Re: [CAD_CAM_EDM_DRO] reprap build started

On Sun, Oct 9, 2011 at 3:27 PM, Kevin Martin <kpmartin@...> wrote:

> Actually the easiest solution (now) is to use the maple/aluminum extruder
> body I have already made. Heat leakage (in the amounts we got) is no longer
> an issue.
> If we had been able to print a new extruder body in ABS, we would not have
> had a problem to solve.
>

That works too. :-)

> The filament we have is 3mm ABS, and we find that 200C is a bit cold. We
> found that one sign of overheated ABS is that the extruded plastic contained
> tiny bubbles. Underheated ABS gives opaque milky prints (we have transparent
> ABS) and poor layer bonding. As just the right temperature the prints are
> crystal clear, to the point where we're tempted to print optical elements.
>

I certainly did not mean to imply that everyone will print under 230C. It
completely depends on your materials. Some ABS formulations need even higher
heat-- for example the Stratasys Dimension and Up! printers both print at
about 275C. Transparent ABS is a huge wildcard. I have never heard of anyone
using it, so you are going to have to learn from trial and error. 230 may
well be too high, too low, or just right.

DC extruders as implemented so far are obsolete, but nothing prevents

> someone from using a servo-controlled DC motor which would be just as
> reliable (if not more so) than the stepper feeders.

Actually, there is one thing preventing it... Why bother? Servos have
basically no benefit for the application, so the effort and cost in doing so
would be wasted. Sure, running the system closed-loop would be nice, but it
is easy to do that with steppers, so going to more expensive servos would be
a waste of money. In fact, for the extruder, running a servo would have even
less benefit since a filament jamming/slipping is far more common than
missed steps, and running the motor closed loop would not detect that.
Instead, you want your motion encoder to read off the filament movement, not
the motor movement.

Besides, if you did re-implement the extruder system to use a DC servo, it
would still share basically no code with a plain dc motor, so my point
stands regardless.

People regularly suggest servos as a way to improve the performance of a
reprap (mostly for the XY, not usually for E, though it has been discussed),
but I have yet to hear anyone say why it would be a better solution to what
we have now. If we need more torque or speed (and for the extruder we really
don't), the obvious solution would be to go up to NEMA 23 steppers, not
servos. Servos make perfect sense for a router or a mill, but a 3d printer
has very different requirements. Is there some reason that they would be
better that I am failing to see?


[Non-text portions of this message have been removed]