Print issue

[jens55]

I could use an opinion on a particular print issue I am having.
This is printed at 80 mm/sec using petg at 240C with 20% cooling. There are actually a number of issues in this picture but I am particularly puzzled by tiny gaps/pock marks in the print and the giant section of crap at the top middle/right of the model. The big crap issue doesn't happen anywhere else but the dropouts/pockmarks happen throughout.
I am not sure if this can be seen in the photo but the giant crapfest is missing a lot of filament and you can almost see through the wall in some places. .

Here is another picture showing a section of the inside wall towards the back which is even worse .....
note that there is crap on the right side of the inside wall that is just filament that collected on the nozzle edge being shed and that removes easily.

Thoughts / ideas ?
Sorry about the lack of focus ...

There is also a mechanical issue that shows heavily on the X axis but I am not concerned with that at the moment. I suspect a bad bearing in an idler pulley.

[cosmowave]

@jens55 said in Print issue:

I am particularly puzzled by tiny gaps/pock marks in the print and the giant section of crap at the top middle/right of the model.

Can it be that at this place are some retraction / unretraction moves?

[jens55]

@cosmowave, I don't see how this could be a retract/unretract issue as it is happening all over the print including the middle where you would not expect a retract or unretract. Z seam is set to 'sharpest corner' so nothing should pop up in the middle of a straight line.
Of course I might be missing something .....
Thanks!

[chrishamm]

@jens55 Check that the nozzle is 100% perpendicular to the bed, properly fixed, and that any cut PTFE tube is flush in the nozzle block. Maybe the same effect that causes the slight X shifts makes things worse in Y, too.

[jens55]

@chrishamm, I will double check. The hot end is a Dragon hot end so no ptfe tube involved (there is a tube but only to guide the filament into the feed gears).
What are you referring to when you say 'slight x shift'?

[jens55]

Here is another print with some back lighting showing the major print issue

I did find another area with that same kind of problem but only about half the size and on an inside wall and not getting all the way through the wall.

[Norder]

@jens55
This looks like a temporarily clogged nozzle.
You should clean the nozzle (change it if necessary) and the extruder (abrasion on the teeth).

How high is the retract set ?
Is it a direct extruder or bowden extruder ?
If the retract is too high, the filament will be pulled back too far... cools at the tip and can cause clogs.

You should also check the E-Steps after cleaning the extruder and the nozzle... are 100mm filament really moved if you enter 100mm extrusion in the DWC?

All mechanical issues also need to be fixed...loose screws, loose printhead, geometry (is everything fitted straight), belt tension, idler pulleys etc pp.

Dry the PETG filament to about 30%
However, other problems can also be the cause here, so you should proceed step by step.

Google Translate
-- Original Text --

Das sieht erst einmal nach einer zeitweisen verstopften Düse aus.
Du solltest die Düse (eventuell wechseln), und den Extruder (Abrieb an den Zähnen) reinigen.

Wie hoch ist der Retrakt eingestellt ?
Ist es ein Direktextruder oder Bowden-Extruder ?
Ist der Retrakt zu hoch, wird dass Filament zu weit zurück gezogen... kühlt an der Spitze ab und kann so Verstopfungen verursachen.

Du solltest nach dem Reinigen des Extruders und der Düse auch die E-Steps kontrollieren... werden wirklich 100mm Filament bewegt wenn Du im DWC 100mm Extrusion einträgst ?

Es müssen auch alle mechanischen Probleme behoben werden... lose Schrauben, lockerer Druckkopf, Geometrie (ist alles gerade montiert), Riemenspannung, Umlenkrollen etc pp.

Trockne das PETG Filament auf etwa 30%.
Hier können aber noch andere Probleme die Ursache sein, daher sollte man Schritt für Schritt vorgehen.

[jens55]

@Norder, retraction is 6.5 mm ( a bit high I think) at a speed of 25 mm/sec. The thing that I find curious is that multiple prints of the same object have produced the same flaw at the same point.
It is direct 'Dragon' extruder. I have a roll of filament coming out of the dryer this morning and will give that a go.

[cosmowave]

@jens55 Do you have these problems only with this object? Or generally with every print?

[jens55]

@cosmowave, I can't really answer that question. I tend to print a given object multiple times and then not print another object for a long time. In this particular case I am printing two objects, the case which has the problem and the lid which is totally different and does not have the same issue. The two can't really compare though.
I suppose I could try and make up some kind of a test object to print ....

I just started another print with filament that was dried over night....

[jens55]

The thing that I don't understand is why the problem shows up at the same spot in different prints. I would expect any sort of clogging issues to be distributed randomly. I would expect a model related issue if the same thing happens in the same print position in different runs.
Now I do have the pockmark issue and generally shitty print that can easily be related to something like a partially clogged printhead since it happens randomly.
BTW, I had the issue with two different makes of petg .... but again, I wouldn't expect the issue happening at the same spot in different prints.
I could see a over/under heat issue that somehow might occur at the same spot although I am having difficulty believing this as the issue only happens during maybe 20 layers and then reduces or goes away.
I will try to experiment with different printing temperatures. I will also bring the wall count up from 2 to 3 which would change the way the model is printed to see if that affects anything.

[jens55]

Dried filament (different filament) and the exact same two void areas in the exact same model. BTW, the voids are diagonally opposed from each other and both appear at a corner or possible 'start' position ie where the printhead did (or could have) done a rapid direction shift. Why would it only happen at these two spots if the exact same direction shift happened below or above the flaw.
Out of desperation I have installed PLA, reduced the temperature to 210C and left everything else the same.
Stay tuned .....

[gloomyandy]

@jens55 Check the gcode for the object and look to see what is happening at those points in the model. It sounds a lot like that is a place at which retraction is being used to me....

[cosmowave]

@jens55 Are you sure that the extruder is transporting the filament properly? Eventually it slips during some moves.

[mikeabuilder]

The same void or bump defect at the same place on two prints suggests to me some setting issue. And when I say "same defect" I mean really the same - photos of both are hard to tell apart - not just "crappy at this corner". And consistent wiggles on straight lines don't count and are probably mechanical.

But assuming really the same, I've had issues like that which were fixed by adjusting Pressure Advance and Input Shaping. Possibly also max acceleration and jerk settings as these affect the beginning and end of lines. Those things show up in places where the print head is accelerating or decelerating. I've used this part to investigate. It has some straight sections to allow max speed, some sharp corners to test braking and a lot of corners with decreasing radii on subsequent layers. I found that when I could print this well, I could print everything else. decreasing raduis part.stl attached part to

[jens55]

Thanks guys, lots of things to try out here.
I did check the gcode and found a possible connection between a whole bunch of very tiny infill moves and the printer then starting to print the inside and outside lines. I have just started a print in petg with the walls set to 3 which eliminates those tiny infill moves. I am looking forward to see what the results of this test will be.
I previously said that I tried printing the same model in PLA - as it turned out, the two spots where there had previously been a lack of filament were still present although much less noticeable. I also did some maintenance on the printer and hopefully this latest print will show an improvement in the overall surface finish.
With this issue happening after a large number of very tiny infill prints, I am wondering if there is a retraction/unretraction issue after all. I don't recall when and in which context but I seem to recall a previous situation where the unretract moves did not keep up completely with the retract moves causing filament starvation.
Just for curiosity sake - I have a retraction of 6.5 mm happening at a speed of 25 mm/sec. Unretract happens at the same speed. Does that seem to be viable setting or am I maybe unretracting too fast causing a tiny jam causing starvation?

[jens55]

Good news - the version of the model printed with 3 walls printed without the issues that the 2 wall model had. I will look into my retract settings to see what I can improve.

[Norder]

@jens55
The fact that this error always occurs in the same place, I believe it is due to the combination of the geometry of the component, the direction of movement of the nozzle and the Z-hop with retraction values ​​that are too high.
6.5mm with a direct extruder is clearly too much.
The retraction length on my direct extruder is set at 1.5mm.
I would not go beyond 2mm.
I would set the retraction speed to 35mm/s.
6.5mm is more of a value that fits a Bowden extruder with a 1m+ long Bowden.

The print on the photos doesn't look very nice either, it has a lot of artefacts (vertical lines) which are probably caused by the defective ball bearing of the idler pulley.

Also, the print looks like over-extrusion, these things could be the cause...

1. The entered diameter of the nozzle or filament in the slicer is not correct.
2. The flow in the slicer is set too high (over 100% or factor 1.0).
3. The E-Steps from the extruder motor are wrong.

If it's not overextruding, the nozzle could also be too hot for that filament.
It is best to take the average of the temperatures specified by the manufacturer. If it says 210°C to 230°C, then take 220°C.
If the print looks how it should look, then use test prints to determine the exact flow for the respective filament. But you should do a PID tuning beforehand, then print a heat tower to find out the ideal temperature for the respective filament and the E-Steps and all other settings (filament diameter, nozzle diameter etc.) should also be correct, otherwise has determining the flow makes no sense and only creates new problems.

The solution cannot be to switch from 2 wall lines to 3. Even with 1 wall line, there must be no holes.

Google Translate
-- Original Text --

Das dieser Fehler immer an der gleichen Stelle auftritt, da glaube ich es liegt an der Kombination von Geometrie des Bauteils, Bewegungsrichtung der Düse und dem Z-Hop mit zu hohen Rückzugswerten.
6,5mm bei einem Direkt-Extruder ist eindeutig zu viel.
Die Rückzugslänge bei meinem Direkt-Extruder ist mit 1,5mm eingestellt.
Ich würde nicht über 2mm hinaus gehen.
Die Einzugsgeschwindigkeit würde ich auf 35mm/s stellen.
6,5mm ist eher ein Wert der zu einem Bowden-Extruder mit 1m+ langem Bowden passt.

Der Druck auf den Fotos sieht auch nicht gerade schön aus, er hat viele Artefakte (vertikale Linien) die vermutlich vom defekten Kugellager der Umlenkrolle stammen.

Auch sieht der Druck nach einer Überextrusion aus, diese Dinge könnten die Ursache dafür sein...

1. Der eingetragene Durchmesser von Düse oder Filament im Slicer sind nicht korrekt.
2. Der Fluss im Slicer ist zu hoch eingestellt (über 100% bzw Faktor 1.0).
3. Die E-Steps vom Extruder-Motor stimmen nicht.

Falls es keine Überextrusion ist, könnte auch die Düse zu heiß sein für dieses Filament.
Nimm am besten den Mittelwert von den Temperaturen die der Hersteller angegeben hat. Steht dort 210°C bis 230°C, dann nimm 220°C.
Wenn der Druck so aussieht wie er aussehen soll, dann ermittle anhand von Testdrucks den genauen Fluss für das jeweilige Filament. Du solltest aber vorher ein PID-Tuning machen, dann einen Heat-Tower drucken um die ideale Temperatur für das jeweilige Filament heraus zu finden und die E-Steps und alle anderen Einstellungen (Filamentdurchmesser, Düsendurchmesser etc.) sollten auch korrekt sein, sonst hat das ermitteln des Flusses keinen Sinn und verursacht nur neue Probleme.

Die Lösung kann nicht sein, von 2 Wandlinien auf 3 umzustellen. Selbst bei 1 Wandlinie dürfen keine Löcher entstehen.

[Norder]

P.S.:
At 80mm/s and 240°C for PETG it should be more than 20% cooling.
With the speed and the not too large component that you are printing there, it can be 100% fan if the fan is normal.

There are also extreme fans that almost blow a component off the print bed at 100%. In such a case, 20% is probably better.

I usually print at 40mm/s and only use a fan for PLA and PETG for small areas (printing time under 10 seconds per layer) and for overhangs.
From 60mm/s, the fan always runs at at least 40% from the 2nd layer.
At 80mm/s it always runs at 100%.

[jens55]

@Norder, thank you for your thoughts. I have just tripped over a significant issue that I was not aware of. My extruder motor is rated at 1000 ma but I had set the config file for 800 ma. As it turns out, at just a tad under 5 mm/sec extrusion rate into the air, I start skipping steps. With the motor rated at DC input voltage, I could in theory go up to 1400 ma and still be within specs since I am feeding AC to the motor and RMS of 1400 ma is about 1000 ma. I am not clear on what happens if the motor is stationary but has not yet reduced to idle current (does the motor receive a DC drive voltage when stationary which could theoretically be 1400 ma which is well above rated current)
Anyhow, I have increased the stepper current (M906) to 1000 ma and that seems to solve some of the skipping issue. I am in the process of printing another copy of the model.
I am going to do another test with 2 mm retraction. What is your un-retract speed set at?
Yes, the printing artifacts have been bothering me for a long time. I took the time today to check each idler and clean any belt residue from the idlers. I also had a slight setup issue with one of the idlers. I did a print after the tear down and the artifacts seem to be significantly reduced. It is too early to tell for sure but I think that one of the things I did today significantly improved surface finish.
You misunderstood my thinking about going from 2 walls to 3 walls - what that switch did was to change the way the model is printed. I realize there shouldn't be holes in either scenario but doing the switch got rid of all the tiny infill steps and also completely changed the way the model is printed. Going to 3 walls, with the changed way the model was sliced, eliminated the two large holes.
Anyhow, I am making great progress and I am optimistic that both the underextrusion as well as the printing artifacts can be eliminated.

Ps.: I just looked at the print that is currently printing and it would appear that things have improved 'dramatically' !!!!