Floppy CNC Router (on the Cheap)

[Rance]

This is wierd. On the way to work this morning I was thinking of posting this proposal and lo and behold, not only had it been sortof discussed, but an update was posted to bring it to the top of the stack. (see itsy bitsy tiny CNC engraver (cheap) which references http://cnczone.com/forums/showthread.php?t=5036)

Unfortunately I am one of those types of guys that can analyze something to death but never get anything done toward the goal. With that in mind, I've been thinking about scaling down my first CNC project in order to eliminate as many of my barriers as possible, both actual & perceived. The main one being cost. Although I've got some background in electronics and am on the virge of stepping into the PIC arena, maybe it would be best to keep my dependency on these to a minimum.

OK, would it be possible, no, how practicle would it be to build a smaller (in comparison to most 1st's here) CNC router keeping the following criteria in mind?

Criteria-
1) Follow the Keep It Simple Stupid methodology.
2) Make this as CHEAP as possible, since no matter what you make for your first one, there's probably going to be many modifications or changes that you'll want for your 2nd one.
3) Use stepper motors found inside floppy drives. These can usually be had for free.
4) Also use the controllers found inside the floppy drive to drive the motors.
5) Use rails from readilly available printers (IJ or DM) that can often be had for FREE as well.
6) If new bearings are needed for the rails, rather having to order Delrin and wait on that delivery and bear additional expense, simply use one of the kitchen cutting boards that have been mentioned here before. These would be more readilly available than Delrin for the average builder.
7) Use a lightweight router (Dremel maybe?) This may turn out to be the most expensive part.
8) Use threaded rod rather than ACME rod as lead screws. Can be obtained from your local Home Depot.
9) Usable movement would be maybe 8" x 5" x 2". (Other suggestions?)
10) For me, this first scaled-down machine would probably be used for routing circuit boards and small, hardwood or soft alum. parts.

Benefits-
1) Use of Steppers from Floppy drives helps to keep the cost to a minimum (probably $0).
2) Use of the Steppers from Floppy drives allows the use of the Floppy controller to drive them, keeping folks from having to fork out significant $$ for a decent controller board. The floppy port already has pins 18, 20 for Step and Direction (ref. http://www.eio.com/public/floppdrv/index.html#398). Graham mentioned a good aspect that if you DID end up buying a heftier controller for your first machine, then that controller could be moved to your 2nd machine, keeping the cost of the 2nd machine lower. The problem might be that you probably don't know what your constraints might be for your 2nd machine yet.
3) Another reason for using the controller that comes with the Floppy drive is to possibly allow the controller to be connected DIRECTLY to the printer port. One possible exception to this would be to include buffers of some kind (2803 maybe?) between the controller and the printer port. However, adding anything more than just plain simple buffering would violate the simplistic aspect.

Questions-
1) Would buffers (or level shifters) be necessary between the parallel printer port and the Floppy controller board?
2) How small would we need to 'shrink' our little machine so that the floppy steppers could move our gantry?
3) Are any of my requirements contradictory with each other?
4) Would this setup allow enough resolution for a) etching PCB's? b) smooth wood carving?

Rance