On Friday 15 December 2017 14:20:27 Chris Albertson wrote: > If the goal to build a zero backlash rotary table the gold standard is > a "harmonic drive". These are simpler designs that are inherently > near zero backlash, typically at the arc second level. It works on a > similar principle -- a ring gear and another gear with one less tooth. > These devices just blow away any worm and wheel. > > https://en.wikipedia.org/wiki/Harmonic_drive
My personal objection to the normally built harmonic drive is two fold, first being the up front expense, and second, the flexing of the outer gear has to have an effect on its working lifetime. Even if the flexing of the rim is well within the elastic limit of the steel, the stored stress will lead to a crack failure at some point. Might be several millions of cycles, handily outlasting me and my great grand children, but. I have heard, not recently, of a rifle action failure when it had never been 'overloaded' due to the buildup of stored stress. Its said perhaps 20,000 rounds or less for the larger, and higher pressures developed cartridges, up to perhaps 100,000 rounds for a lower pressure cartridge such as a 30-30. Any of those figures will outlast the barrel. OTOH, my P17, an Eddystone action thats famously hard and therefor brittle, has gone thru its 5th barrel, and close to 20,000 rounds, most of which were 60kcup+ Ackley-06 loads. So its 6th barrel I just put in it, is chambered for the somewhat lower powered 6.5 Creedmoor. > THIS is the project I want to copy https://youtu.be/mmOnktzifeg and > here is part 2 https://youtu.be/xaEhkF9S0mY Excelent videos > > If I could find a harmonic drive cheap enough or if I could make then > want 7 of them for a 7 DOF robotic arm. My current arm uses model > airplane serves and has a 1/2 to 1/4 inch position error and very > little payload ability. > > Autodesk Fusion 360, an animate parts too. That s what I use. It is > it's free (until you use it to make $50K gross income) and prefect for > the above. Yes can output g-code for mills and lathes. > https://www.autodesk.com/products/fusion-360/overview > I just pulled in blender, but its an older version, can import, but not export, .svg's. And inkscape can I am told, then export the g-code. How well that works is TBD. I think there is an export .svg option thats installable, but haven't found it yet. Might take a newer version of blender, there have been 14 point releases since the debian repo version. > On Fri, Dec 15, 2017 at 10:09 AM, Gene Heskett <[email protected]> wrote: > > On Friday 15 December 2017 11:11:32 John Kasunich wrote: > > > On Fri, Dec 15, 2017, at 06:16 AM, andy pugh wrote: > > > > If you look at stock internal gears: > > > > http://www.hpcgears.com/pdf_c33/17.7.pdf > > > > You will find that getting a pair that differ by only one tooth > > > > isn't that easy. And they will differ in PCD. The PCD difference > > > > isn't to hard too deal with in external gears if you can make > > > > them big enough for the corrections not to distort the teeth too > > > > much, but it would be much easier to not have two PCDs on the > > > > planets. So, you would ideally be making your own internal > > > > gears. > > > > > > I believe it would be possible to design a similar drive using the > > > same internal gear for both the grounded ring and the output ring. > > > All the gears would have standard tooth forms, and could be > > > off-the-shelf. > > > > > > The trick is that the planet gears centers would have to be at > > > different radii. So instead of the two sets of planet gears > > > running on the same planet carrier pins and being fastened to > > > each other, you would have two sets of planet carrier pins > > > (probably on opposite sides of a planet carrier plate). The > > > planets would all be independent. One set would have N teeth, the > > > other set would have N+1 teeth. The sun gear would be two stacked > > > and coupled gears, one with M teeth (meshing with the N tooth > > > planets) and one with M-2 teeth (meshing with the N+1 planets). > > > > > > Not sure if what I'm describing is still a wolfrom drive, but it > > > would have the same result - a very high reduction from planet to > > > output ring. > > > > That sounds like a quite practical thing once the math has been > > worked out. And it also sounds like it would be a heck of a lot more > > efficient assuming the planets were running on caged needle > > bearings. Ratios high enough they could be a rotary table drive, and > > most certainly less backlash than my current 4 incher has which is > > tight at some positions, and a good degree at others. I just > > yesterday watched a guy build a std planetary set using blender, > > which in turn allowed the wireframe to be animated for proof of > > concept. I had no idea blender had all those capabilities. I believe > > it can actually export g-code too. > > > > But something along these lines, using gears maybe 1/4" long, sounds > > like it could be stacked in a rotary table casting, one that could > > be moved by a nema23 motor while resisting cutting forces. Mine > > can't even do that when stationary due to the backlash. Piece of > > junk from India for a smidgeon over a $100 bill today, I doubt its > > accurate enough to do a usable gear. Those 1/8" thick, #25 chain > > sprockets I made that one of you kind folks wrote the gcode for me > > several yeas ago, I cut with a new 1/8" 4 flute carbide tool, and > > due to it moving to the other side of the backlash on opposite faces > > of the tooth, cut the teeth wide enough that I had to file the wider > > part of the tooth about 4 strokes of a fresh mill bastard on each > > side of the tooth before the tooth would properly enter the > > inter-roller spacing of the chain. And while it still runs "lumpy" > > it does get the job done, which was driving a taller fence with a > > couple pieces of 1/4" redi-thread on my bandsaw so I could move the > > fence while maintaining the drift angle well enough to cut about > > 3/16" slabs of butternut out to make panel inserts for all our > > kitchen cabinets. Butternut I got from Ray Henry on one of my trips > > to the UP. > > > > I am "intrigued" if you've time to hack up some drawings. I have > > 1/2" thick micarta which could be used for a proof of concept. > > Highland Hardware in Hotlanta has more of it. And I have about 8" or > > so of 1" Acetal rod for the planetaries. I bought it to hot mold > > zero backlash nuts from, then bought some teeny ball screws instead > > from Stuart S. > > > > Heck of an idea John. > > > > Cheers, Gene Heskett > > -- > > "There are four boxes to be used in defense of liberty: > > soap, ballot, jury, and ammo. Please use in that order." > > -Ed Howdershelt (Author) > > Genes Web page <http://geneslinuxbox.net:6309/gene> > > > > ------------------------------------------------------------ > > ------------------ > > Check out the vibrant tech community on one of the world's most > > engaging tech sites, Slashdot.org! http://sdm.link/slashdot > > _______________________________________________ > > Emc-users mailing list > > [email protected] > > https://lists.sourceforge.net/lists/listinfo/emc-users Cheers, Gene Heskett -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) Genes Web page <http://geneslinuxbox.net:6309/gene> ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
