Stuart, Jon Elson, and John, Thanks for the feedback and that's why I posted here to get some feedback before I went and did something crazy.
Stuart - I'd love to see that Cinci in action someday, although I'm in Cali so it might be awhile. Also, short emails escape me and I appreciate your feedback. You're right about 'what do you need' and I think for now I need to see how well I can get the mill working as is. I would say, I don't want to buy anything from that 300 person shop:-) Jon - Backlash is a concern on my 1988-ish mill. When looking at Halscope, the f-error plot is more jagged than John Thornton's f-error from his tutorial and I can't help but think backlash (and friction) in the system might be causing that. Also, with FF2, I can't seem to get the accel/decel as nice and flat as his graph either. I'll take a backlash reading and get my machinist buddy over to look at it. I'm really not much of a machinist, but trying... Also, I stopped tuning after the X, I'll do the Y and Z and see what I get, I bet I learn something from that. John K - I am familiar and like US Digital encoders. I didn't look at ebay, but I was comparing new differential US Digital rotary encoders to the used renishaw 1um encoders and in my head, seemed like a close comparison... I did a lot of forum reading and didn't see anyone go w/ a tach feedback and close both loops, so I'm a little nervous on trying that because it would be new. However, If I try that (least expensive solution), I'd be happy to put together a writeup on it to help others. I think I need to tune all axes and check lash in each axis and see how easy it is to tune and my results. Then, make a couple parts and see how well they turn out. If they're w/ in 0.001, I should be fine, I would think. To your hobbyist point, I guess I'm trying to stretch the hobby part to make some money from it when I quote rather simple mechanical systems. I had a quote for qty 4 of a small part at $95 and I could easily make those on the mill (maybe by hand) and they'd be good enough. So, I hope to have some decent results, but w/ a used mill from 1988, I can only expect so much. Thanks everyone for the feedback! Mark ----------------------------- > > Message: 6 > Date: Sat, 30 May 2015 07:50:17 -0500 > From: Stuart Stevenson <[email protected]> > Subject: Re: [Emc-users] Dither with Anilam linear encoders > To: "Enhanced Machine Controller (EMC)" > <[email protected]> > Message-ID: > <CAHF93hRWqSbfuLFReN2jEiOD63ifpJXus_BNPnYhMEd= > [email protected]> > Content-Type: text/plain; charset=UTF-8 > > Mark, > I understand the desire for as much precision as possible but do you > regularly work to tolerances of less then .001? > If not then dither of a count or two will not affect your part making > precision in the slightest. > > an old racing quip is "Speed costs money. How fast do you want to go?" > apply that to precision "Precision costs money. How close do you need to > cut?" > > My goal with the 5 axis Cinci at MPM was positioning the tool tip within a > .010 volumetric accuracy. I felt I could cut +-.010 parts all day with the > machine at that accuracy. I did not calculate the volumetric accuracy from > the readings I saw on the 3 indicator device but I estimate about .005 > volumetric accuracy. I figured if I put a part on the mill with a +-.001 > tolerance it was my fault for putting it on the wrong machine. We cut very > few parts with a positioning tolerance of less than .005 and almost none > with less than .001? When we do we use other tools and machines. > > When I got the 5 axis Cinci an old shop owner (300 man shop) told me the > check was to run a part and if it sold then it was good enough. I didn't > run quite that loose but I soon reached the point of dimishing returns. > > ------------------------------------------------------------------------------ _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
