Skirt (More precisely, this is a border for additional adhesion of the printed model to the build plate). I stopped the printing manually to make it easier to take a photo. In the slicer, this skirt is straight everywhere, but the printer head always prints it with this defect (a slanted joint). Moreover, the printer head jerks to do this.
I’m interested in understanding whether this is the case for everyone or just for me? This is the beginning of printing the calibration model available in the slicer itself for the volumetric flow test; you can check it on your own, if you have the opportunity.
Yes, you’re right. If you continue printing, after this border, the model itself prints evenly and without that slanted connection. But I don’t understand why the printer head deliberately prints the border this way (with a slanted connection) in the first place. Although, in the slicer, the border is also straight, and there are no such constructions.
Hello,
To me, the appearance of the skirt doesn’t matter at all.
Its purpose is to improve adhesion. I never look at the skirt or the brim when they’re doing their job.
Especially if the rest of the model is good.
JMO
The offset near the radius corner looks like a hardware problem, particularly if the closeup view of the first layer in the slicer doesn’t show the same jagged offset. It looks like the result of backlash in the motion control system. I’d check the following, in this order of liklihood:
Loose idle pulley set screw (aka grub screw) on one of the two stepper motors on the left and right sides on the rear of the gantry. I’d check to see if a screw is loose and allowing the timing pulley to spin slightly before rotating in the opposite direction whenever that motor changes direction, but I’d also back each set screw out, apply non-permanent blue Loctite and torque the set screw to slightly below what would strip the head. Bonus points for replacing these set screws with the best quality set screws you can buy (McMaster-Carr, Fastenal, etc.)
Check the belt tension on those belts. If the belt tension is very loose, the same backlash could occur, similar to the timing pulleys spinning slightly on the drive shaft. The belt tensioners are on the left and right front of the gantry.
Cardoc already provided the needed info on the localized bed adhesion issue that caused the straight corner cutting with one filament on the corner radius. In addition to cardoc’s suggestions, you might also want to increase the slicer’s extrusion width for the first layer to extrude slightly more filament on the first layer only, so the hot filament is pressed more firmly on the critical first layer. If you’re seeing evidence of filament not adhering in a corner, you may have weak bed adhesion for the entire layer even though the filament is deposited where it should be. The first layer needs to be squished into the bed - not too much, but not too little. Too much would cause nozzle grooves embossed in the first layer, blobs of excess plastic on the first layer, rough surface and feed issues when the nozzle is effectively clogged by the build plate being too close that possibly causes the extruder gears to grind the filament that it’s unable to push through the nozzle. Too little extrusion on the first layer results in printing slightly above the build plate with no adhesion or poor adhesion, which can cause the part to detach later while it’s printing.
Thank you for the tips! I’ll run a few more tests after making the adjustments. Right now, I hardly have time to work with the printer — I have a lot of routine and other work to do. I’ll provide feedback once I’m done.
I agree with this. What worries me is the very fact that the printer is deliberately doing this, even though there is no such build route in the slicer.
This behavior may indicate some kind of malfunction, which could later affect the overall operation or lead to other errors. That’s why I started wondering whether everyone prints the border like this on a Sovol printer, or whether this is my unique case.
The set screw is used to secure the timing pulley onto the shaft of the stepper motor. It isn’t an adjustment screw.
The belt tensioners at the front of the gantry are used to adjust the two belt tensions. Simply loosen the mounting screws, pull back on the tensioner and tighten the mounting screws. You can pluck the belts like guitar strings. They should both make approximately the same pitch. Mine are probably 200-300 Hz.
In general, after resetting all settings and parameters, and after a slight adjustment of the z‑offset, the printing improved. However, the printer’s strange behavior when printing the border remained unchanged, and the same problem is still observed.
I’m curious to know whether everyone experiences the same thing when trying to print a test model in the slicer’s calibration section, or whether this feature is unique to me.
@Liberty4Ever @cardoc
I also have a question…Could such defects or incorrect behavior of the printer’s print head be caused by incorrect shaper settings?
I got some rather strange values during automatic calibration using an accelerometer.
## [input_shaper]
## shaper_type_x = 3hump_ei
## shaper_freq_x = 64.2
## shaper_type_y = mzv
#*# shaper_freq_y = 37.6
Perhaps I should create a separate topic related to this parameter? As I understand it, this is a very important calibration, but I don’t really understand it well. Even after watching training videos and reading articles, I still have some questions… On this forum, I didn’t find a topic where this knowledge was explained.
The issue you’ve circled is normal and expected. Even if pressure advance is perfectly tuned the “seam” where one parameter joins itself is always going to leave a visible defect.
As the nozzle approaches the begin point it:
The extruder motor decelerates
The X,Y motors decelerate
The extruder motor stops
The X,Y motors stop. Where they stop is determined by “seam” settings but is some distance away from the start location to account for oozing at the start and end points.
The X,Y(Z) motors jog the nozzle to the start point of the next parameter line. Z moves are optional and a result of Zhop settings.
The extruder motor accelerates
The X,Y motor accelerate
The E motor reaches print speed and stops accelerating
The X,Y motors reach print speed and stop accelerating
The width of the OOZE defect is magnified on the first layer. Higher layers droop into the layer bellow and subsequently not as wide.
Study pressure advance. There is a disconnect in the time domain between E moves and actual flow from the nozzle.
Also take onto account that moves are the center POINT of the nozzle. The extruded plastic is a 3d SHAPE that HOPEFULLY is centered about the target point. Nozzle position and flow rate needed to cause the new extrusion to fuse flawlessly to the adjacent extrusion require some “Guess work”. The initial layer is a case all it’s own and optimizing it will negatively effect layers 2-10^6.
I don’t think that jagged line is a shaper artifact. I’ve seen similar jaggies on the skirt on my prints but didn’t worry about it because that’s not a defect on the part. It may be an artifact of the G code created by the slicer for the skirt that doesn’t compensate for anything to produce a good print because the skirt doesn’t matter. There is some motion planning magic behind the scenes to make the best prints possible, compared to the earliest 3D printers that did their best to move the print head where it’s commanded in an open loop stepper motor motion system without input shaping, etc. The early printers were very slow and still produced poor quality prints.
I don’t see such a defect in the part itself. You’re right, the skirt construction doesn’t affect the model construction.
Apparently, I’m irritated by the slight vibration of the print head when it creates this defect on the border (edge/skirt) of the printed model. But if it doesn’t affect the model itself, then I don’t need to worry about it.
Thank you for your help!
