On Sunday 17 September 2017 11:01:48 Erik Christiansen wrote: > On 17.09.17 10:07, Gene Heskett wrote: > > I also have a good level, actually > > 2 of them now, of that fawncy Starrett, except this one is still > > factory calibrated, I can set it, read the bubble, turn it end for > > end, and the bubble settles at exactly the same spot. The one I had > > been using I'd found in the BiL's garden shed after he'd passed, > > sitting bare on a wooden shelf and was quite rusty on the bottom. > > Enough it had a detectable end to end rock when placed on a sheet of > > glass or on my surface plate. I lapped it wet on a sheet of 600 > > wet-r-dry on that surface plate, took the rocking out of it but > > would have had to motorize it and let it run for days to clean it > > all up proper. Readjusted the bubble cartridge until I could turn > > it end for end without any change, but I still don't trust that > > bottom. > > ISTM that for "levelling" a lathe bed, neither adjustability nor a > pristine base are big assets in a machine level. Just flat enough to > sit solidly, and a large arc vial, so that it can detect minute > inclinations, in the order of 0.1mm/m = 0.006° will do. With one of > them sitting solidly on the top of a flat cross-slide, bubble set to > pretty near the middle of the vial graduations by winding the jack > screw at that end, it's just a matter of winding the saddle to the > other end, and jack to the same setting. > Pretty much what I have in mind Erik, I thought I would loosen the ball joint, and while the saddle is sitting on the ball joint, jack the headstock until its level but I'll have to invent the jacks as all there is is tabs intended to bolt it into the floor, but gravity seems to be doing an excellent job of that, then (assuming the bed isn't warped from the years like my back is) run the saddle to the headstock, and any error there will be front to back differential wear. And only moglice will help that. I suspect it needs applied to the bottom of the front lip in addition to the faces of the saddles v-way, but I do not have the measuring tools that could tell me which needs it worse. The front lip, outside of where the rack WAS, is badly cut into by the saddle locking tab with its bent bolt, but the brass, nearly full width gib I put in reaches a good 3/8" past the worn part of the lip, so all I am fighting with in looking for a reference surface is the leftover paint I've not fully dissolved and removed. Yet...
A much larger problem ATM is all 3 of the live centers for the tailstock, are miss-aligned between the point of the center, and the center of the #2 morse stub, so I can rotate the stub in the tailstock socket and see the tip of the center describing a circle about 30 thou in diameter. And I've no facility to reverse mount it, and try to sand that error out of the morse fittings. The best one I have came with it, tip looks like hell, but its closer to true, so I have adjusted and shimmed the tailstock base so its at least usable if I don't push it too hard. The dead center is a center. > That left the middle of the 1 metre bed off by 0.1mm/m = 0.006°, which > is within spec, IIRC. > > Mine set me back around A$200 or so, but it was worth it. I found this last one at an antique place. Still in its original fitted wooden case that Starrett shipped it in. Quite pristine. I had to be carefull I didn't break the hand reaching for the money when I put it in the hand. :) > > So when I find that famous round tuit, I'll check it, and if off, > > will loosen the lock nut and let it adjust itself. My problem I > > suspect, will be the beds swaybacked wear over the years. > > It's past midnight here, so I'm not recalling the name of that tabular > axis error compensation, and can't remember whether it's just for > screw pitch error, or could also tweak x as a function of y. But that > ought to be able to undo the effect of a bit of saddle tilt, you'd > think. Yes it could, but I've not figured out how to use the z position to correct an x error. If there is a way to cross-couple that without another 20 kilobytes of hal code, I am all ears. I think the limit stems from the same rotate a mills coordinate mapping only about the z axis limit, when on these chinese things, we actually need to be able to rotate the map about the x and y axis's too. One could then compensate for a poorly machined post to base error etc. OTOH, how far should we go in making that silk purse out of a lopsided sows ear? Point of diminishing returns etc. > Sleep beckons. > > Erik > > ---------------------------------------------------------------------- >-------- 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
