Getting reversal working in Netfabb

makerswamp · November 16, 2011

Hi Guys and Gals.

I wanted to create a thread specifically addressing the stringing issue. On my mendel, stringing disappeared completely once extruder reversal was correctly implemented. This feature has been missing from the ultimaker from day one. Anyone want to help me do some tests?

I did some initial experimentation and could successfully get reversal working, however I couldn't get netfabb to extrude extra after a retraction. End result: the filament pretty much keeps reversing between jumps and not enough plastic is extruded.

Any ideas? / Thoughts?

-Makerswamp

konrad · November 17, 2011

Hi Makerswamp,

Sometimes cleaning stringing residures takes longer than printing itself. I use Paul&Florian's (great) settings with small changes. And those settings have all antioozing parameters in material definitions/jumps section zeroed. Sure low temp settings decreases ooze but doesn't eliminate it. It would be great to see this updated and have ooze-free printouts.

As to the problem you described I assume you play with push-on speed, pull-back speed and travel RPM settings. And push-on does not extrude enough material after the jump. Right?

Netfabb's very laconic manual here

http://wiki.netfabb.com/Material_Settin ... for_RepRap

says:

The Pull-back Speed and the Push-On Speed determine the RPM rate with which the material is pulled back before the jump and pushed forwards again after the jump.

I started to wonder what does it mean "before" and "after" the jump. For how long are those RPMs executed? After all the jump is executed immediatly after extrusion ends.

I thought that maybe that is the time when Z-Jump is executed? So if the Z-offset is set to zero pull-back and push-on are effectively ignored and only travel RPM works.

To verify this I created in Netfabb two small cubes with a gap between them to make jumps and generated g-code (standard quality, hollow object) for them with the following settings:

Jump speed 180mm/s

z-jump speed 20 mm/s

pull-back speed 400 RPM

z offset 1mm

push-on speed 400 RPM

Travel RPM 0 RPM

The result for one of the middle layers:

(begin layer 30 at 2.250)G1 X35.0000 Y12.0900 Z2.6800 F232.0000 E690.9806 G1 X35.0000 Y12.0900 Z3.1800 F232.0000 E690.9806 G1 X30.3400 Y12.0500 Z3.1800 F7800.0000 E690.9806 G1 X26.1151 Y12.0214 Z3.1800 F7799.9995 E690.9806 G1 X26.1151 Y12.0214 Z2.7100 F223.0000 E690.9806 G1 X26.1151 Y12.0214 Z2.2500 F223.0000 E690.9806 G1 X30.7900 Y12.0300 Z2.2500 F1200.0000 E692.0870 G1 X30.7900 Y16.7100 Z2.2500 F1200.0000 E693.1946 G1 X26.1100 Y16.7100 Z2.2500 F1200.0000 E694.3022 G1 X26.1100 Y14.6200 Z2.2500 F1200.0000 E694.7968 G1 X26.1100 Y12.0300 Z2.2500 F1200.0000 E695.4098 G1 X26.1100 Y12.0300 Z2.7500 F232.0000 E695.4098 G1 X26.1100 Y12.0300 Z3.2500 F232.0000 E695.4098 G1 X30.7800 Y12.0600 Z3.2500 F7800.0000 E695.4098 G1 X35.0049 Y12.0886 Z3.2500 F7799.9995 E695.4098 G1 X35.0049 Y12.0886 Z2.7500 F232.0000 E695.4098 G1 X35.0049 Y12.0886 Z2.2500 F232.0000 E695.4098 G1 X39.6800 Y12.0900 Z2.2500 F1200.0000 E696.5162 G1 X39.6800 Y16.7700 Z2.2500 F1200.0000 E697.6238 G1 X35.0000 Y16.7700 Z2.2500 F1200.0000 E698.7314 G1 X35.0000 Y14.6800 Z2.2500 F1200.0000 E699.2261 G1 X35.0000 Y12.0900 Z2.2500 F1200.0000 E699.8391 (begin layer 31 at 2.330)

You can clearly see z-offset before and after the jump (Z2.2500 -> Z3.2500 and back). But correct me if I'm wrong: extruder did not pull any filmanet. Pull-back value was ignored!

I did the same code for no z offset with the same result - no fiament retracting.

The only way I could make E values decrease was to set negative Travel RPM. At Travel RPM = -2 RPM code looks like this:

(begin layer 30 at 2.250)G1 X35.0000 Y12.0900 Z2.6800 F232.0000 E407.8387 G1 X35.0000 Y12.0900 Z3.1800 F232.0000 E407.6720 G1 X30.3400 Y12.0500 Z3.1800 F7800.0000 E406.1186 G1 X26.1151 Y12.0214 Z3.1800 F7799.9995 E404.7103 G1 X26.1151 Y12.0214 Z2.7100 F223.0000 E404.5536 G1 X26.1151 Y12.0214 Z2.2500 F223.0000 E404.4003 G1 X30.7900 Y12.0300 Z2.2500 F1200.0000 E405.5067 G1 X30.7900 Y16.7100 Z2.2500 F1200.0000 E406.6143 G1 X26.1100 Y16.7100 Z2.2500 F1200.0000 E407.7219 G1 X26.1100 Y14.6200 Z2.2500 F1200.0000 E408.2165 G1 X26.1100 Y12.0300 Z2.2500 F1200.0000 E408.8295 G1 X26.1100 Y12.0300 Z2.7500 F232.0000 E408.6628 G1 X26.1100 Y12.0300 Z3.2500 F232.0000 E408.4962 G1 X30.7800 Y12.0600 Z3.2500 F7800.0000 E406.9395 G1 X35.0049 Y12.0886 Z3.2500 F7799.9995 E405.5311 G1 X35.0049 Y12.0886 Z2.7500 F232.0000 E405.3645 G1 X35.0049 Y12.0886 Z2.2500 F232.0000 E405.1978 G1 X39.6800 Y12.0900 Z2.2500 F1200.0000 E406.3043 G1 X39.6800 Y16.7700 Z2.2500 F1200.0000 E407.4119 G1 X35.0000 Y16.7700 Z2.2500 F1200.0000 E408.5195 G1 X35.0000 Y14.6800 Z2.2500 F1200.0000 E409.0141 G1 X35.0000 Y12.0900 Z2.2500 F1200.0000 E409.6271 (begin layer 31 at 2.330)

Fiament is retracted during the jump. But since there is no push-on after the jump extrusion starts with smaller E values than before the jump and thus with void in the nozzle.

That would explain the situation you describe, wouldn't it?

It's late and I might be missing something huge but it seems that push-on and pull-back settings are ignored altogether. That would be bug that makes restart of extrusion after a negative travel RPM jump impossible :-( Can anybody check my reasoning?

On top of that there is no comb feature in Netfabb - that makes stringing even worse.

Edit: Only after writing the above I found confirmation from noone else but Paul http://groups.google.com/group/ultimaker/browse_thread/thread/7511401e952b586c/2c7df6ff0e8dca94?show_docid=2c7df6ff0e8dca94&pli=1. Note to self: use more Google BEFORE. It wouldn't hurt, though, if Netfabb manual didn't describe features that aren't implemented/are not working.