On Thursday 01 March 2018 13:36:16 Chris Albertson wrote: > Gene, > > Thanks that is some good real-world data. My plan with this Mini > Mill is to do the conversion in steps. > > 1) Add the z-axis The factory rack and pinyin z-axis setup is not > very good and if nothing else a motorized screw based z-axis is a big > improvement. > 2) Do a quick and simple X, and Y axis conversion using the stock lead > screws and 3D printed parts.
This is std 20 tpi threads, so doing a hot mold in alcetal is likely the best way to reduce the backlash, for a lot longer than adjusting the jackscrews can. > 3) Re-do #2 using metal parts and zero backlash ballscrews If you can find them in small enough to fit sizes, 8mm OD is crowded. The screws I used were the "last of the Mohicans" that Stewart Stevensen had apparently stripped out of a surplus assembly of some unknown sort, I bought the last 3 he had. The nuts and screw fit were such that I was able to refill the nuts with .001" bigger balls I sourced from ebay. 2 of them are in my expended table micromill, and the last is now in the crossfeed of my 11x36 Sheldon. Those nuts were flangeless, so holder cages had to be made. The last "cage" I made used setscrews to clamp lengthwise, with about 3/16" of old felt hat washers of both ends and cone shaped brass pieces to contain the felt, which serves as both oilers and debris wipers. I did somewhat the same in the micromill, but tapped one side of the holder block at 50 tpi, and made a threaded screwin plug to achieve a lengthwise grip on the nut and some of that same old felt hat. Those two have weed eater fuel lines to bring them oil from a manifold on the post, but the sheldon's cage was filled with silicon seal on the bottom, and I remove the bolt thru the top that connects to the crossfeed and refill it via the bolthole in the cage. > 4) Add a 4th axis Stay away from any table made in India. If you think Chinese stuff is schloppy, you'll get quickly informed its worse, lots worse than the Chinese stuff. I was dumb and bought a 4", but when I motorized it, the inconsistent backlash and lack of a set of table locks became quite obvious. And the odd angle the motor sticks out is a PITA, always hitting something. So look for something with a horizontal worm AND clearance enough to clear the table with the motor unless you want to make a pad to gain the clearance. I've been mentally plotting how to make a 45 degree tilt mount so the motor will stick out horizontally, but haven't yet found out where I stashed my round tuit, may have to make more of those if I can ever find a pattern that doesn't have copyright stamped all over it. :) > I want to document the process so if it works someone else can > duplicate it. > Thats to be desired. > I ended up reading a lot of data from the belt supplier's web sites. > Yes GT2 profile should be used in every new design. Only use > trapezoid profile teeth as replacement parts. The newer profile > almost doubles the system performance and as you said, it uses a much > lower belt tension too. > > The major different I find between ball and lead screws is friction. > A ball screw converts about 95% of the applied motor torque to thrust > while a lead screw converts only 40% to 50% of the torque to thrust, > the rest of the power goes into heating the screw and nut by sliding > friction. So in theory I should see at least a doubling on the z-axis > performance you have. I don't know, we shall see. My performance is limited more by low motor voltage (28 volts) than by friction. The diff between 28 volts and 40 volts, using the same 2m542 driver, is eyebrow raising. Same motor, damper & driver on the 7x12's z axis, 2/1 gearing (not belt, gears, 40 on motor, 80 on screw), makes around 90 ipm on the lathe. > My cheap ($190) 3D printer has me spoiled. It routinely operates at > 100 mm per second. (in Imperial units that is 230 inches per minute.) > It can reliably move 30% faster. My eye cannot really follow motion > that quick. It can print anything I can draw. Now I want milling > machine like that. So do I, till I think of the cost of broken tooling. > BTW, I looked at your web site. Noticed the wooden gear model that > gets about 60:1 reduction. That same principle is further developed > by a guy who uses the name "gear down for what" in Thingiverse. He > has a very well engineered system of modular parts. You can assemble > high torque systems with no tools. He uses ring gears that are thin > enough to be flexible, you wrap it around the planet gears then press > fit a round collar over the ring gear that stiffens it up. At first > look it seems more complex than a 15-speed automatic transmission but > after some study is its very simple. He uses herringbone gears which > are like two helical gears, one backwards glued together. > Herringbone gears are very strong and have zero side force but are > near impossible to cut in metal without a high-end 5-axis mill and > tiny ball end. A spur gear version would be much easier. In any > case his innovation was to cut all the sun and plant gears on ONE > SHAFT. And not to use any kind of gear carrier or bearings. The lack > of carriers and bearing makes in very buildable. He has versions > that go from 20:1 all that way to 1200:1 all based on the same design. > It is very compact, no wasted space. > https://www.thingiverse.com/thing:2731585 I'll have to look at that again, since I now have a mill that can probably do that. > > Writing g-code by hand might be a problem for design this complex. > My brain would explode if I tried to write the code for just one > helical gear. But it's trivial to go to the Boston Gear web site and > download a CAD file for one of their stock helical gears. Then I > convert that to g-code. I don't have to understand involute curves. > I just download a stock part from their catalog. The McMaster Carr > catalog works too. The whole process takes 10 minutes and I end up > with a 3MB g-code file. > But its all step and repeat, Chris. 100 LOC at worst. If something needs a tweak, its one tweak, not 70,000. I'll readily admit I've got more than 10 minutes in writing it though. [...] -- 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 Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
