We made gearboxes for many years with hypercycloidal action. The ring and pinion were straight cut gears, some had multiple pairs of gears that would greatly decrease the output speed and often increased the torque. It was a relatively efficient gearbox which was used in a multitude of different end users, from water treatment works, newspaper printing presses, defence contracts.
Wow, I'm just about to make a reducer, and I was worried that my CNC might not give the clean finish and precision I need. I looked at your work, thank you. Your video gave a good kick and accelerated the production of the gearbox. I will do it in the near future! Thanks again for the video!
A brave attempt my friend! I think however you have much to learn in the world of "Machine Shop". About tolerance, squareness, parallelism, fit and finish etc. Also machine design and general procedures. All this has been developed over the last two hundred years or so - a huge body of knowledge and lore. Good luck to you!
it would be interesting to see some measurements. for instance how round did your part come out, how concentric are the bores, how flat. Wondering what the wobble is from and i'm guessing it is from backlash deformation of your toolpaths.
@@dekutree64 it should not wobble at slow speeds. The wobble is due to the eccentric weight distribution, not a function of the mechanism. the output should be non-wobbly with a single disk. The whole thing may wobble at high speed, but assuming perfect rigidity and perfect tolerances the output would stay concentric to the body of the mechanism. in this example it appears the wobble is due to tolerance and likely other manufacturing defects.
@@JesseSchoch The tolerances on my parts are not very good by CNC machine standards. I know that the circles my shapeoko makes are slightly oval and that there must be deviations in other places. Unfortunately, I don't have the equipment to measure exactly how off it is, only a set of cheap calipers. I do a series of spring passes and finishing passes until things fit, so I don't rely on how big my measuring tool or the software thinks my hole is. I essentially use my bearings as the guage of hole size, since they're the most precise things I have around.
@@NathanielBechard A DTI should be all you need. You definitely should pick one up if you don't have one. You can measure the cnc backlash fairly easily and you could measure the roundness of the part on the lathe, just scan the high and low spots of the disk by comparing them at 90 degree offsets. my cnc mill has ~0.05mm backlash in both x and y and i've been wondering what sort of tolerances I would need to make a cycloidal gear.
@@JesseSchoch I do have one of those, I'll test it out. I thought you meant the tolerance within the holes, which is harder to measure. You can have pretty bad tolerances to make a cycloidal gear, but how precise you need the gear depends on your application
making a custom weighted flywheel for your spindle will greatly reduce the chatter and give a smoother cut. Just be sure that you adjust the spindle drive parameters to allow for a longer ramp up time and ramp down time. you can balance the flywheel extremely accurately with a home made needle point balance stand. Feel free to reach out if you have any questions.
@@NathanielBechard That all depends on how your spindle motor is set up. if you dont have the ability to mount it on the top (back side) of the spindle motor, you can make a simple collet extension and mount it on that.
@@MND22 I downloaded a fusion 360 script to generate the drive geometry. Here is the GitHub for it: github.com/woodenCaliper/CycloidalDrive. Now that I think of it, I'll credit this page in the description.
Thanks! These gearboxes are quite good for robots, but they have to be designed with a high degree of precision to have zero backlash. Backlash isn't an issue for my application, and my gearbox has lots of it. What machines will you be making your gearbox with?
@@NathanielBechard I'll be using a cnc dmg mori 3 axis mill. but I have a cnc router just like yours as well but a different brand so i might try it on there as well. I might make one for a combat robot... see what happens.
@@irishracingfling3711 Sounds like a good plan. Are you going to use this gearbox in the drive train of a combat robot, or are you planning on making a robot arm of some sort?
@@NathanielBechard I am making a robot arm, not finished the design though can see it being mostly 3d printed before i go to anything more long term haha. As for the combat robot its going to be used in a grabber/crusher mechanism. The drive will simply have a belt reduction to the wheels powered by a 5045 200kv brushless motor.
That was the Dumbest order of operations i have ever seen.. Start with the center hole, and the pin holes.. then clamp from the inside, without moving the plate, and THEN.. machine the outside..
It may not look like it in the video, but the adaptive toolpaths used in the center and pin holes put a ton of side force on the clamps. When I was testing these adaptive cuts, they were often enough to detach an aluminum plate from the bed (with 4 clamps keeping it down). Using the plywood jig ensures that the part will not move when an aggressive adaptive toolpath is used to open up the pin and center holes.
@@JanBinnendijk Possibly your endmills are sharper than mine, or you are taking a different chip load because I have observed that I need very rigid clamping to stop the part from moving on my machine. I used the technique I showed in the video since it worked well on my machine with my single flute endmill, however, another strategy may be more effective with different combinations of tooling.
@@NathanielBechard If your endmills aren't sharp, Don't even bother using them. I mainly machine titanium, and chipload is as high as the tool can handle.. I can push a 4 mm Endmil in titanium up to 12 mm deep, in one pass.. , and the parts that i machine that way, are clamped on an Expanding Mandrel, with Only one bolt (M5)
We made gearboxes for many years with hypercycloidal action. The ring and pinion were straight cut gears, some had multiple pairs of gears that would greatly decrease the output speed and often increased the torque. It was a relatively efficient gearbox which was used in a multitude of different end users, from water treatment works, newspaper printing presses, defence contracts.
Very interesting! May I ask which company this was done with?
@@NathanielBechard They went out of business around 8 years ago. They were based in SE England. Been trading for just over 100 years.
Wow, I'm just about to make a reducer, and I was worried that my CNC might not give the clean finish and precision I need.
I looked at your work, thank you.
Your video gave a good kick and accelerated the production of the gearbox.
I will do it in the near future!
Thanks again for the video!
A brave attempt my friend! I think however you have much to learn in the world of "Machine Shop". About tolerance, squareness, parallelism, fit and finish etc. Also machine design and general procedures. All this has been developed over the last two hundred years or so - a huge body of knowledge and lore. Good luck to you!
你攻M5的螺纹，还用铰杠？电钻攻螺纹不香吗？钢件我都直接电钻过，M2的我都能直接过，你那么粗的螺纹，我服了。
it would be interesting to see some measurements. for instance how round did your part come out, how concentric are the bores, how flat. Wondering what the wobble is from and i'm guessing it is from backlash deformation of your toolpaths.
Cycloidals are supposed to have two discs to counterbalance eachother. Single disc will always wobble.
@@dekutree64 it should not wobble at slow speeds. The wobble is due to the eccentric weight distribution, not a function of the mechanism. the output should be non-wobbly with a single disk. The whole thing may wobble at high speed, but assuming perfect rigidity and perfect tolerances the output would stay concentric to the body of the mechanism. in this example it appears the wobble is due to tolerance and likely other manufacturing defects.
@@JesseSchoch The tolerances on my parts are not very good by CNC machine standards. I know that the circles my shapeoko makes are slightly oval and that there must be deviations in other places.
Unfortunately, I don't have the equipment to measure exactly how off it is, only a set of cheap calipers. I do a series of spring passes and finishing passes until things fit, so I don't rely on how big my measuring tool or the software thinks my hole is. I essentially use my bearings as the guage of hole size, since they're the most precise things I have around.
@@NathanielBechard A DTI should be all you need. You definitely should pick one up if you don't have one. You can measure the cnc backlash fairly easily and you could measure the roundness of the part on the lathe, just scan the high and low spots of the disk by comparing them at 90 degree offsets. my cnc mill has ~0.05mm backlash in both x and y and i've been wondering what sort of tolerances I would need to make a cycloidal gear.
@@JesseSchoch I do have one of those, I'll test it out. I thought you meant the tolerance within the holes, which is harder to measure. You can have pretty bad tolerances to make a cycloidal gear, but how precise you need the gear depends on your application
Fantastic! How did it work?
making a custom weighted flywheel for your spindle will greatly reduce the chatter and give a smoother cut. Just be sure that you adjust the spindle drive parameters to allow for a longer ramp up time and ramp down time. you can balance the flywheel extremely accurately with a home made needle point balance stand. Feel free to reach out if you have any questions.
Good idea!
Can you describe the process to me in further detail?
How should I attach the flywheel to my router?
@@NathanielBechard That all depends on how your spindle motor is set up. if you dont have the ability to mount it on the top (back side) of the spindle motor, you can make a simple collet extension and mount it on that.
Pretty Cool!!!
Awesome content, Subscribed!
How did you find the dimensions the cycloidal drive?
@@MND22 I downloaded a fusion 360 script to generate the drive geometry. Here is the GitHub for it: github.com/woodenCaliper/CycloidalDrive. Now that I think of it, I'll credit this page in the description.
@@NathanielBechard thank you so much, look forward to seeing you complete the project
9:10 wouldn’t say the part had a high spot. Probably due to the part being fixtured with glue and plywood on a cnc router
43 subs.... make that 44. Very good video. Ill be making one of these gears boxes soon for a robot.
45 checking in
Thanks! These gearboxes are quite good for robots, but they have to be designed with a high degree of precision to have zero backlash. Backlash isn't an issue for my application, and my gearbox has lots of it. What machines will you be making your gearbox with?
@@NathanielBechard I'll be using a cnc dmg mori 3 axis mill. but I have a cnc router just like yours as well but a different brand so i might try it on there as well. I might make one for a combat robot... see what happens.
@@irishracingfling3711 Sounds like a good plan. Are you going to use this gearbox in the drive train of a combat robot, or are you planning on making a robot arm of some sort?
@@NathanielBechard I am making a robot arm, not finished the design though can see it being mostly 3d printed before i go to anything more long term haha. As for the combat robot its going to be used in a grabber/crusher mechanism. The drive will simply have a belt reduction to the wheels powered by a 5045 200kv brushless motor.
看你干活挺累啊，你这也就是做一个玩吧，要精度没精度，要效率没效率，要产量更是没有。这都啥年代了，还整上台钻了，沉孔用台钻手工干，油管一绝啊，大弟！！我会说英语，我英语很熟练。
你这个干法，我看没有什么精度，本身你这台床子精度就不行。怎么的也得用加工中心干啊。我也就看着玩吧，你等我发一个加工中心干的视频
That was the Dumbest order of operations i have ever seen.. Start with the center hole, and the pin holes.. then clamp from the inside, without moving the plate, and THEN.. machine the outside..
It may not look like it in the video, but the adaptive toolpaths used in the center and pin holes put a ton of side force on the clamps. When I was testing these adaptive cuts, they were often enough to detach an aluminum plate from the bed (with 4 clamps keeping it down). Using the plywood jig ensures that the part will not move when an aggressive adaptive toolpath is used to open up the pin and center holes.
@@NathanielBechard no, it doesn't, i use adaptive toolpaths as well, in titanium, with only scrwes holding my part on a wooden sacrificial board..
@@JanBinnendijk Possibly your endmills are sharper than mine, or you are taking a different chip load because I have observed that I need very rigid clamping to stop the part from moving on my machine. I used the technique I showed in the video since it worked well on my machine with my single flute endmill, however, another strategy may be more effective with different combinations of tooling.
@@NathanielBechard If your endmills aren't sharp, Don't even bother using them.
I mainly machine titanium, and chipload is as high as the tool can handle.. I can push a 4 mm Endmil in titanium up to 12 mm deep, in one pass.. , and the parts that i machine that way, are clamped on an Expanding Mandrel, with Only one bolt (M5)
@@JanBinnendijk What machine are you using to machine titanium with a wood spoil board? To me that sounds like a pretty outlandish combination.