Notifications
Clear all

[Solved] First layer issue at some specific bed area  

Page 1 / 2
  RSS
Desti
(@desti)
Eminent Member
First layer issue at some specific bed area

Hi there,

I used to be able to use my MK3 but I recently had to replace the whole heating block plus cables, so now it's replaced and working.

Now I just have an issue with what seems to be extruding in a certain area, as you can see here:

https://drive.google.com/open?id=1VD-bqClxGciNxy2wA0mXqtv8V7_sr11i

There is no issue on the right part of the bed but it seems to get more complicated on the left. I thought first it was about changing the bed leveling for the sides (like higher or lower for the left side) but that never changed anything. The strange thing is that when you check at the video, it's always when reaching the left end that the gear seems to struggle but the nozzle isn't catching the bed.

I tried many things, cleaning the steel sheet with detergent for the adhesion, making a cold pull in case it can be a bit clogged, tried to screw more or less the gears' screws, changed the filament (tried with Prusa silver PLA and Sunlu grey PLA), did a whole XYZ calibration again, but nothing worked.

Has anyone ever had something similar or any leads of what could be the issue? Thanks!

Best Answer by Vojtěch:

Looking at the pictures, particularly the picture with the four prints is very interesting. The lower bottom print shows not just the first layer, but the same problem appearing in the second layer, too. 

I have only one plausible explanation for a XY localized non-extrusion that affects subsequent layer: Broken E motor cable. As the head moves, the strands of copper make and lose contact, resulting in the motor stopping in a particular spot. Broken cables happen, particularly if they're pinched somewhere and moving a lot. A typical place is the thick cable bundle from the printing head at the place where it enters the Einsy enclosure.

@tim-m30 Desti mentions having tried washing the board with a detergent. You have a very good hammer, but not everything is a nail.

This topic was modified 5 years ago by Desti
Posted : 16/05/2019 5:39 pm
--
 --
(@)
Illustrious Member
RE: First layer issue at some specific bed area

Please -- post images inside the forum so they don't disappear in a day or two.  

But poor adhesion, is usually a contaminated bed and a soap and water wash will resolve the problem.

Hot Water wash: often, as needed
Handle the bed only by the edges.
Wash the bed in hot water, use a fresh paper towel as a wash cloth, with a few drops of plain dish soap (Dawn, unscented, no anti-bacterial, etc.). Rinse well in hot water - if you have very soft water, rinse a bit longer.
Dry the bed with a fresh paper towel.
Handle the bed only by the edges.
Place bed on printer.

Alcohol rinse: every few prints
Once in a while, an alcohol rinse is helpful to remove PLA residue. It does not remove finger oils.
Pour a 5 cm puddle of 91%+ alcohol in the middle of the bed, with clean hands use a fresh paper towel to scrub the bed. Wipe up all the alcohol.

Streak test: when contamination is suspected
With a fresh piece of paper towel, and very clean fingers, dampen the towel with 91%+ alcohol, and wipe the bed side to side moving back to front, like you're painting it with alcohol. The alcohol should be thin enough on the towel it quickly evaporates. If you see any streaks, the bed is dirty and needs a wash.

Acetone wash: infrequent
Pour a 2 cm puddle of acetone on the bed, scrub it around with a fresh paper towel. It will evaporate fast as you clean. This step removes PEI oxides that form over time and with heat, and improves PLA adhesion to a like new state.

  

Posted : 16/05/2019 5:44 pm
Desti
(@desti)
Eminent Member
Topic starter answered:
RE: First layer issue at some specific bed area

Thanks for the answer, but that's not an adhesion issue since it's the gears that just decide to struggle only at some area of the bed. I've already had adhesion issue and that's doesn't seem like it.

Also the bed is very new (got it like 2 weeks ago), I washed it with dish soap today during my tests, and also with 99% alcohol.

Sorry for the pictures, I do it here then:

The video: https://forum.prusa3d.com/wp-content/uploads/2019/05/VID_20190516_182544.mp4

 

Btw I just did now for the live adjust Z from this topic: https://forum.prusa3d.com/forum/original-prusa-i3-mk3s-mk3-assembly-and-first-prints-troubleshooting/life-adjust-z-my-way/

But still a similar issue. It starts from the bottom right corner, and until it reaches the line that fails first there is no gear problem at all.

This post was modified 5 years ago 4 times by Desti
Posted : 16/05/2019 6:06 pm
--
 --
(@)
Illustrious Member
RE: First layer issue at some specific bed area

It looks like a combination of effects; but not.  The first layer cal test is showing signs the bed is dirty (skin oil can be an invisible film that prevents PLA from sticking to the bed).  A soap and water wash may help (alcohol doesn't do a good job unless you use a lot of it to rinse). 

If it were simply first layer cal, the skirt wouldn't be printing.  

The only thing making sense is a localized dirt (weird, I know).  Try a soap and water wash using this recipe, and report back.  

Hot Water wash: often, as needed
Handle the bed only by the edges.
Wash the bed in hot water, use a fresh paper towel as a wash cloth, with a few drops of plain dish soap (Dawn, unscented, no anti-bacterial, etc.). Rinse well in hot water - if you have very soft water, rinse a bit longer.
Dry the bed with a fresh paper towel.
Handle the bed only by the edges.
Place bed on printer.

Alcohol rinse: every few prints
Gloves recommended.
Once in a while, an alcohol rinse is helpful to remove PLA residue. It does not remove finger oils well. 
Pour a 5 cm puddle of 91%+ alcohol in the middle of the bed, with clean hands use a fresh paper towel to scrub the bed. Wipe up all the alcohol.

Streak test: when contamination is suspected
Gloves recommended.
With a fresh piece of paper towel, and very clean fingers, dampen the towel with 91%+ alcohol, and wipe the bed side to side moving back to front, like you're painting it with alcohol. The alcohol should be thin enough on the towel it quickly evaporates. If you see any streaks, the bed is dirty and needs a wash.

Acetone wash: infrequent
Gloves recommended.
Pour a 2 cm puddle of acetone on the bed, scrub it around with a fresh paper towel. It will evaporate fast as you clean. This step removes PEI oxides that form over time and with heat, and improves PLA adhesion to a like new state.

 

Posted : 16/05/2019 6:27 pm
Vojtěch
(@vojtech)
Honorable Member
RE: First layer issue at some specific bed area

Looking at the pictures, particularly the picture with the four prints is very interesting. The lower bottom print shows not just the first layer, but the same problem appearing in the second layer, too. 

I have only one plausible explanation for a XY localized non-extrusion that affects subsequent layer: Broken E motor cable. As the head moves, the strands of copper make and lose contact, resulting in the motor stopping in a particular spot. Broken cables happen, particularly if they're pinched somewhere and moving a lot. A typical place is the thick cable bundle from the printing head at the place where it enters the Einsy enclosure.

@tim-m30 Desti mentions having tried washing the board with a detergent. You have a very good hammer, but not everything is a nail.

Posted : 16/05/2019 6:33 pm
--
 --
(@)
Illustrious Member
RE: First layer issue at some specific bed area

Okay - sorry for the duplicate post ... if you are hearing grinding, there are a few possible reasons.  Though NONE of them fit the "it only sputters in this one spot" you are seeing. Oil on the bed does.

The MK3 Bondtech idler gear shaft is too short for the application and has been known to slip out of one side of the door and allow the idler gear to move around, and it will cause clicking. Make sure the shaft is fitted into both "ears" in the door, equally. Flush on one side isn't enough, it must be spaced equally ... [  -----------  ] not   [---------    ]  ...

Spring tension being too loose can cause skipping,  and might cause skipping that is sensitive to travel.  Ensure the two tension screws are set evenly, and just a two turns past spring contact (not too loose, not too tight). End of screws should not extend beyond the surface of the idler door.  Flush is about ideal.

Poor filament routing.  If you have an MMU, old PFTE Bowden tubing gets work and will cause friction as it flexes.

Filament dragging through things like filament guide mods hanging on top the printer frame...

Broken power cable to extruder - that "breaks" when the extruder is in a certain position...

 

Posted : 16/05/2019 6:43 pm
--
 --
(@)
Illustrious Member
RE: First layer issue at some specific bed area
Posted by: Vojtěch

@tim-m30 Desti mentions having tried washing the board with a detergent. You have a very good hammer, but not everything is a nail.

Yes - and posts get interspersed when two people post at nearly identical times.  

Honestly, I am surprised you didn't recommend a full tear down and spout some formulas from a reference book.

This post was modified 5 years ago 2 times by --
Posted : 16/05/2019 6:46 pm
Desti
(@desti)
Eminent Member
Topic starter answered:
RE: First layer issue at some specific bed area

Thank you for your answers, indeed I already washed before with dish doap and hot water, and just tried to do it on the other side of this new sheet which I have never used, and still the result is quite similar:

But it actually looks better than the first try of 1st layer calibration.

I really hope it's not a broken cable, I know it happens, but as I said in 1st post I just repaired my hotend because I broke a cable and having to wait again would suck :'( but if it's really the issue I won't have a choice anyway.

How can I check for the cable? Is it visible or can it be inside the cable? 

Posted : 16/05/2019 6:51 pm
Vojtěch
(@vojtech)
Honorable Member
RE: First layer issue at some specific bed area

Posted by: Tim

@tim-m30 Desti mentions having tried washing the board with a detergent. You have a very good hammer, but not everything is a nail.

Yes - and posts get interspersed when two people post at nearly identical times.  

They do. Not the first post, though. 😉

Honestly, I am surprised you didn't recommend a full tear down and spout some formulas from a reference book.

Perhaps I listened to you after all ...

Posted : 16/05/2019 6:57 pm
--
 --
(@)
Illustrious Member
RE: First layer issue at some specific bed area

Is there any chance you have dirt UNDER the metal sheet?  Something stuck to the fiberglass? 

Also - what firmware are you using?  Have you tried loading 3.7 and using the 7x7 meshbed level?

 

There's one other "out there" problem I've seen before: someone handled the filament spool after applying hand lotion.  Makes for a section of filament that comes and goes.  You can test by extruding a pile of filament and check for diameter changes - this image shows oily filament, some sections are the correct 0.4mm and others are much larger:

Posted : 16/05/2019 6:59 pm
--
 --
(@)
Illustrious Member
RE: First layer issue at some specific bed area

A closer look that the images found this detail: an occasional trace succeeds.  It can't be a debris under the bed or Z variance problem.   It pretty much limits the issue to the extruder stopping when in that region, but not every pass.  Plastic also isn't sticking well... but it's probably secondary.  Either filament is snagging or the extruder stopping.    Have you checked the Bondtech gear teeth for debris? 

Posted : 16/05/2019 7:09 pm
Vojtěch
(@vojtech)
Honorable Member
RE: First layer issue at some specific bed area
Posted by: Desti

I really hope it's not a broken cable, I know it happens, but as I said in 1st post I just repaired my hotend because I broke a cable and having to wait again would suck :'( but if it's really the issue I won't have a choice anyway.

Unfortunately, if your heater block cable was broken, it's quite likely other cables could also have sustained a lot of fatigue at the same spot.

Thinking about it some more, another reason could be that the filament spool is not turning easily when the head moves into a certain direction.

Make sure the filament's way into the extruder is unobstructed and there are no tangles.

How can I check for the cable?

I'd go to Settings -> Move Axis -> E and make the E move a long long distance, without filament in. You'll need to preheat for this to work.

The E motor should be spinning. To see it better, you can draw a mark on the axis with a sharpie pen.

Now try moving the extruder along the X axis by hand or moving the cable itself around. If the motor stops in some positions, it's the cable.

Is it visible or can it be inside the cable? 

It's usually not visible on the surface. There may be some marks if the cable was pinched and the cable is more flexible at the point of break.

If you have a continuity tester or a multimeter, it's possible to use a thin needle to pierce the individual motor cables near the motor and then measure between the needle and the connector.

If you don't have any such tools,  then removing the four cables from the connector, untwisting them and pulling on each one individually will reveal the point where the cable is broken. The insulation will stretch in the spot where the copper is not continuous.

It's fixable using a soldering iron. If you already have one, it's certainly a cheaper way than getting a whole new E motor. 

Posted : 16/05/2019 7:14 pm
Desti liked
Vojtěch
(@vojtech)
Honorable Member
RE: First layer issue at some specific bed area
Posted by: Tim

A closer look that the images found this detail: an occasional trace succeeds. 

Every second trace (mostly) succeeds. That means it depends on motion direction. I imagine the motor not moving in the area at all, but extruder pressure release and delayed buildup causing this pattern.

Posted : 16/05/2019 7:18 pm
--
 --
(@)
Illustrious Member
RE: First layer issue at some specific bed area
Posted by: Vojtěch

Every second trace (mostly) succeeds. That means it depends on motion direction. I imagine the motor not moving in the area at all, but extruder pressure release and delayed buildup causing this pattern.

No - it's missing a lot more than that. No directionality I can see.  It's spatial for sure.

Desti -- Just to rule out electrical -- any chance you can grab and wiggle the extruder cable harness while printing the part.   See if you can get skips anywhere else?

Posted : 16/05/2019 7:24 pm
Desti
(@desti)
Eminent Member
Topic starter answered:
RE: First layer issue at some specific bed area

Okay vojtech I've tried and you might be right for the E cable, indeed when I manually move the extruder along the X axis there's some point where it gets stuck when right next to it it's perfect.

I'm sorry I'm not english native and don't get everything you said for trying to manually repair otherwise by buying a complete new E motor.

So I unfortunately don't have any multimeter or anything else to measure.

If you don't have any such tools,  then removing the four cables from the connector, untwisting them and pulling on each one individually will reveal the point where the cable is broken. The insulation will stretch in the spot where the copper is not continuous.

It's fixable using a soldering iron. If you already have one, it's certainly a cheaper way than getting a whole new E motor. 

I don't get perfectly what you mean by removing the 4 cables from the connector. You mean removing the plug from the motherboard of the E motor cable, then from it removing the cables from the connector that goes into the motherboard, and then separate the cables from each other?

Then I also don't get how from that I can figure where it's broken. I actually think that I might have forced a bit too much on the cable when trying to replug it to the motherboard after the latest hotend repairs.

I have a soldering iron, so I'd like to try that, but after knowing where it's broken, the cables are quite thin and there's still multiple cables so I'm not sure how to properly proceed.

 

Sorry for all the questions, and thank you a lot, I definitely wouldn't have figured anything like that by myself.

Posted : 16/05/2019 7:37 pm
Vojtěch
(@vojtech)
Honorable Member
RE: First layer issue at some specific bed area

Ok, let me try. I don't speak French unfortunately.

A cable is made of copper strands wrapped in plastic insulation. Copper doesn't stretch, plastic does. When you pull on a broken cable, it'll stretch in the spot where it is broken, because only the plastic holds it together and the copper is not. If you pull hard, it'll tear at that very spot. It is a crude, but working method to find where a cable is broken. Once the cable tears, you solder it back together, insulate with shrink wrap (or electrician's tape, or any plastic tape) and it's fixed.

Now this works well on individual cables only, not on a bundle like with the E motor. That's why I suggest to untwist the four cables that come from the E motor. It's hard to pull on single cables when they're twisted together. They cannot be untwisted unless you remove the black plastic connector at the end of the cable. The black plastic can be removed by pushing on the tiny latches in holes in the back of it. If done carefully, you can put the metal contacts back into the connector when done fixing the cable.

If that's still hard to understand in English, I have an E motor here that I can post pictures of ...

This post was modified 5 years ago by Vojtěch
Posted : 16/05/2019 8:08 pm
Desti
(@desti)
Eminent Member
Topic starter answered:
RE: First layer issue at some specific bed area

Ooh okay I see the logic of it now!

Thanks a lot I'll give it a try tomorrow, it's a bit late and I'll prefer to make it in a fresher mood. I'll keep the post updated once I've done it. Thanks again for everything 🙂

Posted : 16/05/2019 8:33 pm
Vojtěch
(@vojtech)
Honorable Member
RE: First layer issue at some specific bed area

Let me add that only one out of four will likely be broken. If it was more, the motor would likely be dead all the time. So don't pull too hard, be careful, it's possible to tear a perfectly good cable, too. 🙂 And keep in mind that you could also tear the cable off the motor - so don't pull by holding the motor.

Good luck!

 

This post was modified 5 years ago by Vojtěch
Posted : 16/05/2019 8:40 pm
Desti liked
--
 --
(@)
Illustrious Member
RE: First layer issue at some specific bed area

Let's not forget to say that wire when first installed is annealed and willing to flex with out issue.  When bent back and forth over time, the anneal fails and the wire begins to work harden; and eventually it breaks. One wire breaking is a sign the wore next to it is also near breaking.  You should replace all four of the motor wires at the same time.  Fixing one will only be a short term fix: with weeks, maybe only days until the next wire fails.

Perhaps someone has  found suitable replacement wire for the various cables.  Perhaps something better, a more flexible stranding.  But the job really calls for a complete crimp set, connectors, housings, and good soldering iron and soldering skills.  What's needed is a Wiki of all the parts and pieces involved.

Posted : 17/05/2019 12:13 am
Vojtěch
(@vojtech)
Honorable Member
RE: First layer issue at some specific bed area

@tim-m30 Shouldn't I be the one asking for a complete teardown and rebuild? 😀 Yes, you're right, there is a risk that some other cables may have suffered similar fatigue, and may fail soon. So examining them all near the failure point is a good idea. But replacing every single cable to the printing head is a bit of an overkill solution.

Anyway. The connector on the motor is a Molex SL connector. The Prusa provided cable is 24AWG. So the right pins are Molex SL 16-02-0097 and the housing is Molex SL 50-57-9404. You may be able to get compatible connectors from other manufactrurers, like TE Connectivity. Any Dupont Crimp Tool will work for crimping the connectors on. It may be easier to just buy a ready made Einsy motor cable, chop the old cable off near the motor and solder the new cable on. I would still recommend to separate the individual conductors of the cable and twist them, a twisted cable is much less vulnerable to bending induced stress failure.

Posted : 17/05/2019 8:07 am
Page 1 / 2
Share: