How to make an easy yet strong robot arm
HTML-код
- Опубликовано: 18 апр 2022
- I'm designing and 3d printing a robotic arm to hold a camera and make graceful camera movements.
In this video I 3d print the shoulder and elbow joints and make initial trials with a camera!
These are the supplies i used:
NEMA 17 stepper: amzn.to/3gtK3Qd
Arduino Mega: amzn.to/3nx3Oui
5mm steel rod: amzn.to/3HMBGeV
Thin wall bearings: amzn.to/3gtYLqG
and: amzn.to/3H7JPZY
1.75mm PLA+ filament: amzn.to/3HvzVT5
The PTFE grease came from eBay - search for "50g silicone PTFE grease" 
Наука
Awesome job on the double cycloid design! Very smooth, would be neat to have a desktop robot to help in the lab!
You're going to have a ton of subs in no time, your videos are so excellently done
I appreciate that!
Absolutely fantastic work! I'm so grateful for your willingness to share this. I'll be sure to include you in the acknowledgements of my work and give you a glowing recommendation.
Brilliant Work, I was struggling with building my robot, few of changes you mentioned and did in your robot will help me out fix things !! Thank you so much !!
That is a great application!! So cool to see it works out with no jitter and super smooth movement!! 🤩
thank you! I’m really pleased with it so far.
Amazing. Well done
Bloody Brilliant.
excelent work, congratulations. 👏👏
Great project
Looks great, love the smoothness. and I've printed a few of the bearings at about 66mm OD. They seem fairly decent. thanks
Thank you! I'm glad the bearings are working for you - I'm going to use them in the next video for the base of this arm.
Amazing ! Great ingenuity and craftmansship ! Well done ! I ordered my brushless motors a while ago. Still waiting to receive them. I plan to use them in a shoulder joint for a robot arm attatched to a mannequin. Can't wait to start.
Thank you! I really want to see this arm of yours...
very nice, looking forward to the next video
Thanks!
Nice work! I'm not 100% sure on this, but I think you could reduce some of those subtle high frequency vibrations in the footage by using drivers with higher microstepping values. At this point some of the TMC drivers capable of 1/256 microstepping are only a couple dollars each.
Hmm, Interesting idea, thank you very much! I'm currently running at 1/100 microstepping, but I do have a spare TMC driver lying around so I'll give you idea a go at 1/256 and see if it makes a difference. If it works I'll give you a shout out in the next video.
I thought that the camera made you invisible, but blown away, is also an explanation...
Darn it! You uncovered my secret invisibility research!!!!!
This looks fantastic so far! Congratulations! I'd love to know more about why you went with this design over the eccentric cycloids
Thank you! While the eccentric cycloids were a lot simpler than the traditional cycloid drives, i found that for the same amount of torque they required a larger size and I wanted the joints to remain compact. I'm thinking of using the EC gearing for the base, and can definitely find more uses for them in the future.
It's great to see your robot arm working! I've found that a ball joint adapter works well for mounting the camera, as you can rotate it separately from the end of the arm.
Oh, nice idea!
Well done - very nice ideed!
Thank you very much!
Wow, that's smooth. I thought it was impossible to make stiff gearboxes with fdm. Great work with circumventing material limitations :D
Thanks - yeah, I was impressed with how stiff the gearboxes were too! Bizarrely, I think most of the wobble is coming from the non-gearbox components - time to stiffen them up!
After running your first cycdiol driver, I decided to make this robot arm. But I couldn't run the cycdiol driver. I've been watching this video 10 times trying to find where I missed it. greetings from Türkiye
Awesome! You could try a cheap commercial drone camera gimbal as the wrist (hold a GoPro or similar) or make your own larger one. While auto stabilizing, most can be manually controlled with normal servo signals (pwm) simultaneously.
That is a genius idea! I’ll see if I can find one!
Hi, what a wonderfull job, so many people on the internet did a bulky job with tons of modifications ans your's works fine by staying simple. Mabe you could use carbon square shaped tube for axis. It could save some weight to release your cicloidal drives. Plus by multiplying your fixing points you should reduce the mobility of the camera. A good trick is to use 2 springs locked with the neck stand
Thank you! that's a great idea - I hadn't thought of carbon fibre - I'll see if I can source some!
So we'll thought out!
Nicely done. Sub'd, liked and looking for more.
Amazing video, thanks for sharing! Do you intend to do a video about the code you've made to control the steppers?
thank you. I'm definitely going to cover writing the code for this!
Probably focus more on the code now since it is built. We could give some pointers maybe on smoothing out that motion and reducing the harmonics.
smoooove😀
great work
like to make one
Instructions not clear.
I accidentally made a thermonuclear bomb. And im now wanted by the FBI.
Hey, fantastic work! I'm an last semester electronic engineer student, and I know a lot of Control Theory, Inverse Kinematics, Direct Kimenatics and Trajectory Planning and now I'm looking to learn how to build from scratch a robotic arm. I would appreciate if you can suggest me any source to learn how to design all this mechanical stuff, I know the basics of Mechanics.
Thank you. I'm afraid that I don't have any magic "how to design a robot arm" sources. I'm trying to do it by starting small and iterating through ideas - each time I do a project I make mistakes and the next version is better. There are loads of resources online for specific problems (for example, how to control stepper motors), but nothing I can point you to for a comprehensive step by step guide.
I'd suggest that you take a similar path to what I've been doing - first start with something really small and simple and then approach each component in turn, getting more complex as you learn more.
I'm happy to help if you have any specific questions or want any pointers. Good luck!
Well, obviously now the camera has to be mounted in an actual motorized gimbal, keeping the center of mass of the camera aligned to the tilt and pan axes. I think I have seen some neat RC servo modding hack recently that enabled them to be used for high precision tasks.
Good call! on the Gimbal - maybe I could use that off the arm too I'll check out the modded servos and see if I can use them
good
you should have put the second motor next to the first one on the bottom and use a chain to move the forearm, so the main arm wouldn't have to carry the forearm's weight.
Where are the STLs. I have a 4 dof arm, but I like your design using double cycloid gears. I have all the parts already (surplus from other projects), but no 3D files.
Any chance you will share the 3d file?
I'd be really curious to see if the eccentric cycloidal designs introduced any less vibration. I agree with IV projects on the microstepping probably introducing the most vibration, but we could always test that with a high speed accelerometer at the end of the arm 👀
hmmmm, accelerometer. now there's a thought. I had put the vibration/wobble down to the flexibility of the materials, but if I used an accelerometer I should be able to prove it one way or the other. I'm guessing that if it is the microstepping or cycloidal drive then I should be able to graph the output of the accelerometer and see the frequency of the oscillations match a harmonic of the number of (micro)steps per second. What do you think?
@@roTechnic I was thinking along the same lines as you- by driving the stepper at a known and constant speed you should be able to calculate the frequency of the microsteps and look at the magnitude of that frequency on an fft graph, and you could compare that to say, the magnitude at the motor rotation frequency, or that of the gearbox output shaft. You could also look at how much the overall acceleration values change when you switch from one microstepping value to the next... this is starting to sound like it's own mini project to me 👀
@@Peter-898 yes, yes it is :D I have a horrible feeling this could be a rabbit hole that leads away from finishing the robot arm. I've done some testing and it now doesn't seem to be related to the motors/drives at all. if i keep the accelerations low, the wobble goes away, so, I'm thinking that it is the momentum of the arm/load and the flexibility of the materials involved that are causing it - I'm working on increasing the stiffness/damping of certain portions of the arm to counteract it.
I may well come back to the accelerometer/microstepping in the future though
Awesome work as per usual :D How many degrees of freedom do you intend to have with the robot?
Thank you :D
Well, next time I'm going to add a rotating base taking it up to 3 DOF. and then I'm thinking of a servo controlled camera mount for 4DOF. Then I guess I'll have to see if I need more to get those smooth camera shots I'm after! I think I'll probably need a roll axis on the camera too. What do you think?
@@roTechnic starting to sound like a proper camera arm that :D cant wait to see the final result! I take it you did all those load tests with this goal in mind?
@@mrteemug5329 yep, while it was fun testing all those gearboxes, seeing them actually do something is really satisfying!
thumbs up for the due self-deprecation for filming in portrait mode. 5:04
Amazing designs but where are the 3d models please
Wonderful project! Btw can it hold something that weighs 6lbs or greater? I've built an eye tracker using open cv and I wanted to use it to automatically move my monitor to where my eyes are and maybe a heavier load may affect the motors performance. Thanks in advance.
Thanks - unfortunately this in limited to about 500g of payload. to lift 6lbs, you'd need much bigger motors, or a LOT more reduction. I'm currently working on a 100:1 gearbox which might be able to lift that much - I'll post a video on it with load tests at some point
@@roTechnic Thanks. Also, what do I need to look for to know the right kind of motor to use for a specific payload? or are there some calculations needed to be done?
Hi, how do i know if a robot arm can push down a heavy lever? Thanks
Hy dude.....aufregend.
Hi! Loved your project!!! If you can share the 3d printing file, it would be great.
it's all at github.com/roTechnic/Robot-Arm
what 3d printer are you using?
Nice work, can you share the STEP or STL files
Impressive mechanisms. Your voice is quite scary, though
Yea, so.. 8 months and this looks like I would LOVE to do it. Drop a small arduino cam with motion tracking just below the camera so it automatically adjusts position based on motion detection.. all you really need is a rotating joint to change x/y positioning so it can actually turn and follow. Are your plans, code and files available?
Hi, yeah, the plan is to bang a motion tracking cam on it eventually! CAD and code are available and linked in the latest video on the arm!
A year later, and I'm wondering how feasible it would be to use something like this as a 3d printer. Replace the camera with a hotend, add some IK, make it understand gcode...no problem! :P
Can someone help me, where can I find the files for this amazin project?
What CAD software do you use for the design?
Fusion 360.
Hope if you can make 6Axis Robot Arm
Just done that in my latest video! If you want, you can take all the credit for the idea :D
Great vid. Why do u want it to be back drivable, out of interest?
If it's backdrivable, I can move it around by hand and "teach" the arm where to go.
Nice work! Could you share the STL files?, I would like to build my own robot arm :D
The CAD files are shared in later videos! Good luck!
W
Can you share 3D design for printing?
it’s not ready for release yet, but I’ll be releasing it once it’s done!
@@roTechnic where can we get it when it is released?
@@jasonng2628 It's available on my github github.com/roTechnic/Robot-Arm
@@roTechnic thank you
Face reveal!!!!
YUP!
why not sell the 3d models in a patreon?
Slow