Any advice on strange issue?

[Waterboy]

I’m reaching out to the community for some guidance and/or advice as to which path to further explore. I’ve been having some “issues” that I have been unable to resolve.

My setup:

• A350T with single extruder, firmware 1.18.1 (tried updating to latest firmware unsuccessfully, more details below)
• Slicer: Orca Slicer (tested with Prusa, Cura and Luban as well)
• Connection: Raspberry Pi4B with Octoprint (but tested prints by uploading Gcode to controller and printing from controller as well)
• Filament/s: eSun PLA+ Orange and Black but also witnessed issues on X3D PETG
• Other:
  • Linear Rail Mod has been installed
  • Printing from Eibos dryer via PTFE tube to extruder

Issue description:
My prints started exhibited some “strange” blemishes after fair bit of relatively seamless printing. As part of organising my setup I’ve been printing a large number of Jerrari’s Gen2 drawers ( Gen2 Modular Lack Enclosure Drawers by Jerrari | Download free STL model | Printables.com), Figure 1 below shows setup, which also contains some of the own designs I adapted to fit his system, for those that are looking for something, I cannot recommend these enough.

Figure 1 - Gen 2 Drawer system

I was about 75% through when 2 of the orange large tilt drawers ( Gen2 Large Modular Tilt Drawers by Jerrari | Download free STL model | Printables.com) started showing this strange pattern that I have not witnessed before. What I found stranger was that it was in the middle of the print with previous layers really good, and on one of the drawers it appears to “resolve itself”. See Figures 2 and 3 for the first drawer, and Figures 4 and 5 for second drawer:

Figure 3 - First drawer outside

Figure 4 - First drawer inside

Figure 5 - Second drawer outside

Figure 6 - Second drawer inside

I suspected that my enclosure being closed during the Australian summer temperatures resulted in too much heat, started keeping the enclosure open, continued printing the next drawer and the issue disappeared.

At the same time I noticed some strange seam issues which I could not explain either, but at the time I just ignored it. I say strange because the best way I can explain it would be to say that directly following the seam the extruder appeared to be under-extruding, but would then “catch-up” and print to a very high level of acceptable again, as the seam was hidden in a location that I would not see, this did not immediately bother me too much. See Figure 7 below which shows what I am trying to convey:

Figure 7 - Under extrusion following seam

I am currently finding myself in a rare situation where my wife has asked me for a print, and I feel this is an opportunity to redeem the cost of the printer by providing her with something from it. Last night during my test print, the seam issue presented itself again, see Figure 8 below.

Figure 8 - test print with seam issue

What have I tried
Just to cover off the basics:

• E-step calibrations has been done, but I am thinking of doing this again just to confirm. However, the fact that the “rest” of the print is good without signs of under or over-extrusion leads me to believe this is not the issue.
• Flow, temperature and linear advance calibration has been completed using the built in Orca Slicer calibration.
• I am also aware that in the test print shown that I can “hide” the seam in a sharp corner, but I would like to resolve, understand what is happening here

So I went about trying to first replicate the issue and then playing with my slicer settings (originally believing this is where the issue lies) to see if I can resolve it:

1. Printed a hollow cylinder of 1.2mm thickness - To my surprise this had no issues, and following later tests I believe this is because there was no retractions part of this “test”:

Figure 9 - Test cylinder with no seam flaws (at least not the flaw of this topic)

2. I created a reduced version of the original test print (from Figure 8) which is just the upper part (cylindrical shapes) with seam position as “Back”. First print was with my normal settings to use as baseline and confirm the issue is present and then I started making changes. Test file attached for those interested:
   Seam_test.stl (1.2 MB)
   Gcode of seam test file with “issues”:
   Seam test_default.gcode (3.1 MB)

Figure 10 - Multiple test prints

The below is what I tried without avail:
a) Arachne wall generator is my standard, changed this to Classic → The issue appeared to be less prominent but still present
b) “Wipe before external loop” is on, this was turned off → No change (changed back to Arachne)
c) Reduced Acceleration in the speed settings of Orca to be, in some instance half the default → Issue reduced but still present
d) Other settings such as retraction distances and speed also tested also with limited success.
e) Changed to Prusa slicer instead of Orca - Still present. Both Prusa and Orca files print from Octoprint

– At this point I started noticing (I guess I should have mentioned I always print with 2 line skirt loop 5mm from print to ensure first layer is still good) that there is a portion of the first line of the skirt loop not printing.

f) Changed to Luban and Cura to see if I can replicate this and I could. Gcode sent to controller and not printed via Octoprint. I manually inspected the Gcode via text editor and could not identify issues, the code retracted and extruded the correct amount (and at the right time) according to what i can see.

Figure 10 - First part of skirt not printing (top left next to ellipse, print direction is counterclockwise)

So with all of the above in mind I started to think that it might not be a software/slicing issue but rather a hardware/setup issue.

• First I removed the filament from the dryer and PTFE tube, installed on my internally mounted bearing filament roller, and did another test print - no issue. My thinking was that there might be too much resistance for the extruder to pull the filament smoothly, but this does not seem to be the case.
• I started wondering (noting my issue is only present after a retraction and deretraction) if the extruder (or controller was having issues with these commands following shortly after each other. I wasn’t sure how to test this, so I removed the filament and issued (after heating the nozzle) extrude and retract commands from Octoprint. I noticed a significant delay between when the command (for either) was issued and the extruder gear moving. Trying to rule out Octoprint as the issue, movement commands had no perceivable delays, only extruder commands. I then used the Snapmaker controller and the “load filament” and “unload filament” commands but no delay was observed.

At a loss of what to do next I tried updating the firmware:

• The controller became “stuck” and the app had to restart multiple times when I inserted the USB with the latest firmware
  

  
  
  Figure 11 - SnapMaker controller issues

• Power cycled and was able to select firmware, and then controller became “stuck” in the “updating” loop, which is where I am now…

Figure 12 - Controller “updating for past 2 hours”

Experiencing the issues above with the controller has reminded me that in the last couple of months I’ve been experiencing random controller disconnections, where OctoPrint would loose connection in the middle of a print, and the print head would just sit there with its nozzle at 215deg C, or where I would get the SM controller and I would have an error, something along the lines of “Connection to machine lost” and a button on the screen to “reconnect”

So noting all of the above, I’ve started wondering if my issue might be controller related, where it either has a poor connection (I have plugged and unplugged to check but couldn’t see anything) or if the controller itself might have an issue that is causing it to stall, delay to react to “instructions”, specifically when it comes to retraction and that has been part of my issue. However as i am not sure what to explore next (or if I’ll be able to get my controller out of the “updating” loop) I am not sure where or what to test next and would appreciate some guidance/advice.

Thank you in advance

[Hauke]

Regarding the update issue: It may be that only Snapmaker can help here - worst case something is bricked.

Regarding your underextrusion issue: Did you check the mechanics, i.e. the feeder wheels? They can wear out, losing grip, which may cause all kind of effects. Open the flap and look closely on the teeth of the gearwheel feeding the filament - does it show bumps in the teeth?

And I seem to understand that you use retraction. In my case, I add some “extra prime amount” after retractions, because during nozzle travel I find it loses pressure, and adding extra prime amount compensates for that. Only thing that makes me think that this will not help is that you say that the effect was not there before… So I guess that might not be the issue here.

When I saw your picture of the orange drawers, I immediately thought: Overheat problem. It seems you came to the same conclusion.

[clewis]

I get some under extrusion near the Z seam when I’m printing as fast as I can. 0.32mm layer height at 80 mm/s leaves a thinner wall for about 1 cm after the travel. If I lower the layer height to 0.2mm, it’s only about 0.25cm. If I slow down to 60mm/s, it’s hard to notice at all.

Your wife’s part looks small enough that slowing down shouldn’t extend your print time too much. That should let you finish that print while you spend time dialing in your retraction settings. Those drawers are big enough that you don’t want to print them slowly

[Waterboy]

Thanks for that:

• I’ll reach out to support as per your suggestion, I’ve left the “updating” screen overnight just to ensure I wasn’t being too hasty.

• Under-extrusion issue - I have yes, in the portion where I mentioned the delay between issuing the retraction and extruder commands from OctoPrint. From what I can see the wheel is neither worn nor dirty that would cause any type of “slippage”
  

  

• Extra prime amount - to be honest I did consider this, but in another post on this forum, which I am not sure which one right now, if I find it again I will link it, another user made a comment along the line of “if you need to do this you have something else wrong that you need to fault find, this is a setting that should not really be used” which is why I have refrained from testing this. If I am able to get use of my SM back, can you advise how you went about determining the “amount” to add to the extrusion?

• Overheat issue - At the time I thought so as well, I am only starting to doubt myself now following the other (retraction and potential controller) issues if there isn’t something else at play, but I think in that case you are right, it was only heat related and opening the enclosure doors seemed to have resolved this.

Appreciate the time you’ve taken to review

[Waterboy]

I am not saying this isn’t my issue, and I will attempt to reduce speed, however I am not printing anywhere near what I consider to be the machine’s limit. 98% of my prints are 0.2mm first layer and 0.16 for all layers after that, and my print speeds are “moderate” I would say, definitely nowhere near what you are doing (volumetric wise):

Speed settings:

Layer height and line width:

Unfortunately the part for my wife was only 1 piece in the puzzle, and was printed as a test to ensure my measurements and tolerances where good, the overall print is much bigger and will take some time. If I have to slow it down to get the quality I’ll do it, however I am not sure speed is the issue, or if it is, it is only directly following a retraction and extrusion sequence, after which there is no issue. However what I am picking up from both @Hauke and your comments is that I am not the only one experiencing this, which although it doesn’t resolve my issue (yet), does provide me with some comfort.

The wife’s design, for what it is worth, is a nail polish station to house all of her tools and nail polish, which consists of 2 drawers, stands for all the different nail polish (I only drew one of each variation of what she’ll need) and a station on the right for all the tools and liquids she uses, and the intent is for it all to slot together so it is one big unit that can be moved where she needs, so unfortunately it won’t be a “small and quick print”, the part shown was just one of the shelves of which there are 12…
:

[Blockmodule]

Advice from Japan.
Is the bearing in the linear module dead?
Is the X-axis of the linear module loose?

I have had multiple occurrences where the molding has looked like your picture in heavy use over a considerable amount of time.

In each case, I was able to address the problem by disassembling the linear module and fixing the play.

However, it is only a wall issue.

I think the rough skin at the seam joints is a different problem.
Perhaps it is insufficient extrusion when it comes to the next layer of molding.
Might try replacing the nozzle kit.

[Waterboy]

I did give all the modules a feel to try and see if they have any play but couldn’t notice anything (more than normal), but that being said, I need to move house within the next month and was thinking of doing the full maintenance on all the modules before the machine is assembled at the new house, and by that I mean to open all, clean the rails, install new grease, check bearings and adjust as/if required.

As for the controller “stuck” on the updating screen, support advised to:

1. Turn off machine
2. Unplug module head
3. Power machine on
4. Re-attempt to update firmware from USB
5. After firmware power off, reconnect module head and power on

To be honest I just followed steps 1 and 3, the good old turn it off and back on again, and the “new” firmware had been updated already. Thus I guess as a summary:

1. Orange (rough) skin - the concensus seems to be heat, for me it worked to have lowered the bed temp for PLA from 60 for the entire print to being 50 for first layer and 45 for the rest but also keep the enclosure doors open when possible, and lastly I’ve added a start G-code to all my PLA filament profiles to start the enclosure fan on print start and turn it off on print end to assist with cooling if the doors cannot be left open (noise), which also seems to have helped.
2. Under extrusion after seam - the jury is still out and I am continuing testing, however to suggested solutions thus far:
   a. Add extra extrusion after retraction
   b. Lower speed of printing
   I will try these and continue to update for future reference if I find anything
3. Controller stuck - see first paragraph on advice from support

Thank you again to all, your advice and time is appreciated.

[Waterboy]

Following some further testing I am starting to form a “new” hypothesis that I would like to sound board with the group.

I think it is group member @xchrisd that once said, in another post on the forum, if you don’t know what is going on “…just watch your print to see what it does as it prints…” so this is what I did and after which I noted something “small” and I started looking at the Gcode at these “small” instances, and from this my question - What is the minimum or smallest G1 E value the printer can execute?

Using the same Seam test file posted in my original post, I watched the print. I noticed that in between where the cylinders are linked the printer appears to just move there and then do nothing and then moving to another spot, this happens (on my print) a couple of times before the each of the seams with the issue. So, I moved to the GCode and I think I found the below:

Using a random layer of 21, because it has both an inside and outside seam with issues

Figure 1: Layer 21 of test file at 0.16mm layer height and strange seam artefacts

Originally I just looked at the GCode at the location of where the seam was and what was following the seam however, as can be seen at the bottom of Figure 2 (the GCode section) there is nothing “strange” there and the “E” values was similar to areas without issues. So I started looking at the GCode leading up to the seam, and I refer you to Figure 2 as a picture speaks a thousand words.

Figure 2: Locations of very minimal printing before seam

As mentioned earlier the printer, when viewed, seemed to “do nothing” at point 1, then move to point 2 and do nothing again, before moving to point 3 and then under extruding. Viewing the GCode, the inner wall print (yellow) is so small that it is not even indicated at points 1 and 2, and only the retraction (purple/pink) and de-retractions (light blue) are shown. Viewing the “Printing at Point 1” and “Printing at Point 2” portions of the GCode, it can be seen that the “E” values are a order of magnitude smaller than “normal” printing (GCode shown in Figure 2):

• Point 1 E-values: 0.00094 - 0.00163
• Point 2 E-values: 0.00096 - 0.00167
• Outer wall E-values: >0.02

Noting this, and this is where my hypothesis commences, if these E-values are too small for the printer to handle (potentially limited to the single extruder), the extruder will ignore these “print” instructions and instead do nothing (in my mind, the Stepper motor does not have a step small enough for this instruction).

IF the above is true then the sequence in Figure 2 above can be summarised as:

1. Do nothing at point 1
2. Retract 1mm and move to point 2
3. Push out 1mm at point 2
4. Do nothing at point 2
5. Retract 1mm and move to point 3
6. Push out 1mm at point 3
7. Start printing at point 3

However as the printer is using “Relative E values” potentially the controller is not seeing the motor as doing nothing, but rather sees it as:

1. Extrude 0.0042mm at point 1 (the summation of previously “ignored” E-values)
2. Retract 1mm and move to point 2
3. Push out 1mm at point 2
4. Extrude 0.0043mm at point 2
5. Retract 1mm and move to point 3
6. Push out 1mm at point 3
7. Start printing at point 3

In the example above, the difference between the controller and what really happened at the extruder is approx. 0.0085mm. And my theory is that this is the “under extrusion” witnessed following the seam at only certain, or specific instances, but not all as it is dependent on the amount of “too small to print” steps leading up to the seam and is potentially aggravated by the retractions and de-retraction.

It is difficult to show the image and the GCode for the seam at the inner wall of the Cylinder (not the inner wall of the print) but the “same” occurrence happens here, refer to Figure 3, but in this instance, there are 4 instances of the same sequence happening rather than just 2 as in the example above before it starts printing the wall on the inside of the cylinder at point 5:

Figure 5: Sequence of events before inner seam with problems.

GCode for the above sequence below:

;WIPE_START
G1 X193.956 Y179.229 E-.45
;WIPE_END
G1 E-.55 F2700
G1 X159.866 Y178.591 F6000
G1 E1 F3600
;WIDTH:0.37962
G1 X159.801 Y178.629 E.00172
G1 X159.866 Y178.666 E.00172
G1 X159.887 Y178.629 E.00099
;WIPE_START
G1 X159.866 Y178.666 E-.03237
G1 X159.801 Y178.629 E-.05607
G1 X159.866 Y178.591 E-.05607
;WIPE_END
G1 E-.8555 F2700
G1 X124.956 Y177.446 F6000
G1 E1 F3600
;WIDTH:0.519692
G1 X124.896 Y177.446 E.00196
G1 X124.865 Y177.499 E.00196
G1 X124.896 Y177.551 E.00196
G1 X124.956 Y177.551 E.00196
G1 X124.987 Y177.499 E.00196
;WIPE_START
G1 X124.956 Y177.551 E-.0455
G1 X124.896 Y177.551 E-.0455
G1 X124.865 Y177.499 E-.0455
G1 X124.874 Y177.483 E-.0135
;WIPE_END
G1 E-.85 F2700
G1 X124.959 Y172.424 F6000
G1 E1 F3600
;WIDTH:0.527654
G1 X124.898 Y172.424 E.00202
G1 X124.867 Y172.478 E.00202
G1 X124.898 Y172.531 E.00202
G1 X124.959 Y172.531 E.00202
G1 X124.99 Y172.478 E.00202
;WIPE_START
G1 X124.959 Y172.531 E-.04625
G1 X124.898 Y172.531 E-.04625
G1 X124.867 Y172.478 E-.04625
G1 X124.874 Y172.465 E-.01126
;WIPE_END
G1 E-.85 F2700
G1 X159.87 Y171.331 F6000
G1 E1 F3600
;WIDTH:0.3745
G1 X159.806 Y171.368 E.00167
G1 X159.87 Y171.405 E.00167
G1 X159.891 Y171.368 E.00096
;WIPE_START
G1 X159.87 Y171.405 E-.03189
G1 X159.806 Y171.368 E-.05523
G1 X159.87 Y171.331 E-.05523
;WIPE_END
G1 E-.85764 F2700
G1 X142.361 Y191.907 F6000
G1 E1 F3600
M204 P500
;TYPE:Outer wall
;WIDTH:0.419999
G1 F3000
G1 X141.527 Y191.88 E.02141

So again, my question: What is the “smallest” E-value that can be printed?

I am however not experienced enough to know if there is merit to this discussion and as such, reaching out to the more experienced/knowledgeable members for comments?

Thank you in advance.

[xchrisd]

Only to be said, from my point of view, this is a slicer issue.

I mostly use simplify 3d so maybe no really help here.
Look for parameters or rules which could limit single extrusion, like minimal single extrusion length.
This could lead to skip this “empty extrusion moves” and improve the print quality.

I use retractions as less as possible.

The minimum extrusion length can’t be quantified, it depends on the layer height if it’s usable. I have no real experience on this, 0.0008mm extrusion length is extremely tiny…
This need some calculation, what’s the steps per mm of the extruder?

[Hauke]

OK, your feeder definitely does not look worn, so that’s not the problem.

@Waterboy

user made a comment along the line of “if you need to do this you have something else wrong that you need to fault find, this is a setting that should not really be used” which is why I have refrained from testing this. If I am able to get use of my SM back, can you advise how you went about determining the “amount” to add to the extrusion?

That’s a bit harsh, saying that extra prime amount is a sign for something wrong. I consider it a necessity - because here is what happens:

• Retraction pulls back the filament
• A bit of the molten plastic will stick to the print and miss in the nozzle
• After retraction, you re-prime your nozzle, but plastic needs a bit of time to build up pressure again, still the printer assumes full flow
• In some cases, you even have oozing, again making your nozzle lose plastic

Extra prime amount takes this into account. I determine my amount by trial and error, and ended up at 0.05-0.1 mm³.

EDIT: not sure if you mentioned it - but did you try with retraction turned off completely? Is the effect still there then?

[Tigercjn]

Just a thought as it looks like melting. Is the fan working in the extruder. Do you have it at 100% in the slicer. Octoprint should show what it is set to. Can you feel the fan working? Is the vents clean?

[Waterboy]

I agree, to me this seemed to get resolved when I changed the following setting (for Arachne walls) “Perimeter transition threshold angle” which has a default of 10deg, and upon changing this to something higher like 30deg, the slicer no longer looked to do these “non-prints” in this area.

With regards to your “steps per mm of the extruder” question, is this the E-steps (M92 value)? I’ve been curious what the represents, but I guess that makes sense… My value is 219.91 (as opposed to the default of 212). I know this is “low” compared to what other guys are getting but have tested it more than once (with the extruder removed) and have gotten this value or very close to it every time.

[Waterboy]

This makes entirely sense to be honest, and it would explain the “gap” in the skirt… Would I be correct in saying that this would be more prominent if you were to have z-hop enabled?

Do you mind sharing your trail and error process? did you do this with the stringing models? My (maybe unreasonable) fear would be that in some places one will start introducing over-extrution…

On your last question, yes with stringing turned off there were no longer any gaps or under extrusion in either the print or the skirt.

[Waterboy]

Yeah, I tested it by utilising the terminal from Octoprint, issuing M106 commands to see if I could hear and feel a difference in the velocity of air extruded, and I could.

Looking at issues other users might have (including the Bambulab X1C users who are lifting the top glass lid of their printer to prevent the chamber from getting too hot), I suspect it was just as simple as the enclosure getting too hot for PLA with the bed at 60deg, as well as all the heat the linear rails and nozzles are giving off. But opening the enclosure for PLA prints, at least for me has not resulted in any warping and the issue has not occured again.