On Tuesday 16 October 2018 03:33:30 theman whosoldtheworld wrote: > unfortunately I'm not always at the pc ... I'm often away for work ... > so .... but your problem made me think of something that I did a while > ago ... a vision system that chose cards from a pile .... one of the > problems was to find the symbols of the playing cards .... a basic > part of the program was to identify the circle as a geometric figure > .... so I thought: you fix the laser to the spindle ... shoot at a > webcam that is on the tailstock .... at this point the center does not > matter and even the cone of light with opencv we always identify the > center of that kind of circle that the laser prints on the webcam ... > the center will be very precise and you with a piece of iron you can > always get to the center of the design you get on the webcam ... there > is a bit of code to do ... but it can be a nice tool. Having said that > today's LDVTs are very precise, they already have on board an > electronics dedicated to the measurements and can have both digital > and analog outputs ... so that you can use the mesa Analog input card > > The 7i90 has no analog input. Any A/D would have to be fed into the pi in 8 bit pieces, 16 or 24 bits accurate and let the pi do the calcs. And it doesn't have the iron for that job. And googling for LVDT gets me to the omega pages, where a non-contact is $290 USD, but in a machine with over 1 thou of backlash, probably a decade of accuracy overkill for this job. I was able to get it within a thou Sunday with the laser, and thats close enough for the girls I go with. I made all the corrections with Z moving left and x moving in. With the backlash comp in use, that should also get me close enough to fine tune an internal MT5 taper. But I've already made it fit the taper I have after grinding the bent spindles internal taper concentric again, and I'm out of steel in that adapter getting there. To get the wobble I'm getting reduced further will likely need to have both backing plates removed from the chucks, remounted and the face fined up with the toolpost grinder. They were last touched with a freshly sharpened HSS tool, leaving a fine powder floating in the air.
Fixing bent spindles is a cast iron bitch. But I am gradually getting it usable. > Il giorno lun 15 ott 2018 alle ore 00:13 Gene Heskett > <[email protected]> > > ha scritto: > > On Saturday 13 October 2018 18:43:25 Gene Heskett wrote: > > > On Saturday 13 October 2018 16:10:16 andy pugh wrote: > > > > On Sat, 13 Oct 2018 at 15:03, theman whosoldtheworld > > > > > > > > <[email protected]> wrote: > > > > > I think for your pourpose the better idea was using a very > > > > > precise LVDT and put the measure of your axis in a table for > > > > > make a good backslash linearization kinematics file .... > > > > > > > > Or use a dial indicator and a notebook. > > > > > > And a long, perfectly straight rod, turning slowly enough not to > > > beat the $100+ dial to bits, and try to get an average reading. > > > I'd need a 34" rod, a dog to turn it and an MT5 dead center, none > > > of which I have. > > > > > > And from my experiments this afternoon, I am going to have to find > > > a single spot beam green laser, this red line pistol cartridge > > > model isn't going to give me good data, ever with its rotating - > > > pattern beam. I've put a layer of cloudy poly on the cds and it > > > helps, but it all goes to hell the instant the knife on the > > > toolpost starts to interrupt the beam since the beam isn't round, > > > and there does not seem to be a null at exactly half the beam > > > blocked. I changed the drive belt to get more spindle revs, but at > > > 1200 revs the front bearing is smoking in about 6 or 7 minutes > > > with the bolts loose. And the reading still wanders around since > > > the spindle is not synchronized with the meters A/D. So I've got > > > to find a different way. I was hoping I wouldn't have to putz > > > with building a camera, but that might yet be the only way. > > > > > > I'm getting my cataracts fixed starting next week, so this project > > > is on hold for the time being. > > > > Actually, blame it on oldtimers I guess, but I thought what would > > happen if I put a honking big cap across the CDS cell, so I went out > > and stuck a 4700 UF cap across it. This made the settling time > > several seconds, but worked a treat, so now I have it well within a > > thou from just clearing the 4 jaw with the side of a carriage, which > > is -3.500 in machine coords based on my homing, on out to around 23 > > inches. There seems to be a bump there, or the bellows is pushing, > > or perhaps I don't have that screw bearing properly adjusted. Shop > > made bracket by an amature, me. :( Thats about 1.5" from running > > into the tailstock. So I feel like I actually accomplished > > something for a change. Murphy was present and up to his usual > > mischief, the mouse battery died in the middle of it, and of course > > I had to come into ther house ands get one out of the coyote.den, > > and I'd give my left one to have a switch on the DVM to disable its > > timeout shutoff feature. Batteries for it I have. > > > > FWIW, the maximum correction was about 3 thou, in the middle of the > > bed, most corrections were under a thou. there's a bump in the > > middle of the ways that I can't see, or feel thru a paper towel > > wiping it. > > > > But its better than it was. > > > > > Many many thanks for all the help and ideas I've been offered, > > > even if my explanations lacked clarity, its appreciated, a lot. > > > > Thanks again everybody. > > > > -- > > 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> > > > > > > _______________________________________________ > > Emc-developers mailing list > > [email protected] > > https://lists.sourceforge.net/lists/listinfo/emc-developers > > _______________________________________________ > Emc-developers mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/emc-developers -- 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> _______________________________________________ Emc-developers mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-developers
