Hello,
I've noticed an underextrusion problem in my printer. I've already replaced extruder and nozzle of my original Geeetech Prusa i3 Pro B, so it's unlikely to be related to that specific model. That's why my question is more generic I guess.
I noticed underextrusion in one single use case - printing a cylinder with 1.2mm thick wall (0.4mm nozzle). There's no such problem when printing straight walls (also 1.2mm thinkess).
Underextrusion happens after nozzle travel, when it starts printing a curve straight after that. There's absolutely no dependency on Z-hop and retraction (I tried printing with and without them with exactly same results). It only happens after travel - all the rest of the layer is printed fine with no artifacts, so unlikely to be related to any bubbles or anything else related to PLA condition. I tried different PLAs as well.
I use Ultimaker Cura 4.7.1 as a slicer and Repetier Server running on Raspberry Pi as a host program.
Given that it only happens on certain types of lines, it makes me think that there might be a bug in the slicer or some settings, but I'm not sure. May be there's something that I can check in the generated GCode?
Have you by any chance seen any similar behavior or have any ideas regarding what could be causing this?
Thanks!
Underextrusion on curved parts after travel
Underextrusion on curved parts after travel
- Attachments
-
- Straight wall with no artifacts (while the cylinder was part of the same print had them)
- 05.jpg (166.57 KiB) Viewed 4334 times
-
- Underextrusion on the inner side (no retraction, no z-hop)
- 04.jpg (142.97 KiB) Viewed 4334 times
-
- Underextrusion on the outer side (no retraction, no z-hop)
- 03.jpg (161.4 KiB) Viewed 4334 times
-
- Underextrusion on the outer side (retraction, z-hop)
- 02.jpg (170.33 KiB) Viewed 4334 times
-
- Underextrusion on the inner side (retraction, z-hop)
- 01.jpg (183.2 KiB) Viewed 4334 times
Re: Underextrusion after low flow
Just posting an update to my old post, in case it becomes helpful to someone else.
Originally, I said that my underextrusion was happening on curved parts only, however later I observed it on other parts as well. After trying all sorts of alterations (firmware, filament, extruder, hot end, slicer, jerk settings, linear advance, cooling, retraction, speed, extra prime, flow) and not seeing any difference, I observed one distinctive feature which seems to be the key points, but which I don't know how to fix yet.
So the problem is: The underextrusion happens after long periods of small extrusions.
The test is:
1) make a cylinder and slice it with Cura, no infill, wall thickness 0.4mm (nozzle), no skin. It will print just a tube with one-line wall and it will simply do it printing ring by ring. This prints perfect (the left one on the photo)
2) make a module of a tube 0.4mm thickness (basically same shape), tick "print thing walls" in Cura and slice. This once prints really badly (the right print on the photo). All layers are screwed up expect the first one.
The difference is - in the second case, Cura always tries to make the second inner line with a reeeeaally low extrusion flow. And even though I know that the first way is the correct one, but it really highlights the issue in general.
It looks like this micro steps in extrusion commands accumulate into a floating point operation error inside the firmware and as result, when it comes to the actual extrusion, the E parameter is already too high and thus it skips some extrusion commands in the beginning of the next layer print.
The solution which I found in the Cura was "Minimum Wall Flow", which by default is 0%. I put it 50% for starters which seems to prevented Cura trying to print some lines which don't make the difference in the printed model, but screw up E position calculation.
Hope it might be helpful to someone
Originally, I said that my underextrusion was happening on curved parts only, however later I observed it on other parts as well. After trying all sorts of alterations (firmware, filament, extruder, hot end, slicer, jerk settings, linear advance, cooling, retraction, speed, extra prime, flow) and not seeing any difference, I observed one distinctive feature which seems to be the key points, but which I don't know how to fix yet.
So the problem is: The underextrusion happens after long periods of small extrusions.
The test is:
1) make a cylinder and slice it with Cura, no infill, wall thickness 0.4mm (nozzle), no skin. It will print just a tube with one-line wall and it will simply do it printing ring by ring. This prints perfect (the left one on the photo)
2) make a module of a tube 0.4mm thickness (basically same shape), tick "print thing walls" in Cura and slice. This once prints really badly (the right print on the photo). All layers are screwed up expect the first one.
Code: Select all
G1 X116.359 Y88.516 E39.6612
G1 X116.365 Y88.455 E39.66125
G1 X116.462 Y87.514 E39.66204
G1 X116.464 Y87.472 E39.66211
G1 X116.469 Y87.399 E39.66222
G1 X116.525 Y86.508 E39.66304
G1 X116.525 Y86.485
G1 X116.526 Y86.409 E39.66314
G1 X116.542 Y85.588 E39.66396
G1 X116.545 Y85.502 E39.66407
G1 X116.544 Y85.425 E39.66432
G1 X116.528 Y84.576 E39.66492
G1 X116.526 Y84.508 E39.665
G1 X116.525 Y84.49 E39.66501
G1 X116.522 Y84.436 E39.66508
G1 X116.465 Y83.521 E39.66573
G1 X116.462 Y83.486 E39.66575
G1 X116.365 Y82.539 E39.66647
G1 X116.359 Y82.484 E39.66653
G1 X116.225 Y81.559 E39.66733
G1 X116.214 Y81.486 E39.66736
G1 X116.049 Y80.598 E39.66811
G1 X116.029 Y80.494 E39.66817
G1 X115.832 Y79.633 E39.66883
G1 X115.804 Y79.513 E39.66892
It looks like this micro steps in extrusion commands accumulate into a floating point operation error inside the firmware and as result, when it comes to the actual extrusion, the E parameter is already too high and thus it skips some extrusion commands in the beginning of the next layer print.
The solution which I found in the Cura was "Minimum Wall Flow", which by default is 0%. I put it 50% for starters which seems to prevented Cura trying to print some lines which don't make the difference in the printed model, but screw up E position calculation.
Hope it might be helpful to someone
- Attachments
-
- IMG_20210225_194406.jpg (162.56 KiB) Viewed 2392 times