On Thursday 08 August 2013 12:31:30 Jan de Kruyf did opine: > Gene, > I would say, after reading through your code and looking at the pictures > of your setup, that perhaps your table is not as square as you like to > believe;
I can believe that, however the jig is pinned to the table's t-slot with a low profile fin on its back that fitsd well enough I had to press down several pounds to seat it in the t-slot. And it has not been moved since, the pocket for the board was milled insitu. I can put a .0001 dial indicator on the head, adjusted to finger the edge of the pocket, and run the table the length or height of the pocket showing perhaps .0002" of noise from the fiber content of the micarta. I realize this doesn't prove its square since such a measurement is self compensating. The lexan test pieces are cut on my table saw, which is known to be pretty square, and there is no gap that I can see with a lens about 20x better than a jewelers loupe between the reference edges of the board, and the pallets pockets edges. So I have to conclude it well within a degree. I tried the feeler blade, but my thinest is .0015, and it won't enter anyplace along the reference edges of the test piece. I don't want to waste the micarta pallet, but at this point I think I may as well make another, all in one clamp up once the fin on the back has been made, first installing the reference contact, then using its results to place the pocket at 0,0 to 1.375,1.250, not forgetting to use tool compensation. I forgot that the first time so my pocket is .125 over sized, no biggie just cut the raw board bigger. My real problem as I see it is an error in one of the 'fudge' figures defined near the top of the file, so I may waste 2 more pieces of lexan to see if I can arrive at a better offset calibration those 2 'fudges' represent. IMO, if I wasn't square enough for the girls I go with, I would be seeing a location error in a row of holes that varied over the length of that row of holes, and looking at that, its dead true. Looking at the lexan test sample at a 45 degree angle in the x plane, the hole alignment if perfect, turn it to look at it from the Y axis at a 45 degree angle, and the offsets visible are very consistent at about .010" of what is supposed to be a .03125 hole. So it seems square enough and should be fixable with the fudge factors, starting with the y, which becomes x when the R270 is applied. > and at the same time the pc board you were trying to drill is > not as square as you believe. > Then when you flip it, it will not be in the exact same position on the > table, because the skewness that fitted perfectly into the jig first > time around, causes a gap at the top or at the bottom (as you see it, > along the Y axis) for the bottom-up run. I am using the same two edges of the board, left end always, registered to the back of the pocket for the top, and when flipped ion the x axis for the bottom, its wedged to register against the front of the pocket. > Or the other way round. > > With a perfectly right-angled pc-board that would not happen, but who > makes those? ;-) > Try running a very thin feeler gage along the left edge with the board > in both positions and see if it does not go in at the top or at the > bottom. > > To prove the squareness of yr table, mill a square or rectangular plate > (perhaps screw it down in the middle with 2 screws into the T-slot) and > machine a perfectly round hole in it somewhere (reamer?) perfectly in > the middle would be nice, but again that is difficult to archieve. Then > put it in your jig, and clock on the hole with your fancy rotating > probe and your program. > Then flip the plate and clock again. I bet you will find the same > registration error as you find now. > > And here is a last thought: > After you have worked out the unsquareness of your machine you can write > a kinematics routine to adjust for it, and then polish up the pcb jig > so this does not bother you ever again. Unforch, the jig/pallet will be a from scratch effort, this is the smallest of about 6 boards the fellow has designed, but the others are I believe, going to be as big as the free eagle can do. So this particular jig is specific to this board, and it it works, he will get some offshore company to make 20 or so, which is probably an optimistic estimate of the market unless it turns out to be the magic mousetrap for a 30 YO computer that a lot of us still use. We don't expect it to be... So I make 2 boards, he gets one and I frame one copy for me, and this pallet is going in the toolbox drawer, possibly forever. Its a labor more of love than any potential profit motive. Thanks Jan. [...] Cheers, Gene -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) My web page: <http://coyoteden.dyndns-free.com:85/gene> is up! My views <http://www.armchairpatriot.com/What%20Has%20America%20Become.shtml> Debian is the Jedi operating system: "Always two there are, a master and an apprentice". -- Simon Richter on debian-devel A pen in the hand of this president is far more dangerous than 200 million guns in the hands of law-abiding citizens. ------------------------------------------------------------------------------ Get 100% visibility into Java/.NET code with AppDynamics Lite! It's a free troubleshooting tool designed for production. Get down to code-level detail for bottlenecks, with <2% overhead. Download for free and get started troubleshooting in minutes. http://pubads.g.doubleclick.net/gampad/clk?id=48897031&iu=/4140/ostg.clktrk _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
