Homebrew PCBs

> Can you give me a rough idea of what sort of considerations are
involved in the machine to work at that kind of scale? I don't think
I've ever seen a  mill that small and my drill bits only go to #80. 
:-) I guess spindle runout is a major factor there? <

Roy, sure, trying to stump me right off the bat, eh? <g

I purchased a Sherline tabletop mill and lathe and later converted
them to CNC using a "turnkey" system from FlashCut. (www.sherline.com)

The Sherline machines, roughly, each are the size of a sewing machine
and are driven with an IBM 600e laptop I purchased used. 

I bought Sherline because, when shopping, all the various reviews I
read were glowing about the quality of the machines. Perhaps as
importantly, I bought Sherline's "ultimate" package that included,
among other things, quality accessories such as Jacobs chucks and also
quality collets. Quality accessories, as I understand it, have much to
do with a machine's cutting accuracy.

I use collets only with miniature cutting tools; I've read that
collets provide greater accuracy than chucks.

I'm a member of the yahoo SherlineCNC list
(http://groups.yahoo.com/group/SherlineCNC/?yguid=225428826), and
spindle runout is seldom mentioned there as a problem. More likely,
threads will deal with problems such as Z axis backlash compensation
and the limitations of leadscrews vs. ballscrews. A number of smart
and talented Sherline owners have documented, and posted at the site,
 machine upgrades that they've developed.

But, I don't think spindle runout is a problem. Let me give you an
example. I have milled a negative mold of the side of a 40-foot long
highway trailer in 1:87 scale (about 5-3/8" x 1-3/16"). Within this
space are about 40 horizontal corrugations (grooves) that I cut, iirc,
with a .012" ball mill. 

In between each corrugation (a space also .012" wide), in vertical
strips, are rivets drilled with a .005" end mill (about 36 vertical
rows, or more than 1400 rivets). Each rivet is precisely placed; none
"slide" into the corrugations. To me, this is the kind of accuracy I
require -- and get -- from this tabeltop unit.

BTW, I've drilled hundreds of holes on this or that workpiece with #97
(.0059") drill bits and have yet to bust. Any sort of runout problem
would probably snap these, but I'm guessing only.

Let me ask you a couple of questions, Roy. What are the narrowest
traces that you would need to mill on a PCB surface (am I thinking of
a copper-clad PCB ready for etching)? In between traces, what are the
narrowest clear areas required; that is, what is the minimum
separation required between traces?

Did I answer your questions?

-Brian

Brian Chapman
Evansdale, Iowa

PS - I'm hoping, once Curt comes back from the holiday, he can advise
me whether or not I can easily adapt the CNC Sherline mill for use as
part of the EDM system I'm intending to assemble. 

The EDM plans I have include a logic board for driving the EDM head,
but I wonder if I can't build the power side of the EDM system only
and use the mill and CNC software for the rest. . . . I'm way over my
head at the moment, but I hope to break the surface and tread water
while I look around and figure out what to do next. <g


---