Alan,
Nice design! Very practical. I like the use of the electical boxes
as structural elements. I've been collecting parts to build one for
a while now, but have hesitated because of the apparent tolerances
needed. Most of the info, comes fromm the CNC website, and most CNC
designs are 'over engineered' for pcb drilling.
Looks like you have a winning design.
Congrats.
Myc
--- In Homebrew_PCBs@yahoogroups.com, Alan King <alan@n...> wrote:
> Robert Hedan wrote:
>
> > Thanks Alan.
> >
>
> > I knew I could compensate for the total distance, that was
figured out
> > already. I thought a thou or two on the platform, two more on
the gantry,
> > another on the drilling rail and you ended up with one big
sloppy machine.
> >
>
> Only precision needed is X and Y, taken care of in the rails
and screw. Z is
> critical in the alignment, so that the shaft of a drill travels
precisely in
> through the hole cut by the tip. Distance up and down is fairly
non-critical.
> Board being flat is fairly non-critical. A rail on either side of
the board,
> and make sure that they are precisely aligned. Couldn't care less
what lies
> between them as long as they are aligned, the rest of the machine
rides on their
> alignment, not what they rest on etc.
>
>
>
>
> >
> > I've read posts (somewhere) that had really made this
application turn into
> > a woodworker's nightmare. Maybe when I said I'd like to do mild
CNC
> > (plastic) didn't help. But I do need a drilling station and
fast. Your
> > comments are encouraging in that it is within the reach of an
ordinary
> > do-it-yourselfer to build this.
> >
> > Robert
> > :)
> >
>
>
> I think people doing it in wood are nuts :). Or maybe just
love woodworking.
> I sort of figured why make a straight surface when you can go
buy one,
> probably about as cheap as the wood too. Plus for me drilling a
hole and
> putting a screw and nut through it about 100 times easier than
making wood joints.
>
> http://home.nc.rr.com/alan69/CNC/
>
> Board is the first hmm no second time I drilled a pattern, on a
junk board.
> Only etched that board because it was the first one to have enough
transfer
> pattern to etch and see. Next board was useable. I didn't know
what I was
> doing yet eye-balling it, so the upper right holes have a few
misaligned. After
> that it did ok, all off but that's me not the machine they're all
about the same
> off. I didn't bother much either it was just a test.
>
> Controller is my own controller, MOSFETs and PIC and not much
else. Works good
> enough, have done a better layout since it needed a bit of
improvement. New
> layout can take 10 FETs per motor and run 5 phase too. Or just 8
or just 4,
> depends on what you need to run.
>
> Motor is a motor mount and coupler, total of about $5 including
the motor,
> coupler, and mount. Coupler has two set screws on the other side.
>
> Whole is the whole thing. Really need to cap off the vertical
rails, I cut them
> off and they're sharp, usually keep a box over them. Hasn't been
used much in a
> while, the board is out of the clamp. Just clamp the board and PC
board and go.
> Does 1'x1'x2 or 3 inches, could make it more easily by mounting
the side rails
> up a bit and extending the Z axis. Only used 1 8' piece of
aluminum angle, and
> the inner 2 rails from 3 packs of the KV drawer slides. Note that
they are
> really the straightness, the angle is simply for 90 degree
mounting and has two
> holes in it the rail is screwed through.
>
> Xaxis is the gantry. Easiest to access the board from the open
end, so the side
> rails become the Y. Motor is mounted in box with cutout for
shaft, other end is
> just a bearing mount.
>
> Yrail is just the end of one of the rails.
>
> Zaxis is two metal electrical boxes. Easier and fills more space
so more
> accurate, the rail would only go up 1 1/2". Boxes are all kinds
of not square,
> but board travel is the only thing that matters so they're just
canted slightly
> vs the X rails to bring the mounting board into square.
>
> I could probably eliminate 90% of the play that is there,
simply by canting
> things just slightly to work the parts against each other. But
never bothered,
> it basically centers itself within the play, and with everything
aligned right
> there is very little force acting to push it out of line.
>
> Alan