On Oct 15, 2007, at 10:05 AM, Peter Clifton wrote: > > On Mon, 2007-10-15 at 09:30 -0400, DJ Delorie wrote: >>> Can you enlighten me: Why should the distance to soldermask be >>> greater at corners than at straight copper edges? >> >> Pythagorean theorem. If the mask opening is rectangular, the corner >> is sqrt(2) times further away. >> >> IMHO the right solution is an round-cornered mask as well. > > See my point about X-Y axis of registration. I think for most > machines, > these are likely to be independent, this square mask clearance > would be > better.
Depends on the nature of the registration error distribution. If the errors are Gaussian, the probability contours are round, not square, for independent errors. Why? Because Exp[-x^2]*Exp[-y^2] == Exp[-(x^2 +y^2)]. The isotropic multidimensional Gaussian already accounts for Pythagoras, no need to invoke him twice. > Even when producing by hand, the square + square should be > easier to line up acceptably well. > > Peter > > > > > _______________________________________________ > geda-dev mailing list > [email protected] > http://www.seul.org/cgi-bin/mailman/listinfo/geda-dev John Doty Noqsi Aerospace, Ltd. http://www.noqsi.com/ [EMAIL PROTECTED] _______________________________________________ geda-dev mailing list [email protected] http://www.seul.org/cgi-bin/mailman/listinfo/geda-dev
