UMs5 Wall Separation Issue

[matt_dt3d 1]

I've been having an issue with my s5 since the beginning of the year. The issue is that the outer walls are printing separated, which leads to weak, inaccurate, unsightly parts. I've tried talking it through with Ultimaker / FBRC8 support, and nothing we've come up with has helped after months of chasing the issue.

 

The problem started after I replaced all of the sliding blocks. One of the blocks had completely lost its grip on the print head axle, so it was due. At the same time, I thought it would be a good idea to clean my s5 thoroughly, as it was entering its fourth year of service.

 

Here is a list of things I've tried to get the printer to run correctly:

 

0. Deep cleaning the entire motion system before replacing the sliding blocks. This included disassembling the print head to clean its components, cleaning the axles and pulleys thoroughly, reassembling and oiling the motion system, all according to the directions in the knowledge base. 
1. Realigning all axles, checking for axle play, tightening belts, grub screws, and motor mounts per the knowledge base.
2. Different filament and printing hotter. I have multiple known-good filaments that print just fine on my other Ultimaker 3 machines. Doesn't matter if it's PLA, ABS, or PETG. Same result.
3. Cleaning the filament feeders (Ultimaker Support suggested).
4. Cleaning the Print Core by hot pulling and then cold pulling until fully clean.
5. Swapping to a known-good Print Core from another Ultimaker.
6. Realigning all axles, checking for axle play, tightening belts, grub screws, and motor mounts per the knowledge base. Again.
7. Trimming the Bowden tubes shorter to eliminate filament play. This resulted in crisper prints, but the issue didn't go away.
8. Replacing the short belt loops (Ultimaker Support suggested).
9. Shimming the X-axis printhead bearing to eliminate play in the bearing (Ultimaker Support suggested).
10. Going through and carefully torqueing all of the printhead screws to the very light torques specified in the knowledgebase.
11. Disassembling, cleaning, and reassembling the printhead according to documentation provided by Ultimaker support.
12. Made sure that both motors are mounted perfectly level and not tilted (Ultimaker Support suggested).
13. Downgrading Cura to 4.13.1 to see if the Arachne Engine was glitching.
14. Slicing the test file in PrusaSlicer, then editing in the gcode header from Cura, and printing the resulting gcode.
15. Cleaning and inspecting the filament feeders for damage. Again (Ultimaker Support suggested).
16. Re-calibrating the XY Offset.
17. Disassembling the motion system. Realigning all axles, checking for axle play, tightening belts, grub screws, and motor mounts per the knowledge base. Once again.

18. Taking the bearings out of the printer frame, cleaning them thoroughly, and repacking them with grease. Two of them had debris that was making them crunch.

19. Disassembling and comparing the filament feeder to one that hadn't been disassembled. Turns out I reassembled the filament flow sensor wrong.
20. Installing a new X axis bearing into the printhead (Ultimaker Support suggested). Now the wall separation is very well-defined. (Teal test print)

 

At this point, if you told me that sacrificing a lamb under the blood moon would fix it, I would try it.

 

I'm running firmware version 7.1.3. I don't plan on upgrading any further, because:

 

A. I can't downgrade later if I want to.
B. The new features and bug fixes don't apply or matter to me since I don't use Digital Factory, have no addons, and don't have an s7.
C. Local Management Console features were removed in 8.1.0 that I use regularly.

 

I'm at the stage where I'd be relieved to sell the s5 for parts if I could get most of my money back out of it.

 

Attached pictures:

1. Teal PETG test print from today, 06-05-23.

2. Candy Red PLA test print from 05-01-23.

3. White PETG test print from 04-13-23.

4. Orange ABS print fail from 04-13-23.

5 and 6. Examples of what a PETG print failure looks like when I grab the part and twist. It should be rock solid, instead, this happens...

 

Any and all guidance is appreciated.

Matt

 

Edit: The test print is the Micro All in One 3D Printer Test by Majda107. I used the bone-stock Ultimaker profiles for PLA/PETG/ABS as applicable. I changed two settings: 3 walls, with Z seam in the rear. No other changes. I've shared my project file with Ultimaker Support, and they either didn't print it, or it printed fine for them. I've also added the project file to this post.

 

UMS5_AIl_In_One_3D_printer_test_micro.3mf

Edited June 6, 2023 by matt_dt3d
Add details

[Dustin 130]

Would say it looks like your having flow issues as its all tested material types.

I would tell you to clean your print core but it sounds like that has been done already?
But just to confirm, per this guide: https://support.makerbot.com/s/article/1667411136256
The cold pull portion of the guide is the most important.
Please feel free to share pictures of your pull results for feedback if your unsure. 

Some other things too look into. 

• Try disconnecting your bowden tube on both ends and just manually push filament thru it and make sure its not being restricted. 
• Are you seeing any overly chewed up marks on your filament in the bowden tube during these prints? Please share pictures.
• Your front fan might be dirty or about to die. Just because its spinning does not mean its reaching its full speed. If this is the case it will result in heat creep which will cause flow issues. 
• Check both sides of your print head cable to make sure the connection is secure. 
• Have you sent your machine log file to UltiMaker after completing one of these prints in question? 
• You said you contacted UltiMaker and Fbrc8, can I get your most recent case number? I will reach out to the Support Manager to find out what might be going on.

[matt_dt3d 1]

4 hours ago, Dustin said:

I would tell you to clean your print core but it sounds like that has been done already?
But just to confirm, per this guide: https://support.makerbot.com/s/article/1667411136256
The cold pull portion of the guide is the most important.
Please feel free to share pictures of your pull results for feedback if your unsure. 

I followed that exact guide. On more than one occasion, I've hot pulled until there was no residue left from the previous filament on the end of the (natural, non-colored) PLA, about three or four times. Then I cold pulled until not only was the tip coming out clean and perfectly formed, but I could sight down the printcore and see a 0.4mm pinhole of light straight through it. Clean as a whistle. I've also used the Ultimaker official cleaning filament. I'll do another round to show you the results.

 

Quote

• Try disconnecting your bowden tube on both ends and just manually push filament thru it and make sure its not being restricted. 

I've tried that, and there was some friction. I've cleaned the tube out and added a couple of drops of oil to free it up. It's a trick I learned from gr8 for attempting to print TPU on an Ultimaker. If that fixes it, I'll upgrade to Capricorn tubes.

 

Quote

• Are you seeing any overly chewed up marks on your filament in the bowden tube during these prints? Please share pictures.

No I am not.

 

Quote

• Your front fan might be dirty or about to die. Just because its spinning does not mean its reaching its full speed. If this is the case it will result in heat creep which will cause flow issues.

I have a spare fan, I'll add that to my list of variables to check.

 

Quote

• Check both sides of your print head cable to make sure the connection is secure.

The cable end inside of my printhead comes loose with a pull, like the retention clip isn't engaging. I'll try designing and printing something to hold it in place if/when the other checks don't work.

 

Quote

You said you contacted UltiMaker and Fbrc8, can I get your most recent case number? I will reach out to the Support Manager to find out what might be going on.

My request number is 65865.

 

 

[matt_dt3d 1]

Here are the results of my print core cleaning. I used Ultimaker official cleaning filament and did 3 hot pulls, then 4 cold pulls. I can sight down my print core and see a 0.4mm dot of light at the bottom. It doesn't show up well in a photo, but I tried my best.

[matt_dt3d 1]

The cleaning did not help.

 

I installed the new fan. Here are the (same) results.

 

The retention clip on the cable inside the printhead is sheared off almost perfectly, so that the connector is fully intact, but isn't being retained. I've rolled up a piece of paper that is the right size to hold the connector in place and reassembled the printhead. I re-ran the test piece, and I'm still having wall separation issues. 

 

It's got to be either something wrong with the mechanics of the motion system, or a problem with the feeder.

I'm going to run a vase-mode backlash test per a user's recommendation from a parallel thread I started elsewhere. I'll report back tomorrow.

[Dustin 130]

Your cold pulls are shaped well.
But unless it's the angle of the camera, they appear still very dirty. 
Your printed part likewise appears to be full of dots of carbon.

If your cold pulls were still dirty and not translucent (assuming continued use of the cleaning filament) then your core is still not clean. 

 

Cold pull does 2 things.

• Cleans out filament stuck in the tip of the print core and down in its inner edges.
• Cleans out the walls of the print core

If the cold pulls are still coming out with dark marks on the sides (again assuming cleaning filament, or white pla was used)

Then this means there is still material or carbon build up. Which means keep doing cold pulls until the core is clean or replace the print core. 

The dark spots in the print suggests that carbon is coming off the walls of the print core and are being forced thru the nozzle opening.. When this happens it restricts filament flow which causes under extrusion.

 

Last print core I cleaned, I wasn't done until about 14-15 cold pulls because I kept pulling out carbon.
(I was printing a lot of ABS prior..)

However, depending how old this print core is and what kinds of filaments you have been using. Its possible the nozzle is worn out as well which will yield unexpected extrusion widths 

As for your loose connection at the print head. 
I recommend replacing the print head cable. That only has bad things waiting for it.
Without a log file I can't say off exactly, but it could be that your having massive temperature swings due to the cable issue.
These would eventually trigger thermal runaway conditions which would cause the firmware to just stop your prints completely.

 

I recommend getting this machine serviced, as it appears its been thru quite a bit.

[matt_dt3d 1]

The dark spots in the print are part of the filament. It is "granite" PLA, which is white PLA with added dark speckles. I thought I had mentioned this here, but I must have said that in the parallel thread.

 

In the parallel thread one of the users was demanding that I use PLA for testing, as it is "more forgiving." The only other PLA I have in 2.85mm is a dark blue with glitter, which would not be any better for troubleshooting.

 

I've switched to bright white PETG for further testing. 85%+ of the printing I do is with PETG anyway, and I've shown this underextrusion problem with PLA, like five different colors of PETG, and ABS, so it's not the filament.

 

To be clear, I'm not seeing any errors pop up on the printer's screen like Err34 for the print head cable.

 

It is not the print core being worn out. I have 6 different AA print cores, one of them barely used, and all of them exhibit the same problem on this machine.

 

Here are the results of 11 more cold pulls. 
 

The last (15th) cold pull:

 

 

 

I see in another thread where a user is having a very similar problem on their s5R1, that they bought new print cores, and it didn't solve their problem. So far, just on this machine, I've bought new sliding blocks, new short belts, a new X axis bearing, and a new fan, and none of it has gotten the machine running properly. FBRC8 is making a killing selling me 2-3x overpriced replacement parts, and I still don't have a working printer, it's very frustrating. And now you want me to pay for your technicians to work on it, but my faith is running out. I'm not willing to spend $500+ to have the machine serviced in an open-ended kind of way. The other user is in the same boat as me, and he's given up hope. I can already smell the money burning and see the emails that say (in effect) "we thought it was X Y or Z, but after replacing those parts, it still doesn't work. Can we bill you for more time while we replace parts until something works?"

 

I've already had a bad experience with the service center. They took my machine for networking issues, literally did nothing to it but run a ping test, and waited two weeks. Then told me they can't find anything wrong with it, and they won't/can't test it with Windows because they only use MacOS, and sent it back to me. Insultingly bad service. So in this case, If I could get a written assurance that I won't be billed for anything that doesn't fix my printer, then and only then would I be willing to send it in. If the technicians want to replace parts until something works, I will not pay for the unnecessary parts and labor. I need a concrete solution, and we're just guessing at this point.

 

Anyway. Do you want me to post a log? The guys working my support ticket have never asked for a log.

I'll keep an eye on the temperature graph during my next print and see if there are temperature fluctuations.

 

PS the backlash tower I attempted to print overnight failed with massive underextrusion.

[Dustin 130]

@matt_dt3d I understand this has been frustrating for you, and frankly I was also having issues not long ago that was much like yours.

I have been in touch with the Support Manager.
Please watch for a response to your noted Support Case, and please be friendly to the person on the other end.
They are some of our best team members and do truly only wish to help.

[matt_dt3d 1]

Update and solution:
I have received my printer back from the Ultimaker Service Center. They replaced:

1. Sliding blocks (maintenance)
2. Timing belts (maintenance)
3. Printhead Cable (broken retention clip)
4. Printhead PCB (the threaded insert was about to break off)
5. Printhead capacitive sensor (the old one had a loose wire)
6. Front Axial fan - (updated to the new version)
7. Nozzle seal - (maintenance)
8. Both Printhead side fans (the slack wires were hanging out of the back of the print head)
9. Printhead side fan bracket (had a chip on the bottom)
10. Both Bowden tubes (maintenance)
11. Both Feeders (maintenance / update to the new fine-knurled version)
12. Installed a new AA 0.4mm PrintCore (maintenance)
13. Front Axial fan (again - the other one died during testing)

And the test prints they ran were turning out well.

 

However, upon receipt of the printer, I was still having the wall separation issue, no matter what version of Cura I used (I tried 4.13.1, 5.2.1, and 5.4.0).

 

The fact that the issue was still occurring and was happening on my other Ultimaker machines led me to investigate the PC I was using to slice the files. All of the hardware is recent except the graphics card (an ASUS GTX 660, model GTX660-DC2O-2GD5), so I purchased and installed a new graphics card (an MSI GTX 1650).

 

Now everything I slice is printing fine.

 

So, a lesson hard won: the graphics card you use in your PC for slicing matters!