bearing in a roller bearing generally have a cage guiding them so they do not rub against each other, and to facilitate assembly as the reduced number of bearings can be placed easily without forcing. What about using a flexible cage ?
Hi @bschwand I already tried that, the cage has advantage for the assembly but what I found, space between balls gets bigger so the contact pressure between the flexspline and the circular spline decreases. The pressure is not that high and plastic is flexible, so its not as stiff as metal so teeth skip easily when the wave generator is pressing that area with no ball.
Great design ! Do you think that teeths this small can support big loads ? In a robot arm for exemple. Also, is there a advantage to using an elliptical wave generator instead of two simple ball bearings pushing on the flexspline ? (I'm currently making my own harmonic drive) Thanks.
Thanks!!! I was about to model from scratch! Do you have a .step file included to make alterations? If not, i’ll just modify .stl’s. Thanks!!! Charge more for these files! Worth every penny. Keep up the good work!
Did you try to use PETG for a flex part? It seems like a logical choice. I made a very early prototype and PETG probably works, but I did not try to print it with PLA and compare
Hi Dmitriy, I didn’t try PETG but indeed It could be a better option. Its more flexible than PLA and more resistant. If you try please let us know if it works :)
Can you elaborate on the tooth profiles between the wave generator and the circular spline? I know modulus is .8 and tooth number difference is 2, but are teeth the same size or is one set reduced ?
Dear ZeroBacklash, It's amazing to see your content on YT! Currently I'm building a delta robot with rotoidal inputs. I would like to know which gear reduction system should I use, in order to get to a 8:1 or even 12:1 ratio. We ahve already designed a 4:1 planetary gearbox attached to a 3:1 belt transmission... We haven't tested it yet, but I'm a bit skeptical about clearences. That is why I would ask your opinion, just to know if we are on the right path!
desired payload and speed control everything else. If you start at the motor you're at the wrong end, like building a car starting with the steering wheel.
Hi @Federico as @Dan mentioned. You should start defining tour payload, but also your reachability. The approach I follow is once these two targets are clear, I start calculating from further joint to the closest near the base. Then with every calculation you know the weight of the motor that your next motor needs to hold. In case of delta robot, if it has 3 arms you can follow this approach for one arm, the others will be identical taking into consideration that the payload will be split between the 3 arms. For the torque calculation always take the worse case scenario. Furthest point with max load.
@@ZeroBacklashRobotics Nope. I tried to print the wave generator and the hub together with the balls placed in halfway through but that was also unsuccesful. I'll try a different material, my PLA seems to be way too brittle.
@@mrraimundo130your print settings may be off. You can adjust horizontal expansion a tiny bit (if you are using a 4mm nozzle) but i'd recommend a smaller nozzle for this print. I will probably change my nozzle to print the flex splines.
I really appreciate sharing your knowledge soooo much. You really contribute a lot the community with this and your other videos!
Thanks a lot!
Thank you Haensel, happy to help people to learn a bit more in this field :)
bearing in a roller bearing generally have a cage guiding them so they do not rub against each other, and to facilitate assembly as the reduced number of bearings can be placed easily without forcing. What about using a flexible cage ?
Hi @bschwand I already tried that, the cage has advantage for the assembly but what I found, space between balls gets bigger so the contact pressure between the flexspline and the circular spline decreases. The pressure is not that high and plastic is flexible, so its not as stiff as metal so teeth skip easily when the wave generator is pressing that area with no ball.
@@ZeroBacklashRobotics makes sense. Maybe a thin cage would work, keeping like half a ball diameter space between each ?
Maybe but bear in mind needs to be easy printable otherwise it will be weak.
Im so glad to that I can purchase those cool stuff! I want to make robotic arm with this!!😅😊
quite interesting.. waiting to see it deployed in an harmonic drive and stress tested. hopefully minimal backlash. Keep us informed :)
Nice design 👍
Great video! That helped a lot.
Thanks Henrique!
Great design ! Do you think that teeths this small can support big loads ? In a robot arm for exemple. Also, is there a advantage to using an elliptical wave generator instead of two simple ball bearings pushing on the flexspline ? (I'm currently making my own harmonic drive) Thanks.
Yes, more teeth engaged so more torque for the same size before they skip or break.
Hey any updates on kickstarter campaign??
Thanks!!! I was about to model from scratch! Do you have a .step file included to make alterations? If not, i’ll just modify .stl’s. Thanks!!! Charge more for these files! Worth every penny. Keep up the good work!
Did you try to use PETG for a flex part? It seems like a logical choice. I made a very early prototype and PETG probably works, but I did not try to print it with PLA and compare
Hi Dmitriy, I didn’t try PETG but indeed It could be a better option. Its more flexible than PLA and more resistant. If you try please let us know if it works :)
Can you elaborate on the tooth profiles between the wave generator and the circular spline? I know modulus is .8 and tooth number difference is 2, but are teeth the same size or is one set reduced ?
But that belt? Where it comes from? With the funny semicircular back surface
3d printed
Yes, PLA 3D printed :)
Did you do torque tests. At what force do you lose the teeth ?
Dear ZeroBacklash, It's amazing to see your content on YT!
Currently I'm building a delta robot with rotoidal inputs. I would like to know which gear reduction system should I use, in order to get to a 8:1 or even 12:1 ratio. We ahve already designed a 4:1 planetary gearbox attached to a 3:1 belt transmission... We haven't tested it yet, but I'm a bit skeptical about clearences. That is why I would ask your opinion, just to know if we are on the right path!
desired payload and speed control everything else. If you start at the motor you're at the wrong end, like building a car starting with the steering wheel.
Hi @Federico as @Dan mentioned. You should start defining tour payload, but also your reachability. The approach I follow is once these two targets are clear, I start calculating from further joint to the closest near the base. Then with every calculation you know the weight of the motor that your next motor needs to hold. In case of delta robot, if it has 3 arms you can follow this approach for one arm, the others will be identical taking into consideration that the payload will be split between the 3 arms. For the torque calculation always take the worse case scenario. Furthest point with max load.
So you sell the models of the arm ?
Mine broke when I tried to push in the second to last one.
Hi Szilágyi, could you finally manage it?
@@ZeroBacklashRobotics Nope. I tried to print the wave generator and the hub together with the balls placed in halfway through but that was also unsuccesful.
I'll try a different material, my PLA seems to be way too brittle.
Ok, thanks. Keep us posted!
@@mrraimundo130your print settings may be off. You can adjust horizontal expansion a tiny bit (if you are using a 4mm nozzle) but i'd recommend a smaller nozzle for this print. I will probably change my nozzle to print the flex splines.
Just show us the 'wrist' workings of your mini-kuka copy 6dof arm you big tease...
Too bad your designs aren’t open source.