"rangerssuck" wrote in message
...
On Monday, October 24, 2016 at 6:13:41 PM UTC-4, Jim Wilkins wrote:
"rangerssuck" wrote in message
...
On Sunday, October 23, 2016 at 9:27:13 PM UTC-4, Martin Eastburn
wrote:
How big a board are you talking about ?
Will a cnc machine do the job - there are companies that make
enclosed
lasers.
Martin

No particular size, though my work usually fits in 8" x 10" or so.
This was more of an exploration into the practicality of producing
quick protos in my shop rather than sending them out and paying big
bucks or waiting a long time for them to come back.

when you say "cnc" I assume you mean milling machine. That would
work
for some boards, but I don't know what kind of resolution you can
get
that way, I find it hard to imagine 6 mil traces and 6 mil spaces
from
a milling machine, though I haven't looked in that direction
recently.

Another possibility is using a laser to remove a resist layer
applied
to the board. People are getting promising results that way, but
it's
not easy to control a chemical etching process with such small
features. It would seem (to my uneducated mind) that it would be
easier and more repeatable to dial in the parameters for a laser
etch
than for chemicals.

The machine I saw a price for was from LPKF. I did not find an
online
price, but I did find a public record contract from a college in
the
Pacific Northwest to purchase one. Delivered and installed, with a
day
or two of training, the prices was around $250K. A bit out of my
price
range.

I do have access to a 40W laser cutter at the local makerspace, and
I
may take a whack at the etching of a chemical resist (most likely
black Krylon) when I get some free time.
=========================================

I used these with generally acceptable results:
http://www.t-techtools.com/store/

At the time, the 1990's, 6 mils was difficult but possible for
them,
IIRC 10 mils was easy. The Z height adjustment of the tapered
engraving cutter relative to an adjacent foot that pressed the
board
against the platen set the space width. It was capable of milling a
GPS receiver. I used Mentor's PADS for PCB design .

--jsw

interesting. They say they do 4mil trace/space, which is pretty
damned good. I'll be contacting them for system pricing, though I'm
guessing it aint cheap.

But still, I'm thinking about this more as a diy project.

Mechanically the T-Techs I used consisted of a smal high-speed motor
plus 1/8" shank engraving cutter on an X-Y bridge mount. The head was
spring-loaded down and located in Z by a foot resting on the board,
which compensated for warpage, until a dull cutter began raising
burrs. A solenoid raised it. Blank boards were located with dowel
pins, so they could be accurately flipped, and held with masking tape.

To some extent you could do the same job manually on a mill with the
high speed motor or Dremel spring-loaded below the spindle. I think
the main difficulties would be translating the PC artwork vectors and
adjusting the cutter depth to the intended width of cut precisely
enough.

A Gerber file consists of line endpoint pairs with an associated track
width, originally meant to be photoplotted with light through a round
aperture. You could have undercutting at corners if you only made the
tool paths the track centerline plus/minus half the track + space
width.

--jsw