RE: [CAD_CAM_EDM_DRO] reprap build started

I'm not sure how is has settled down in existing hobby 3D printers, but ultimately, for a given extruder/material combination, there is a ratio of so many steps for a certain volume of plastic extruded. This ratio is one of the values that I would like to see as a runtime-settable extruder parameter.
At least one of the GCode models uses an additional axis (usually the E axis) to control "motion" (i.e. cumulative volume extruded) of the extruder. Again, in my ideal world, this E value would have definite sensible units (such as cubic millimetres if the other axes count in millimetres). From this is is easy to calculate, given the geometric length of a move and the change in the E value, the cross-sectional area of the plastic bead laid down, and from that and the height of the nozzle above the surface, the width of the bead (within a reasonable range of heights). By considering the E axis as just another axis to accelerate and decelerate, one gets an even bead of plastic.

An older and generally inferior model was to use a special command to set the extruder speed (in arbitrary and probably non-linear units), and make sure the geometric motion keeps up to the correct speed. To do this you have to cap the top motion speed so there is only very short accel/decel at the end of each move (during which time the extruded plastic will be extra-thick because the extruder does not slow down). On the other hand at least one GCode generator takes advantage of this by generating accelerating moves which results in the extruded bead undergoing gradual and continuous change in thickness. This is something that cannot be done with the extra-axis model.

Perhaps what *should* be done is that the extra axis should not be cumulative volume extruded, but cross-sectional area (i.e. volume per unit geometric motion). This would allow both worlds to coexist.

All this glosses over the anti-drool mechanisms (on our printer, when extrusion stops, the extruder actually retracts a tad to make sure the bead is cut off, and when extrusion starts again, it starts with a quick feed to cancel out the retraction). I believe this is currently coded in the GCode (at least for our printer) but again I believe this should be in the smarts of the extruder controller with a handful of appropriate runtime-settable tuning parameters. Note that just raising Z to cut off the bead is insufficient because in the idle nozzle the filament continues to melt a bit, expanding in the process, causing some extrusion without any stepper motion.
-Kevin Martin
the Papertrail Handmade Paper & Book Arts
New Dundee, Ontario
518-884-7123
www.papertrail.ca

-----Original Message-----
From: CAD_CAM_EDM_DRO@yahoogroups.com [mailto:CAD_CAM_EDM_DRO@yahoogroups.com] On Behalf Of Alan

What's the basic relationship of feeding stock to the travel in X &Y?

Sounds like it would be a direct relationship. So one could calculate a
"feed distance" for the stock, and assign it to another axis to drive the
extruder (scaled, of course)...

Seems like if the X axis needed to move 5" to lay down the filament, then 5"
of stock (or some scale of that) would need to be extruded at the same time.
Just like a "coordinated move"! Has someone already thought of that?

I could see a little program that took normal Gcode, and generated the
"pseudo axis" for the coordinated move.