Often you can get drills for free because people throw them out when the batteries no longer work. Alternatively you can look for motors online, such as here www.mfacomodrills.com/, but I don't have any suggestions for which ones will work.
budget plz
I'm on the road right now and don't have access to the info, but I'm guessing it's at around $200 by now. I'll try to remember to put up a budget on my website in the next few weeks, somewhere under rimstar.org/science_electronics_projects/bb-8_star_wars_droid_v2.htm. If I don't remember, feel free to remind me.
damm daniel
Someone already build thisn
+Danyal Ahmad A few people have already built this, though no two were done the same way.
If your going to build something like this, why not just spend a little and buy a good hobby grade Tx and Rx.
+Joe B Actually I was thinking of going the smart phone and bluetooth approach at some point. That seems to be the trend these days, otherwise I would do what you're suggesting.
+RimstarOrg yes, good idea. you already have an arduino, a bt module would be the best imo.
voici une recommandation que j'ai eu après cette vidéo. Je pense à une nouvelle chaine ua-cam.com/video/-MPm36buRwY/v-deo.html
YES FIRST COMMENT