@Lord-Of-Light
Don’t forget that if you want the gears to last, they need to be made using prime numbers of teeth. Otherwise they will ware unevenly because the Sam teeth will keep meeting up. Using prime numbers, the point of impact will rotate ground the gears resulting in an even ware pattern. This will smooth the teeth and help them run better over time rather than creating a ware pattern that will speed up the parts failing point.
@BuffRobotiX
Excellent video, I was very pleased with the results! If you like the belt drive you might want to try a capstan drive which uses cords or wire and has virtually no backlash.
@HoldenArt
Mr. Mechatronics is an international treasure.
@ovidiurosu6632
amazing, so comprehensive and attentive to details; I loved it; keep up the great work
@arthursidortschuk5835
6:20 just dont. If the contour is not closed than the slicer cant compensate or applys the contour value although you indendet for hole compensation because the CAD had a slot going into the hole. It should be done in CAD with a offset Surface but what do i know i dont even use the superior Onshape. Just be aware that it might mess with you.
@TheMarbo74
Beautiful! Thank you for that. Your belt tensionners are holding with friction. Try using a a screw to apply force around a pivot and then lock that screw. That way the counter-force will not try to fight friction but rather compress a screw. It also allows you to fine tune the force on the tensionner.
@newmonengineering
You should try CAPSTAN drive and see how it compares. To build a good one you just need a rope/string with minimal creep. They are fairly compact, easy to manufacture and have decent reliability when built properly. The neat thing is how simple they are and you can make unusual joint motions by changing their shape like a CAM shaft works. They are not perfect, none are but they are an interesting choice and lightweight. Great video. This is just a suggestion to think about.
@airgavvy
Amazing video. Thank you so much for this. I am also building a robot arm. Improving from basic servos to one of these. I am probably gonna use the planetary gearbox for shoulder and belt drive for other two, since they need lesser gear ratio and it was lighter.
@derekfinch9586
Very informative video. Had not seen your channel before. I'd like to comment on the Cycloidal drive design. It may not change the overall result but should move this style much closer to the others and permit it to be a lot smaller. (1) The eccentricity appears to be less than the maximin possible. The greater the eccentricity the greater the 'bite' of each tooth. For example we used a 48:1 cycloid drive on an FRC robot with a 1 mm offset. It eventually gave up (like yours did). I was able to change to 1.2mm offset (a 20% deeper tooth engagement) and it did not fail. (2) A material like PETG will perform much better than PLA for the gears. (3) Since the cycloid teeth virtually roll on the pins, you can get away without the bearings. We used 3/16" diameter plain steel pins. This permits a smaller, more compact drive. (4) We also used the same type and diameter of steel output pins but increased the quantity to the maximum possible. This provided the maximum surface area for the torque to be transmitted (minimizing wear and deflection). (5) This may be the most significant difference. A two gear cycloid is not fully balanced. However a 3 gear system with each gear set at 120 deg is virtually vibration free. Also, since almost 1/3 of the teeth are engaged at any moment, a three gear system provides nearly 100% of tooth contact at all times. (6) How you slice the gears has a significant effect on performance. We went with 100% solid. (7) As one person commented, use lubrication. You may be quite surprised at the improvement provided by the above changes. With all the pins, the part count will always be relatively high but you might find it could 1/4 to 1/8 of the volume - a real plus for a robotic arm. Cheers.
@LarsPantsonFars
Fantastic video!!!!! The only addition i would add is a noise factor. Cycloidal drives tend to produce much more noise than the other two drives, and may not be desireable.
@SippoCoffee
Very cool and thorough video Just 2 little things I noticed: when you compare the weight you show the total, while one of the main advantages for the belt drive is the capability to mount the motor further away on long arms decreasing the torque requirement for the previous stage. The other thing is: why do you use pla? I can't see any advantage over other filaments in wear and temp resistance other than maybe printability.
@xs1l3n7x
I wasnt looking for a video like this but was expecting it to pop up at some point in time, great stuff
@victorantony7563
this is such a useful video, thank you - did you measure loudness by any chance?
@GMx_1
Excellent video - very interesting and i think the testing made sense and was well executed. Not a surprise though as belt drives really are simple and when designed properly are quite efficient.
@caslor2002
Thanks for the informational video one more time. Consider for the update test to use silicone lubricant for the Cycloidal and planetary drives as and in real life the use lubricant to these setups, will help the friction and the wear of the Pla i think . Good idea to use next time a tension pulley instead a single bearing to the belt drive . Good point also to make the Cycloidal drive and the planetary with very tight fit and then use them for some hour to wear in the right dimensions (maybe that would be good idea if when you use the lubricant inside with your suggested dimensions have not good result in wear out Thanks again all this time for your videos , have helped me a lot in some of my own projects
@yonathandevash7657
he is back!
@zarigata
to be fair in defense of the planetary gearset, if constructed with a more resistant material it will outperform the rest i believe, as the cycloidal is very demanding even in the metal counterparts and there is nothing to improve in the bets as in the belt material, it will be the weakest link in certain cases
@GGGG_3333
He is finally back 😭
@Mulakulu
I wish you volume-normalized and weight normalized the components. Maybe making one set that are all equal volume and one set that are all equal weight, and see how well they compare. While I dont think the cycloidal drive would have impressed if you made it larger, I think it still affected the results significantly. Maybe its performance/size ratio is great?
@RentableSocks