Layer shifts

[berga 0]

On 4/30/2021 at 5:31 AM, foo_bar said:

which settings did you use for the ship?

I'm still facing problems with the original vase gcode from sd, I tried to tighten the Y belt (1-2mm) but the drifting seems worse than before:

 

 

any hint?

Thank you

 

[gr5 2,093]

I created a new topic for you as your post was off topic for where you posted.

 

You don't want to tighten the belt - if anything that could make layer shifting worse.  Instead you want to tighten the set screws on the pulleys.  tighten the hell out of those.  Usually it's the pulley on the stepper itself.

 

What is happening is called a layer shift and can be caused by missing steps on your stepper (quite likely) or a loose pulley (also quite likely).

 

You can figure out which by using a sharpie to mark the shaft of the stepper and the pulley to see if it moves over time during a print that has major layer shifting (like the picture above).

 

If it's losing steps then you need to reduce friction or lower the acceleration for the axis that is slipping.  Maybe add a drop of oil or something to the bearings/rods.  What kind of printer is this?  Is this the Y axis?  Does your printer have a moving print bed?  What is the acceleration for that axis?  It's set on the printer itself - not in cura although cura can possibly also set the acceleration but different printer types use different commands to set this parameter so I'd do it on the printer, not in cura.

[berga 0]

Thank you, my printer is Selpic Star A, as mr. FooBar one's.

 

When I tried to tighten the pulley I noticed that some of the bed screws were too loose so I tightened them a bit.

I decided to do so to fix the bad result (1st vase image) and because mr.foobar suggested to do so in a post.

 

I really didn't know how much I should have tightened the pulley (the regulation appear under the bed) but it seems to be a bit looser than the X axis pulley, I'm going to check all and update the firmware to Marlin 2 that allow to change accel params without reboot, the current values I used are the default by vendor fw.  

[gr5 2,093]

There's confusing between "tightening a belt" and "tightening a set screw on a pulley".  You want to tighten the set screws that keep the pulley from slipping.  But again - it could be that the stepper just can't move all the wait of the print bed and you need to lower the acceleration.  Very easy to do in the menu system on the printer.  If it's at say 1000mm/sec/sec then try 500mm/sec/sec.  Just cut the existing acceleration in half and see what happens.

 

Also maybe ask other people with the same printer what they set the Y acceleration to.

[berga 0]

Finally something is going nice:

the short story: pulley and belt check and firmware Marlin 2.0 by Mr. foo bar

the long one: I've bought the printer from kickstarter with optionals (heated bed, lcd and laser), the vendor told to update original firmware to use heated bed, after that all the trials to print the Rocket gcode weren't successful although the vase gcode were printing but badly.

Now I ran out of the white spool and I printed the vase with the wooden filament (1kg with a bad spool holder):

 

[GregValiant 1,141]

Progress is good.  Now that you have a handle on the layer shift in the X Y you get to go to school on Z banding.  There looks to be spots where the Z didn't respond correctly and the result is layers that got mashed together.  It's very noticeable at the wide portion of both vase models.

[geert_2 557]

I don't know your printer, but to see if there are no *mechanical* blockings, try switching it off, and move all things by hand. They all should move smoothly, without requiring brute force, and they should have no hard stuck points. In this way, you can often feel if something is wrong: broken bearings, incorrect axis-alignments, gears or belts blocked by dirt, too dry rods and linear bearings, too much play, etc... Things should have a precise fit with no play, but still move quite freely and smoothly, similar to cameras and microscopes, or like the mechanics in car engines. It are precision instruments. In this way you can separate mechanical from electrical/electronical issues, hopefully.

 

[jahaskalja 0]

I'm very novice with Selpic Star-A or printing in general. Actually I'm trying to get my first model printed, after "the vase".

I had very bad constant shift on x-axis: 11mm per 8mm of z-axis for a wall that should be straight along y and z. 

From bottom up the three layers/steps (paused, drove home between them) are:

1. Device assembled as instructed, except:

 -  extra belt head inside printing head cut away

 - cardboard box of 1kg spool, 2 holes cut and felt pen pushed through works as spool holder

2. Cable sleeve moved so that it doesn't go behind vertical rods, but partially lays on floor on left of the main structure

3. I cut the filament from spool, allowing about 20cm to stand/swing on air

 

My preliminary conclusion is that major problem is to drag the filament away from spool. Reasons can be inertia of the 1kg wheel or it's sides being dragged against cardboard. To mitigate the first, I guess either smaller spool or slower x-motion or acceleration would help, for latter a better spool holder.

 

To be continued whenever wifey allows... 🙂 

[GregValiant 1,141]

If the spool was mounted above so the filament feeds straight into the extruder it might be a better look.  With the spool below (like in the Selpic ads) when the print head moves outboard in the Y it looks to be a definite problem.  I think it needs some type of Strain Relief to eliminate the sharp turn into the extruder.  Mounting the spool up high might be the easiest fix.

[jahaskalja 0]

5 hours ago, GregValiant said:

If the spool was mounted above so the filament feeds straight into the extruder it might be a better look.  With the spool below (like in the Selpic ads) when the print head moves outboard in the Y it looks to be a definite problem.  I think it needs some type of Strain Relief to eliminate the sharp turn into the extruder.  Mounting the spool up high might be the easiest fix.

Pre-emptively I've got this Ikea overhead guiding holder for Clas Ohlson spool holder already before Selpic was published 🙂  I should have results within two hours, will edit to this message.

 

[GregValiant 1,141]

I like it - but you've still got a bit of a kink where it passes through the IKEA guide hole.  If it works you can print something up.  With some thought it may be something that will allow you to put the printer on a table.  Getting down on your hands and knees to level has got to be tough.

[jahaskalja 0]

On 9/7/2021 at 9:04 PM, GregValiant said:

[..]Getting down on your hands and knees to level has got to be tough.

Got my first nozzle jam and distracted from there, so it took a while to get the results. There was improvement, but constant drifting. At some point I went and gave some loose for the filament, and without driving home/recalibration of position, the shift jumped back. So maybe the the structure bends of belt stretches or something, jumping back is hard to explain otherways.

And you're definitely right about getting onto floor, I also have some sort of nerve entrapment and my right hand aches all the time so getting down and up are not something I like to do 😄

[GregValiant 1,141]

This is a mechanical issue but it's possible there is something you can do.

I'm attaching a calibration shape that I came up with.  You can bring it into Cura and set the print speed to 20mm/sec and enable acceleration and Jerk control.  Set all the acceleration numbers to 100 and the Jerk to 5.  Generate a Gcode and print the file.

 

I've decided to also attach a gcode file of that shape (I scaled it a bit).  I used my Ender 3 Pro as the printer and after the 3rd layer it just does the outer walls.  I moved the model to near the LF corner of the build plate and I think it should print on your Selpic.  If it doesn't then just abort it and print one of your own.  I set up Cura for a slow print with no banging around.  According to Cura the print time is 1:11.

 

If it is still drifting and shifting at those low numbers then it is certainly mechanical.  Make sure your filament feed system allows scope for the print head to move back and forth as well as freedom of movement of the filament itself.

GVCalibration_Shape.gcode GVCalibration_Shape.stl

Edited September 11, 2021 by GregValiant