make a 3d printer. . . . . .yes or no

Take a look at Trinity labs Aluminatus A1 for ideas.
It is in beta development  $2000

Linear actuator X Y     12x12x12 print volume

http://trinitylabs.com/pages/aluminatus-overview

Making a printer? Yes -in fact I'm planning on doing this myself!!

for your other questions....

Ball screw drive. accurate yes, fast -not so much, and rather expensive.

Everyone knows that the current rep-rap style drive with stepper, no feedback and just hope that no steps are missed system is less than perfect, (just search for threads here where people talk about shifting axis and missing steps).
but closed system electronics just is not yet available for these machines on the same sort of price point as the open loop electronics like the solidoodle has.

with the build area, when you start getting very large you're going to find it harder to avoid flexing in whatever gantry you have, need thicker rails, which increases weight and may decrease speed, also keeping your build area warm will get increasingly more difficult as the volume increases.

so far as what carriage to move, or how to group the carriages. the idea is that as you stack stages you increase error.

so assume that your model stays completely still, and the print head moves in X, Y and Z, all the Z wobble that has been talked about will now occur directly at the print head, not at the platform. the print head must be movable, so how do you restrain it to reduce the wobble?

which is why generally people making these machines only have two axis of movement on any part
(e.g head may move in X and Z, or X and Y or Y and Z, or just Z, or just Y or just X with the platform moving in the other axis).

which is best? it really doesn't make a whole heap of difference.

if you look at the original reprap darwin it moved the print head in the Z axis, whilst the table moved in X and Y. this means it's footprint was at least 4 times the footprint of the largest printable part (plus a bit extra for the framework.

if you look at the darwin machine the platform moves in the X direction, whilst the printhead moves in Y and Z. this means that the printers foot print is slightly over twice the foot print of the largest build.

The solidoodle machine where the print head moves Z and Y, with the platform moving in Z means that the footprint of the machine is around the same size and the largest printable object. (obviously plus a bit extra for the frame etc)

so for issues of space you may chose to go with one particular gantry system over another. -that's really up to you though!

with standard milling, you generally move the part being made as this is a lot easier to move than the head of the machine where you may have a large motor and gearbox system attached to the spindle. (and the part you're making may be only a few inches square piece of steel.) -basically, with standard milling you're moving the part that is easiest to move!

this is the one i was thinking about when I made a post before about multiple print beds. You could have a gcode trigger to push another glass bed in every inch or 2 inches int he z direction. This so you can print many flat/small parts in one go

hi folks

note on the picture. . . .there is a detail that gives me a hard on ! !

look at the extruder, and how they deliver the filament to the hot end:  far away from the majority of the printers.  i love that.

anyone know and can tell me if it work good, are there any problems with the extruder etc etc etc.
this is a mod i would like to try on my solidoodle.

Im working on a kossel now just finalizing my parts choice and will most likely have a build thread once I get everything