@@stepbystep-robotics6881 By the way, I created a Python app that might help other with their cycloidal drives. Let me know if you want me to share it with you.
This video was a huge help. No one ever mentions the pre-calcs that help it run. Apparently SW has some order of operations issues and the equation straight out of the design sheet everyone has been using doesn't work as pasted.
Thanks a lot for your job and sharing this information. The order of the content is very cleary and eficently. Python + SolidWorks?!! Wow... Every day is a learning experience! Greetings form Mexico!!
@@m_ja1n I think this is pretty much trial and error for 3d printing. But from my experience I will reduce the profile a bit like ~0.3mm to get actual size.
Hi! I'm struggling at it right now. I have problem with the function driven curve which returns unbalanced shapes. Like squished and progressively degenerating. Have you changed something in the function formulas? Thx for reading, thxxxx if you can take some time to reply
@@yannmegard4008 I guess, you are using Inventor as well. The problem is that Inventor and Solidworks use some different expressions. For example in solidworks it is "arctan" and in Inventor you have to use "atan" I can send you a link to the python script I used to generate the equations
@@danielh.8836 Yes, i'm using inventor. I found out for the atan expression, but nevertheless it doesn't work properly. I'd love if you could send me a link to your python script! Thank you so much, it's very kind of you
@@yannmegard4008 somehow RUclips is immediately deleting all my comments with the drive link in it... If you want you can send me your Email and I can try there
@@danielh.8836 Probably they do not want us to exchange information!! I will give you my email again replacing the usual at sign by "(A)" so there little bots do not analyze my email adresses for what they are. There you go: ymeggg(A)gmail.com and alternatively: yann.megard(A)gmail.com .I hope it'll work. Other option: Going back to school, but i'm 52 and my neurons are vanishing very fast!
Gracias por la informacion, entiendo muy poco sobre motores y reductores pero la meta es hacer un brazo robotico y meterle inteligencia artificial. Saludos desde Peru
Nice. nice, nice! I have a self-made CNC, its accuracy is 0.05 mm. Will this be enough to make such a gearbox? Is it better to order on professional equipment? To make, for example, 0.01 mm. I'm thinking of using aluminum bronze to reduce friction.
I see you are using a lot of bearings around the cyclodial disk in your metal version. But you didn't use them in your 3d printed prototype right? Is there a big difference in accuracy or something else? Because I would like to make a 3d printed cyclodial gearbox for robotics application (similar to your prototype) . Would you recommend using these Bearings or could I leave them out like you did in your first version? (would it be a good place to cheap out on?) Are the CAD files for the Cyclodial Gearboxes you made for download somewhere? I really appreciate the content you make!!
there is some research paper mentioned about using rolling components like pin or bearing can give more efficiency (less internal friction) than a solid lobe, you can try check on my description. But for prototype and simplicity, the solid one should be okay.
Thanks for the video, i'm planning to get position from the outshaft, i'm checking the magnetic encoder AS5311 and a magnetic ring. They use the same methode to get position from a rotary table of a handmill
Looks like a pretty slick design! From the shots at the end where you move the arm around by hand, it looks like it backdrives pretty easily; do you have any idea what the backdrive efficiency and starting torque are?
Thanks! I never done the experiment of efficiency before, but for back drivability we need to use the servo that can control stiffness and damping while it’s in position mode (like mini cheetah actuator). I did some simple test on cycloid drive with two links arm before, please check out here ruclips.net/video/CJyCcXob0c8/видео.html
Hello there, Excellent presentation... Can I learn how you do the torque calculation? I want to calculate for my 1 to 30 ( ratio 1:30) project. thank you so much.
Thank you!, ratio is defined by N-1 where N is the number of outer roller, in my video I was using N=20, so I can get 19:1 ratio. Torque depends on your motor spec.
Great job! why are you using two rows of bearings (for each gear). There are some restrictions of using one row of bearings (for both gears? It seems, that cycloidal gears are moving with same speed
Thanks! Using two rows of bearing because there is no a bearing that thick enough for two cycloid disc gear. So even a solid lobe instead of bearings also okay, but with bearing it will give more efficiency.
@@stepbystep-robotics6881 HK0408 bearing will fits your model perfectly. Are you planning to use single bearing for test? one more question - the shafts for holding surrounding bearings were milled with one case with their holder or they can are separately produced and assemblied?
I found it easier to generate the cycloidal profile in solidoworks and then exported it to fusion 360 and the rest of the components were completed in fusion 360.
Your video is very good. Lots of tutorials only cover the cicloidal gear and not the entire drive but yours is very complete!
Wow! Didn't think anyone read my tutorial! I'm glad it was useful dude!
Oh I am glad you found me!
Your doc is very useful for me, it can help me a lot when I started making cycloid drive, thank you a lot for your work.
@@stepbystep-robotics6881 By the way, I created a Python app that might help other with their cycloidal drives. Let me know if you want me to share it with you.
@@omaryounis3849 Great work sir, respect +++
@@omaryounis3849what is this app
I think these are getting more and more popular. It's good to see new videos on them!
Thanks!
-TheRainHarvester
Yes, it's becoming much popular :)
This needs more views! Great explanation. Do a simulation next since that's trickier than the geometry.
This video was a huge help. No one ever mentions the pre-calcs that help it run. Apparently SW has some order of operations issues and the equation straight out of the design sheet everyone has been using doesn't work as pasted.
Thanks a lot for your job and sharing this information. The order of the content is very cleary and eficently. Python + SolidWorks?!! Wow... Every day is a learning experience!
Greetings form Mexico!!
Excellent presentation, thanks for putting this together!
THANK YOU SOO MUCH! I am very new to all this and i have been looking for something like this for sooo long. Thank you soo much for the effort!
You're welcome! good to hear it's useful somehow :)
@@stepbystep-robotics6881 Btw what should we change before 3d printing this? How much tolerances to give?
@@m_ja1n I think this is pretty much trial and error for 3d printing. But from my experience I will reduce the profile a bit like ~0.3mm to get actual size.
Great video...keep up the work
It took me a while but i finally got the method working in Autodesk Inventor as well. Thanks for the video, it really helped a lot!
Hi! I'm struggling at it right now. I have problem with the function driven curve which returns unbalanced shapes. Like squished and progressively degenerating. Have you changed something in the function formulas? Thx for reading, thxxxx if you can take some time to reply
@@yannmegard4008 I guess, you are using Inventor as well. The problem is that Inventor and Solidworks use some different expressions. For example in solidworks it is "arctan" and in Inventor you have to use "atan" I can send you a link to the python script I used to generate the equations
@@danielh.8836 Yes, i'm using inventor. I found out for the atan expression, but nevertheless it doesn't work properly. I'd love if you could send me a link to your python script! Thank you so much, it's very kind of you
@@yannmegard4008 somehow RUclips is immediately deleting all my comments with the drive link in it... If you want you can send me your Email and I can try there
@@danielh.8836 Probably they do not want us to exchange information!! I will give you my email again replacing the usual at sign by "(A)" so there little bots do not analyze my email adresses for what they are. There you go: ymeggg(A)gmail.com and alternatively: yann.megard(A)gmail.com .I hope it'll work. Other option: Going back to school, but i'm 52 and my neurons are vanishing very fast!
Gracias por la informacion, entiendo muy poco sobre motores y reductores pero la meta es hacer un brazo robotico y meterle inteligencia artificial. Saludos desde Peru
Nice. nice, nice!
I have a self-made CNC, its accuracy is 0.05 mm. Will this be enough to make such a gearbox?
Is it better to order on professional equipment? To make, for example, 0.01 mm.
I'm thinking of using aluminum bronze to reduce friction.
thanks fot the content, I really appreciate it
I see you are using a lot of bearings around the cyclodial disk in your metal version. But you didn't use them in your 3d printed prototype right?
Is there a big difference in accuracy or something else?
Because I would like to make a 3d printed cyclodial gearbox for robotics application (similar to your prototype) . Would you recommend using these Bearings or could I leave them out like you did in your first version? (would it be a good place to cheap out on?)
Are the CAD files for the Cyclodial Gearboxes you made for download somewhere?
I really appreciate the content you make!!
there is some research paper mentioned about using rolling components like pin or bearing can give more efficiency (less internal friction) than a solid lobe, you can try check on my description. But for prototype and simplicity, the solid one should be okay.
Are the scripts you mention available anywhere? I didn't see them linked on the cycloidal drive section of your site, unless I just missed it?
Awesome work!
awesome work thank you
Thanks for the video, i'm planning to get position from the outshaft, i'm checking the magnetic encoder AS5311 and a magnetic ring. They use the same methode to get position from a rotary table of a handmill
Thanks for the info :) I will have a look on it for a new version.
great elegant design. i imagine this is not cheap to get made. do you have your own cnc mill?
Could you share the python script for the cycloidal disc equation generator
very good explanations thank you! What material are the cycloidal disc made of? they seem to have a yellow-ish color..what material are they??
Thank you! It’s aluminum with anodize :)
Strange that Im using the exact same method with exactly the same parameters but solidworks runs into an error saying the equations are invalid
Hi ! What is the estimated backlash/lost motion that you get with your reducer ?
Me in with the same question.. What was the efficiency and backlash? And any ways to reduce backlash..?
Looks like a pretty slick design! From the shots at the end where you move the arm around by hand, it looks like it backdrives pretty easily; do you have any idea what the backdrive efficiency and starting torque are?
Thanks! I never done the experiment of efficiency before, but for back drivability we need to use the servo that can control stiffness and damping while it’s in position mode (like mini cheetah actuator). I did some simple test on cycloid drive with two links arm before, please check out here ruclips.net/video/CJyCcXob0c8/видео.html
Do you have any progress in this project after publication?
Thanks. Good job. Please CAd files?
You are MAN. Thanks.
Hello there,
Excellent presentation...
Can I learn how you do the torque calculation?
I want to calculate for my 1 to 30 ( ratio 1:30) project.
thank you so much.
Thank you!, ratio is defined by N-1 where N is the number of outer roller, in my video I was using N=20, so I can get 19:1 ratio. Torque depends on your motor spec.
Thank you.
If according to your project it was N 30, what would the torque be?
Great job! why are you using two rows of bearings (for each gear). There are some restrictions of using one row of bearings (for both gears? It seems, that cycloidal gears are moving with same speed
Thanks!
Using two rows of bearing because there is no a bearing that thick enough for two cycloid disc gear. So even a solid lobe instead of bearings also okay, but with bearing it will give more efficiency.
@@stepbystep-robotics6881 HK0408 bearing will fits your model perfectly. Are you planning to use single bearing for test? one more question - the shafts for holding surrounding bearings were milled with one case with their holder or they can are separately produced and assemblied?
Hanzhen harmonic drive gear , over 30 years experience in industrial robot arm gear reducer
I found it easier to generate the cycloidal profile in solidoworks and then exported it to fusion 360 and the rest of the components were completed in fusion 360.
hi
can someone share python script how to do the equations conversion ?
thanks a lot *_*