Skipped layers

[Mjolinor]

I don't see the problem at the place where holes start at all. If it is the steppers then you will just see a repeated pattern that is related to the steps per mm and you layer height.

Print a 10 mm square 50 mm high without any holes or sticky out bits or anything, see how that looks.

Print it solid then hollow then with say 15% infill. Do it as three separate prints, not all on the same print.

[Reaping miner]

I don't see the problem at the place where holes start at all. If it is the steppers then you will just see a repeated pattern that is related to the steps per mm and you layer height.

Print a 10 mm square 50 mm high without any holes or sticky out bits or anything, see how that looks.

Print it solid then hollow then with say 15% infill. Do it as three separate prints, not all on the same print.

Here are the testprints that show the skipped layers, but not show the most obvious flaw present in the extruder, which i believe is an unrelated issue.



I am now printing a 15x15x50mm bar at the same settings as before, except for the infill, only 15% instead of 30%. (10x10 is too small, that will print too hot. A nice stepping effect is seen as the area of the layers get smaller, especially in the negative control "pyramid". Again this is why i was printing the extruder body in the first place.)

It looks like the corners of the objects are curved, this is an artefact of the camera, the corners are straight, only a little bit of elephants foot is present. This is caused by the overextrusion at the start i mentioned before.

Also one arrow in the right "pyramid" points at a line in the Z axis at an area transition layer.

Some thoughts...
To eliminate one possible issue, try a print without heating the bed. I realize you have an aluminum heatbed and it might make a difference, but multiple people have demonstrated print artifacts being caused by flexing of the bed heater as it cycles on and off using the normal Marlin bang-bang heater control loop. Migrating to SSR and PWM control of the bed heater is the solution for those artifacts.

What are you using for a filament spool mount? The filament spool needs to unwind very smoothly. If it doesn't (like if you are using the MakerFarm scheme where the filament spool sits on a fixed bar), the filament might be tugging on the extruder until the spool springs free and rotates a bit to relieve the tension in the filament. IMO, the extruder mount on these printers isn't rigid enough to not be displaced by this tugging. Is your hot end a snug fit in the extruder body? A loose fitting hot end would make this worse. Many people have added a thin spacer above the hot end to provide a more solid fit.

I will give printing without a heated bed a try as well. I do want to add that i am actually printing on a piece of glass clamped on the aluminium heatbed (MUCH better adhesion of PLA). Also i do think i am using the fancy mathematical kind of BANG-BANG temperature control (PID?), not the dumb heat ON if its under and heat off when it reaches the temperature. I did notice wobbling/ribbing in the prints of extruder gears. This wobbling/ribbing is entirely different from the problem described in this topic though. That wobbling is much smoother and very regular. (Remarkably this wobbling seems to only be present on the gear surface and not the central structural bit holds the hobbed bolt and the nut.)

I am using a threaded rod with 2 nuts to lock it in place as a filament holder, mounted on top of the printer. The issues are not caused by the filament spool, i have had that problem before (filament unwound and got tangled), that causes more serious and irregular faults, it is also very audible when sitting next to the printer when it untangles or if it stays tangled the print will fail entirely and the filament will obviously be stuck when you look at it later. The hotend is fairly snugly fit. However the extruder (+carriage) itself is not as rigid as i would like. This will be replaced together with the extruder. I plan to print or design a carriage that can better clamp the IGUS RJ4Jp bearings. (I really dislike the zipties.....)

[Reaping miner]

Here is the bar, in my opinion, no big problems. (yeah layer alignment could be better i suppose. But this is nothing compared to the extruder body i showed before.

[printbus]

Moisture in the filament leads to steam bubbles that pop from the nozzle, often resulting in small bumps or zits in the print.

Re-reading the thread, I don't follow the height error you describe in post 1. The gcode snippet starts at a height of 3.65mm. A z-hop up of 0.75mm for moving to the next layer gets you to 4.40mm. The move to the starting location for the new printing layer includes moving Z back to 3.75mm, which would be the original 3.65mm height + 0.75mm z-hop hop - 0.75mm z-hop down + 0.1mm for the new layer height. How is 3.75mm the incorrect z-height for the new layer?

A new thought - PLA can tend to curl up at corners and edges - especially if the temp is a bit high or you're over-extruding. At your through-holes, perhaps this leaves a slight ridge that catches the nozzle on subsequent passes. This catching could cause a mechanical twitch throughout your x-carriage and extruder that leaves the nozzle a bit out of place until additional movement lets things settle back into place. Your 0.1mm layer height would be more susceptible to this than say a layer height of 0.2mm. Adding an effective print cooling scheme is very helpful in improving PLA print quality, and might be part of the solution you are looking for.

[Reaping miner]

Moisture in the filament leads to steam bubbles that pop from the nozzle, often resulting in small bumps or zits in the print.

Re-reading the thread, I don't follow the height error you describe in post 1. The gcode snippet starts at a height of 3.65mm. A z-hop up of 0.75mm for moving to the next layer gets you to 4.40mm. The move to the starting location for the new printing layer includes moving Z back to 3.75mm, which would be the original 3.65mm height + 0.75mm z-hop hop - 0.75mm z-hop down + 0.1mm for the new layer height. How is 3.75mm the incorrect z-height for the new layer?

A new thought - PLA can tend to curl up at corners and edges - especially if the temp is a bit high or you're over-extruding. At your through-holes, perhaps this leaves a slight ridge that catches the nozzle on subsequent passes. This catching could cause a mechanical twitch throughout your x-carriage and extruder that leaves the nozzle a bit out of place until additional movement lets things settle back into place. Your 0.1mm layer height would be more susceptible to this than say a layer height of 0.2mm. Adding an effective print cooling scheme is very helpful in improving PLA print quality, and might be part of the solution you are looking for.

That is the reason i am not doubting my slicer program. The G-code looks correct, but the nozzle seems to be physically higher than it should be. Later i looked at position of the Z-axis couplers to more easily see if the Z-height physically does change like it is supposed to, it looks like it does. (1/2 of 1/4th turn is 1/8th turn translates to a Z difference of 0.1 mm which is correct.).

I actually was printing the extruder body to be able to add upgrades (read: layer cooling fan.).

So i bought new calipers today (Digital, 0.01mm). My old calipers seem to be a piece of junk.... The diameter of the PLA (BQ, pure white) and ABS (FFFworld ABS++ black) are very similar. The measurements of both rolls were:
1.76mm on average over the roll, with mostly 1.75 plusminus 0.01 mm in 9/10 measurements, one outlying measurement at 1.8mm.

Surprisingly though, i did get overextrusion symptoms. I am going to check the Z stepper voltage first.

Update:
Upped from 2.2A to 2.5A, still "skipping layer".

Update 2:
Switched to 0.4mm nozzle, same settings: issue persists but seems to be dampened or masked a whole lot better. At least this is more useable to print parts for things such as cooling solutions.

Update 3:
Switched to 0,2mm layer height, same settings: seems a little better than update 2. Infill seems to have improved a lot.