Hey all,
I have been doing calibration print after calibration print trying to get things perfect. I've never gotten holes to be the right size, so that is what I was working on when I found this problem. I made a simple calibration piece with four holes and four pegs. I'm getting a consistent row of blobs in the holes.
It's not wipe, retraction, z-lift, extra length on restart, avoid crossing perimeters, or anything else I could think to turn off; the offending move is still present when I turn all of that off.
Here's a good example of what happens in the gcode
This example has the retract (but not wipe) enabled to show why this is a big problem. On the problem line, the extruder moves part of the way out into the hole and allows plastic to ooze out forming a blob before it has a chance to do the retraction. The feedrate matches what I have set for non-print travel, but I don't know why it chooses to move there into hole before it does the rest of the things. It could be that Slic3r is trying to move the print head into the part before doing retraction but is mistaking the outside for the inside.
I can attach files and pictures if that would help. I'm thinking this is worthy of a bug report. Hopefully someone here will have an answer instead.
EDIT: Commands very similar to the problem line seem to exist after most (all?) connected paths before non-extruding moves. Most of the time they move the print head into the part (good) but in these holes, the moves should be in the opposite direction.
I have been doing calibration print after calibration print trying to get things perfect. I've never gotten holes to be the right size, so that is what I was working on when I found this problem. I made a simple calibration piece with four holes and four pegs. I'm getting a consistent row of blobs in the holes.
It's not wipe, retraction, z-lift, extra length on restart, avoid crossing perimeters, or anything else I could think to turn off; the offending move is still present when I turn all of that off.
Here's a good example of what happens in the gcode
G1 X87.448 Y83.330 E1.59741 ; the last move of the arc that makes up a hole I'm printing - head is at the 270º point in the circle G1 X87.662 Y83.634 F12000.000 ; THIS IS THE PROBLEM LINE G1 F2700.000 E1.09741 ; retract G1 Z0.400 F12000.000 ; lift Z G92 E0 ; zero extrusion G1 X87.583 Y83.007 ; non-extruding move to the next area to print
This example has the retract (but not wipe) enabled to show why this is a big problem. On the problem line, the extruder moves part of the way out into the hole and allows plastic to ooze out forming a blob before it has a chance to do the retraction. The feedrate matches what I have set for non-print travel, but I don't know why it chooses to move there into hole before it does the rest of the things. It could be that Slic3r is trying to move the print head into the part before doing retraction but is mistaking the outside for the inside.
I can attach files and pictures if that would help. I'm thinking this is worthy of a bug report. Hopefully someone here will have an answer instead.
EDIT: Commands very similar to the problem line seem to exist after most (all?) connected paths before non-extruding moves. Most of the time they move the print head into the part (good) but in these holes, the moves should be in the opposite direction.