Crazy issue, trying to rule out printer.



  • So I'm trying to rule the printer out of the equation for a problem I've been experiencing. In the image below you can see diagonal extrusions going bottom right to top left instead of going up and the left and finishing the wall. This slice was from S3D 3.1.1 and I kept trying to fix the issue by messing with the bridging settings. It didn't dawn on me that the printer was going diagonal instead of up and over. It's exactly like it just skipped that one spot. The weird bridges after are another mystery, almost like everything got skewed. Keep in mind I produced the same results multiple times with slightly different speeds, extrusion multipliers on bridging, etc.etc. So it's not as if I tried the same g-code over and over and expected different results. I didn't notice the issue once I switched to 2 walls. I can try to reproduce the issue by slicing the same object with 3 walls again and doing a test print. Outside of that I've lost any of the g-code.

    So, is it even possible at all this was a printer related issue ? I've already posted to the S3D forum.

    Thanks,

    Jeff


  • administrators

    I guess you are trying to make something that needs to be hollow, otherwise you would use infill and you wouldn't need to bridge.

    Evidently the ends of the bridge are not adhering to the walls, and the filament is shrinking and pulling the strands away from them. I suspect excessive print cooling. What material are you printing?



  • Well it's a fan duct , so yes air will pass through the channel. The weird part is how in the world would it 5 times go diagonal instead of up and the left. You can see the 3rd wall is missing an instead is a diagonal.

    This is an ABS print at 235 C, part cooling was off until it went into bridge mode which wasn't until After the weird diagonal move. The S3D preview on the moves showed it should have gone up and then left to complete the 3rd inner perimeter followed by a back and forth bridge fill. I can see once or twice that it failed to stick and then pulled to a diagonal but not 5 times? I also re-sliced each time changing some bridging parameters. I guess i need to reproduce this and see what the g-code shows. I'm trying to rule out the printer and just blame the slicer.

    Edit: I also tried this at 245C with the same results. I also watched the printer move diagonal instead of up and then left. It's as if it skipped the move.

    Thanks,

    Jeff


  • administrators

    Leave the part cooling off even in bridge mode, or else reduce the fan speed a lot.



  • Ok, but why did I see the printer move diagonal when it should have moved up and then left to finish the perimeter wall ? You can see in two places where there is a diagonal extrusion, in the case of the bottom section it's where all the bridges are hanging off of. That line doesn't exist at all in the S3D preview of moves, nothing even resembling that. I'll try to reproduce the issue which I did 5 times and post the g-code. I should have pointed to the problem areas on the photo, always short on time.

    Thanks,

    Jeff


  • administrators

    Perhaps it did move up and left, but the fan came on because bridging was due to start shortly, and the resulting over-cooling caused the filament not to stick to the previous layer but to shrink and take the shortest route instead.



  • @dc42:

    Perhaps it did move up and left, but the fan came on because bridging was due to start shortly, and the resulting over-cooling caused the filament not to stick to the previous layer but to shrink and take the shortest route instead.

    I actually witnessed the diagonal move, also why would this happen 5 prints in a row ? At first I thought it was just an odd retract, but not after a few times. Once I noticed it actually move diagonal and lay down plastic while doing so I gave up and ultimately switched to 2 walls instead of 3. I had my ABS set to 3 walls as I often find myself running taps for screws and 3 just works a bit nicer. Setting it to 2 walls actually changed the behavior and it no longer did that diagonal move.

    Again, not blaming the printer I'm just trying to rule it out. The S3D preview didn't show a diagonal move at all. This is not a case of plastic being dragged left and coming unstuck it literally moved diagonally, not just once, but 5 times after different slices.

    I guess I'll blame the slicer for now until I can get a test g-code file and send it over. The proof is in the print, also notice it did the same thing on the top and bottom sections.

    Fan settings were changed between prints, the first time this happen the fan's were only set to 15%, the last one was 80% on bridging. No change in that move.

    Thanks,

    Jeff



  • I see what you're talking about. It's because the printer is not doing a retraction for that move, and a huge line is going where the travel move is. Fix your retraction settings in S3D.



  • @bot:

    I see what you're talking about. It's because the printer is not doing a retraction for that move, and a huge line is going where the travel move is. Fix your retraction settings in S3D.

    That's what I first thought as well, however I witnessed the printer move diagonally while lying down the plastic. If you look at the rest of the print you'll not see any signs of retraction issues. I'm going to try and reproduce the g-code and problem and view it in another g-code previewer instead of S3D. I doubt the issue is the printer however I witnessed the wrong move (multiple times). It's as if it skipped one move and jumped to the next which resulted in a diagonal move instead of an L (flipped).

    I'm still assuming the printer would be incapable of making such a mistake so I need to get the g-code produced again so I can confirm it's the slicer.

    Thanks,

    Jeff



  • Ok, I managed to get this to happen again. Virtually identical results.


    If you open this image in a new tab you'll get full rez.

    G-code used to create this mess: dual_fan_duct.gcode

    and a couple of S3D shots to show what should have happen.

    Notice nothing has diagonal lines on the inside perimeter.

    Feedback welcome, just trying to get to the bottom of this problem. It's too repeatable to be a detach issue, if it is I'll be really surprised.

    Thanks,

    Jeff



  • Ok, testing something. I suspect it's air printing that line. I just set S3D to do Outside-In for outline direction instead of Inside Out. If my suspicious is correct that should solve this issue…maybe. Printing 2 Outline/Perimeter shells does solve it which would also make sense. My apologies if it turns out to be an air print, hopefully it can save someone else some headaches.

    Edit: Changing to outside in didn't change the behavior at all surprisingly.

    Jeff


  • administrators

    I still don't know where the diagonal extrusions are that you are talking about.



  • @dc42:

    I still don't know where the diagonal extrusions are that you are talking about.

    I think we are supposed to be looking at the pale blue blue lines in the S3D pics which show an unsupported inner perimeter - hence the problem getting it to print IMO.
    Edit. I can't see a diagonal extrusion in the first print picture but I can in the second. Difficult to spot with everything that is going wrong.



  • The diagonal extrusion is sitting under the pile of what should have been bridges. I think the bottom inner extrusion is unsupported and of course not bonding to the side, I'm printing at 80mm/s with 3000 acl and 150mm/s travel moves. So watching it the bottom inner extrusion gets pulled away by a fast travel move which gives the illusion of the printer moving diagonal. Just a case of the mind playing tricks on me seeing something move but not seeing the true source.

    If I print with 2 inner perimeters instead of 3 then I don't get the hanging perimeter and all is well.

    I appreciate everyone looking at this mess, I'll resort to highlights and arrow for future posts of this nature, which I hope I never have.

    Thanks,

    Jeff



  • Sorry I wasn't more clear before. I know exactly what you're talking about, and it IS a retraction issue.

    Under the "Advanced" tab of S3D slicing parameters, you'll find "ooze control" and an option "only retract when crossing open spaces." Make sure that option is not checked.



  • @bot:

    Sorry I wasn't more clear before. I know exactly what you're talking about, and it IS a retraction issue.

    Under the "Advanced" tab of S3D slicing parameters, you'll find "ooze control" and an option "only retract when crossing open spaces." Make sure that option is not checked.

    If you look at the second print, open the image in a new tab and zoom in you'll notice if it's an Ooze it's the largest in recorded history 🙂

    I also tried a print with and without that setting as I to was thinking the same thing. Closer inspection reveals it's a normal extrusion that just didn't stay glued. I could probably get it to hold against the side while floating in air as it is by playing with the temperature and cooling a bit more. The crazy part is I can print this part over and over with the same g-code and get the same results each time.

    Thanks,

    Jeff



  • Do you have the option enabled or disabled? It should be disabled, if my hunch is correct. Did you try printing the object with that disabled?

    Also, do you have a bowden tube? It is not abnormal, at all, for enough pressure to build up to expel that much filament during a move, especially considering you say your print feedrate is so high.

    Edit: Oh I just saw your last explanation. That seems to be it.


Locked
 

Looks like your connection to Duet3D was lost, please wait while we try to reconnect.