Ender 3 Distorted Calibration Cube

4

Issue: My Ender 3 is creating distorted prints with layer separation and deformations.

Example Here

Can anyone point me in the right direction?

Setup:

  • Ender 3
  • TH3D EZABL Auto Bed Leveling
  • 1.75 mm ABS filament (245 °C)
  • Heated Bed (100 °C)
  • Sliced in Ultimaker Cura (with 1.75 mm filament diameter and 0.2 mm layer height)

What I have tried:

  1. Tightening Z axis screw
  2. Tightening Y axis belt
  3. Tightening X axis belt
  4. Switching to a different spool of ABS
  5. Printing a temperature tower (same problem across different temperatures)
  6. Turning off the auto bed leveling.

Update:

  1. I've measured the temperature of the hot end, it is reading around 205°C +- 20°C
  2. As per suggestion from the TH3D support team, I tuned the PID of my hot end. Unfortunately the results did not turn out much better (1,2) and the support technician is suggesting(a long with many people from the comments) that I should try to replace the thermistor. As per suggestion from @Trish, I measured the impedence of the thermistor and it does seem to be somewhat off from stock (118kΩ vs 100kΩ). Will update again once the replacement arrives.
  3. As requested, here are some more photos front, back, left side, right side. The cube isn't hollowed out just to save material while I calibrate the dimensions.
  4. I printed the same cube out, rotated 90 degrees. I got similar results though: front, back, left side, right side.
  5. Swapping the nozzle out solved the issue! Thanks for the help everyone!

Dr. Mantis Tobbogan

Posted 2019-02-26T18:43:25.330

Reputation: 405

1Welcome to 3dPrinting.SE! – Pᴀᴜʟsᴛᴇʀ2 – 2019-02-26T18:49:01.580

Thanks @Pᴀᴜʟsᴛᴇʀ2! – Dr. Mantis Tobbogan – 2019-02-26T18:56:16.277

1What is the diameter of the filament in Ultimaker Cura, default this is 2.85 mm. This looks like an underextruded printed part. – 0scar – 2019-02-26T19:05:20.560

1

245° ABS? The average ABS is listed somewhat lower, about 230 °C.

– Trish – 2019-02-26T19:21:48.810

@0scar, sorry, I failed to mention that the filament is set to 1.75mm in Cura. – Dr. Mantis Tobbogan – 2019-02-26T19:35:17.247

@Trish, that is what I was thinking as well (even though I've gotten good prints at that temp in the past). I can try printing the same cube at 230 °C though. – Dr. Mantis Tobbogan – 2019-02-26T19:37:36.803

1Have you done successful prints with another filament type (e.g. PLA)? – Mick – 2019-02-26T19:53:14.553

@Mick, I've successfully printed PLA in the past. If I can get my hands on another spool I will give that a try as well. – Dr. Mantis Tobbogan – 2019-02-26T20:00:46.537

1carefully inspecting the photo, it seems that yoru filament is either too hot or you have underextrusion for other reasons. – Trish – 2019-02-26T22:55:17.537

1Can you check with a IR thermometer or a thermocouple what is the real temperature at the nozzle? I believe you got the configuration of your thermistor wrong in firmware. – Roberto Lo Giacco – 2019-02-26T23:57:48.793

@Trish I think you may be onto something. I tried printing at 230°C and notice some odd behaviour while it was printing. – Dr. Mantis Tobbogan – 2019-02-27T20:05:00.517

@RobertoLoGiacco, I don't have a IR thermometer or thermocouple but can pick either one up in the next day or two. – Dr. Mantis Tobbogan – 2019-02-27T20:06:22.117

@RobertoLoGiacco, I was finally able to check the nozzle temperature with an IR thermometer it was reading about 205°C +- 20°C. Do you think it makes sense that it is a settings issue and not a hardware issue with the thermistor? – Dr. Mantis Tobbogan – 2019-03-05T00:30:30.310

1@Dr.MantisTobbogan I believe you got your thermistor configuration wrong when building the firmware as you had to rebuild the firmware to get the ABL feature. I would hear other opinions as well: a 20° variation is not normal and I assume the heat block surface temperature should reach a value close to the internal temperature after a short period of time, but that might be a false statement. – Roberto Lo Giacco – 2019-03-05T00:36:18.220

@RobertoLoGiacco, fair enough. I am going to reach out to TH3D (ABL maker) and see what they suggest; their firmware is based off Marlin but I couldn't pinpoint exactly which configuration file I needed to revise. – Dr. Mantis Tobbogan – 2019-03-05T01:14:17.773

1Can we get some more photos of that monstrosity? Also, try rotating the part 90 degrees in the slicer and try the same print settings. Did you print that cube to be hollow for any particular reason? – user77232 – 2019-03-12T20:00:50.700

@user77232, I updated my question with some new images. I will work on printing another cube (hopefully tomorrow) that is rotated 90 degrees. I hollowed out the cube mostly just to save material while trying to calibrate. – Dr. Mantis Tobbogan – 2019-03-12T23:59:17.307

@user77232, I posted some images of the cube rotated 90 degrees. Take a look and let me know what you think. The results are pretty similar. – Dr. Mantis Tobbogan – 2019-03-13T23:09:59.793

1the 90° turn shows clearly, that the behavior is not model dependent but seems to be printer inherent (esp. the back side being a grate in both cases). After considering the thing, what is the resistance of the heater cartridge and the thermosensor at room temerature? Do you have some kind of heater cartridge sock or airstream-breaker to prevent air onto the heatsink reaching the hotend? – Trish – 2019-03-13T23:46:55.450

Hi @Trish, I calculated the resistance of both the heater cartridge and the thermistor. The heater cartridge value read okay but the thermistor definetly seemed a bit off (reading was 118.5kΩ vs the 100kΩ listed in specifications). Unfortunately I don't have any sock or thermally resistant material to block air flow. I am however waiting on a new thermistor and will report back. – Dr. Mantis Tobbogan – 2019-03-17T20:54:42.640

1@Dr.MantisTobbogan a single piece of aluminium foil can help to guide the airflow, but don't have it touch the heater block. Calculating is bad: decouple and measure directly if you can – Trish – 2019-03-18T12:31:57.720

1@Dr.MantisTobbogan, That thermistor reading might be correct. Do you have the means to measure the temp with an external device? – user77232 – 2019-03-18T15:05:57.003

Hi @Trish, that was bad wording on my part. I should have used the word measure (with a multi meter) instead of calculate. I do have some aluminum and will give that a try. – Dr. Mantis Tobbogan – 2019-03-19T13:05:50.597

@user77232, I measured the temp with an IR thermometer and the external temp was reading 205°C +- 20°C. Is there somewhere in mind that you think I should check? – Dr. Mantis Tobbogan – 2019-03-19T13:07:12.420

1@Dr.MantisTobbogan, and that temp was correct on the display was well right? What's the temp of the room in which the printer is located? – user77232 – 2019-03-19T13:36:57.773

@user77232 the temp was reading 245°C on the display. The room temperature is about 21°C. – Dr. Mantis Tobbogan – 2019-03-20T16:40:42.500

1@Dr.MantisTobbogan, I don't think it's the hot end. When you try pushing the filament through the extruder, does it come out straight, or does it curl to one side? If it curls (or sputters) then there is clog in the nozzle. You may have to do a cold pull or use cleaning filament. – user77232 – 2019-03-20T21:00:52.190

@user77232, it definitely does curl to one side. I can try swapping out to a new nozzle (I have a couple spares). Will keep you posted. – Dr. Mantis Tobbogan – 2019-03-21T13:23:45.583

@user77232, you nailed it right on the head. Swapped the nozzle out last night and all is right in the world. Would you have any idea as to why their would be buildup around between the thread of the nozzle? I can post in a new question if needed. – Dr. Mantis Tobbogan – 2019-03-22T16:23:37.983

1Swapping the nozzles might have solved the issue in multiple ways: a deformed nozzle could lead to very fast clogging and non-centric extrusion (might be the reason why 1 wall is "ok" and the others not. I strongly suggest to self-answer this question and tell what and who finally worked and what steps helped to increase the quality. – Trish – 2019-03-22T17:05:08.240

Will do @Trish. – Dr. Mantis Tobbogan – 2019-03-22T18:06:07.460

Answers

2

Resolution: After many trials and errors, I finally replaced the nozzle with one that that was not partially blocked by filament at its entrance.


Likely cause: A careful examination of the old part hints, that the repeated blockage in the nozzle seems to have been caused by a gap between the PTFE tube and the nozzle, which has considerably moved backwards under the stress of printing as one can see here in a photo of the PTFE tube.


It took me about a month and I went down a few rabbit holes until @user77232 made a great suggestion to check the nozzle and see if it needs to be cleaned. Thanks for helping me out everyone!

Dr. Mantis Tobbogan

Posted 2019-02-26T18:43:25.330

Reputation: 405

1Would you have any idea as to why their would be buildup around between the thread of the nozzle? - Dust can get onto the filament (like cement dust) or burnt filament which can clog a nozzle. Metal flakes from the tip of brass nozzle as it wears can also get sucked in when the retraction occurs. – user77232 – 2019-03-22T21:28:49.270

1@user77232 I can think if more causes: the tube might not be fastened correctly and slip ever so slightly, or it was cut slanted. Or, as you also said, the nozzle was bent/worn, creating the blockage – Trish – 2019-03-22T22:09:32.037

@user77232, here is a photo of the nozzle. It's a bit hard to see but there is a ring of plastic at the end of the nozzle. I am guessing that I didn't properly insert the tube back or it shifted out of place somehow (I also swapped the fitting just in case).

– Dr. Mantis Tobbogan – 2019-03-22T23:20:29.537

1the photo of the frizzled PTFE shows, that part of the reason might be the clamp holding the PTFE in the hotend - the tube might have slipped, bit by bit. – Trish – 2019-03-23T21:00:44.903

That is exactly what I was thinking @Trish. I am glad that I had a few spare clamps. – Dr. Mantis Tobbogan – 2019-03-25T19:46:02.180

1

This has been a common-enough issue that there's a well-designed fix for it: https://www.thingiverse.com/thing:3203831 Essentially it involves using two sections of PTFE tube, one inside the hot-end throat and tightened down using a washer and the bowden tube fitting itself. Then the second tube is used as a bowden tube to feed filament into the hot-end. The bowden fitting's ring is no longer responsible for keeping the tube in the throat flush.

– GDorn – 2019-06-30T23:57:22.490