I've been using the gcode interpreter from the latest arduino firmware release on sourceforge, driving it via send.py after creating/exporting gts from AoI and processing via skeinforge latest zip from the forums.
When I extrude it seems like the feedrate is doing the opposite of what it should with the oozebane-generated commands. For instance I have gcode for the beginning of a line segment like this:
G1 X17.94 Y6.92 Z2.58 F60.0
G1 X17.85 Y7.01 Z2.58 F180.0
G1 X17.76 Y7.1 Z2.58 F180.0
G1 X17.71 Y7.15 Z2.58 F300.0
which I interpret loosely as "move from start position to x17.71/y7.15 via four segments with extruder feedrate increasing from 60 to 300".
What's actually occurring is that the stepper speed is increasing instead of the extruder feedrate. So the end result is that the first segment which is supposed to have to lowest feedrate is extruded at the standard feedrate but with the extruder moving at the lowest speed, resulting in too much plastic getting set down there.
So I think there's a bug in the gcode firmware which is using the G1 F param as stepper speed instead of extruder feedrate. Has anyone else experienced this? I'm working on a fix but don't want to reinvent the wheel if someone else has tackled this.
Thanks!
-Adam

F is the head feedrate, not the extruder rate, which is controlled by an M command.

I haven't used it yet, but as I understand it the whole point of oozebane is that because you can't start or stop the extruder quickly, the extruder is turned on and off early and the head speed is modified to match the actual rate the plastic is flowing.

For example, when the extruder is turned off, the barrel has pressurised plastic in it which oozes at an exponentially decaying rate. To compensate for that you need to decide how long you wait, as with it being exponential it never actually stops, but it will become negligible after some finite time. You can then turn it off that amount of time early and slow down the head in a approximately exponential way to match the volume extruded.

A similar scheme would be needed when starting with the head starting slowly and ramping up to full speed as the plastic flow increases.

Edited 1 time(s). Last edit at 10/09/2008 01:28PM by nophead.

---

[www.hydraraptor.blogspot.com]

Ah, I associated feedrate with only the extruder feed, didn't realize that could apply to the movement of the extruder head as well.
I guess I just need to tune this more, right now I have a choice between good edges and extremely sparse infill or tons of ooze/buildup on the edges and solid infill.

I think oozbane is really for fine tuning. You should be able to reasonable results without using it. It is probably easier to start without it to get the basic parameters correct and then experiment with it later to see if you can improve the stringing.

---

[www.hydraraptor.blogspot.com]