The story is this: I live in a small apartment with my fianc=E9e, who is
anything but a computer/electronics nerd. Also, she has strong
opinions about me etching PCB's in the kitchen (due to the lack of
space for a proper "chemistry room" with a place to heat etchant),
because of using acids. Myself I do not mind, but I have to respect my
fianc=E9e.

So I got this brilliant idea of mine some days ago. I will construct
some device that drills/mills the PCB's for me. I have seen a lot of
people building their own CNC-devices, so I thought this might be a
nice idea. Not only because I will have a project active (I tend to
get a lot of frustration when I'm out of ideas), it will also let me
apply theory from my mechatronic engineering-studies. I'm still on the
first year (of three), so even if I don't have a lot of theory in the
luggage, it will atleast help me apply some practical work for the
mechanics course (which I find do a lot of brain damage to me, tough
course that one).

That said, what I'm thinking of is something like this; a metal
construction (looking like a cage), with an X/Y-plane (to put the
PCB's on) and a dremel-tool, or similiar, sitting on the Z-shaft. The
construction will have three stepper motors (one for each axis; two
for moving the plane and one for the dremel/whatever up and down) and
a PIC-based controller. The controller gets its instructions from a
computer using serial link (I'm having MAX232 in mind, but perhaps
parallell transfer is better), like receiving all X/Y-coordinates for
copper tracks, drill holes etc.

Instead of milling a whole board, I was thinking that when the board
is ready, it will - in my CAD-program - be saved in image-format (png
or whatever). I will then write an application that scans the image
and finds all the traces / holes. When a trace is found, it will apply
a graphical outline to it. The outlines will be where the mill will
hit and run on, and something similiar will go for the holes. When the
image is done processing in my program, it will in a way or another
convert all the outlines to coordinates - or similar - and transfer to
the controller, which will then move all the steppers to the correct
position and start the milling (or whatever the correct term is).

What I've written above is probably super basic for any
CNC-experienced person, but since I have zero knowledge about
mills/CNC/drilling/etc I feel the need to describe the project deeply
down to the gory details (since I will, at the same time, describe it
to myself). When (if) finished - it will help me a lot, in a lot of
ways. It will make my PCB's and my fianc=E9e will be more than grateful
when I return all acids and dangerous stuff to the recycling center.

I have seen a few constructions like this, but for this one I don't
want any ready-built schematics or anything like that, I want to go
from square one and reinvent the wheel. Besides getting a good tool, I
enjoy the taste of learning new things and constructing stuff
completely on my own (with other peoples' guidelines, ofcourse). Now,
the only knowledge I have about anything of the above is programming
the PIC. Thus I will have to learn more about stepper motors, about
serial communication (RS-232), reread a few chapters in my mechanics
book about momentum and forces, and a few other things.

So, this is where the experienced PIClist comes in.

* Is my idea feasible?
* What do I want to learn about stepper motors, except how to make them rot=
ate?
* Is RS-232 the way to go, or should I aim at parallell ports?
* Is the idea of making an application that "outlines" PCB-traces a
good idea or waste of time?
* What do I need to learn more about how to pick a "good enough"
stepper motor that can handle the required force in milling a single
sided copper clad?
* What other questions should I ask myself about this project?

I'm looking forward in hearing your comments.


-- =

- Rikard - http://bos.hack.org/cv/

-- =

http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist