Notifications
Clear all

help with lost adhesion please?  

Page 2 / 4
  RSS
rmm200
(@rmm200)
Noble Member
RE: help with lost adhesion please?

Nowhere in your posts did you ever say you scrubbed your build plate with copious hot water and unscented detergent.

If you won't try what we recommend, there is not much we can do to help you.

Every single thing you print leaves a bit of itself behind. That has to be periodically removed. Most of the time IPA works.

Posted : 15/03/2020 7:46 pm
bobstro
(@bobstro)
Illustrious Member
RE: help with lost adhesion please?
Posted by: @justin-adie

[...] I'm open to anything sensible that doesn't involve glue - but really I want an answer to the _cause_ of 'what's changed' rather than a patch to the symptoms

Going by your description, the change is "contamination of the print surface". Whether due to touching the print surface, an ill-time sneeze or other factors, nobody here has any way of discerning. However, the suggestions you are being provide are not merely "patch to the symptoms" but the actual cure for the problem according to the manufacturer and based on accumulated experience of dozens if not hundreds of contributors to these forums. PEI does require maintenance to keep good adhesion. 

 

My notes and disclaimers on 3D printing

and miscellaneous other tech projects
He is intelligent, but not experienced. His pattern indicates two dimensional thinking. -- Spock in Star Trek: The Wrath of Khan

Posted : 15/03/2020 8:05 pm
jpadie
(@jpadie)
Eminent Member
Topic starter answered:
RE: help with lost adhesion please?

@robert-rmm200

And ipa was used immediately after removal of every print.  

Proper electronics grade IPA as I use it multiple times a day for my circuit boards .

Posted : 15/03/2020 8:59 pm
jpadie
(@jpadie)
Eminent Member
Topic starter answered:
RE: help with lost adhesion please?

@robert-rmm200

Hello

 

I said I had done that.  Referencing  fairy  liquid and following the advice of an earlier poster.  Perhaps not all replies are getting through. 

 

 

Posted : 15/03/2020 9:01 pm
jpadie
(@jpadie)
Eminent Member
Topic starter answered:
RE: help with lost adhesion please?

@bobstro

Whilst I doubt that this is the case if you are correct then I  think this answers my basic question. If a build surface doesn't survive a few hours of printing and rigid adherence to alcohol cleaning and protem acetone cleansing then it's probably fair to say that 3d printing on this i3mk3 is not yet plug play and leave and so not meeting my gateway requirements. 

I genuinely cannot afford the time to babysit a non-central tool so as per my earlier post I must thank everyone that helped and simply call it quits. 

From an engineering perspective I just doubt that the build surfa e is the culprit.  As a problem solving specialist (a lawyer and a mathematician) it feels many times more likely that something else is awry.  But either way it's the same result: this just isn't my bag!

 

 

Posted : 15/03/2020 9:08 pm
rmm200
(@rmm200)
Noble Member
RE: help with lost adhesion please?

Sorry to see you go, but do keep in mind Prusa has farms of the same printer you have making parts for printers like yours.

That is at least hundreds if not thousands of machines just like yours, running 24x7.

Posted : 15/03/2020 9:11 pm
jpadie
(@jpadie)
Eminent Member
Topic starter answered:
RE: help with lost adhesion please?

@robert-rmm200

indeed.  that's why i invested 3-4 times the price of an ender 3.  

but the investment of time is vastly more important and outweighs any possible value.

Posted : 15/03/2020 9:26 pm
bobstro
(@bobstro)
Illustrious Member
RE: help with lost adhesion please?

PEI is not exclusive to Prusa, and Prusa printers can use other build surfaces. I've used a variety, including BuildTak (popularly cloned on most low end printers), textured vinyl, blue tape, those free Overture squares and garolite. There are plenty of options. 

However, just as a soldering iron requires a bit of prep, so too do print surfaces. PEI on any other printer requires cleaning. 

My notes and disclaimers on 3D printing

and miscellaneous other tech projects
He is intelligent, but not experienced. His pattern indicates two dimensional thinking. -- Spock in Star Trek: The Wrath of Khan

Posted : 15/03/2020 9:34 pm
rmm200
(@rmm200)
Noble Member
RE: help with lost adhesion please?

Before you go - invest a bit more time and money in the BuildTek print sheet @bobstro mentioned.

One source:

https://www.matterhackers.com/store/c/BuildTakFlexPlate?rcode=PPCADS&gclid=Cj0KCQjwpLfzBRCRARIsAHuj6qXRoKXmaIKnTKlRwoiYmQqTibkNi2Gr7FOW1apZF8Lkx33xx4Xh44kaAqGZEALw_wcB

There are a number of YouTube videos on BuildTak. Any steel plate works fine with the magnet base you already have.

 

Posted : 15/03/2020 9:50 pm
bobstro
(@bobstro)
Illustrious Member
RE: help with lost adhesion please?
Posted by: @justin-adie

Whilst I doubt that this is the case if you are correct then I  think this answers my basic question. If a build surface doesn't survive a few hours of printing and rigid adherence to alcohol cleaning and protem acetone cleansing then it's probably fair to say that 3d printing on this i3mk3 is not yet plug play and leave and so not meeting my gateway requirements. 

Just to be clear: I find I can easily get by for 1-2 weeks between the washes with only minimal wiping with isopropyl alcohol in between. Increasingly, I'm finding a wipe with a thickly folded paper towel sufficient for between-print prep. I've just learned that a wash is usually the fix for any adhesion problems, so start there when problem arise. It's a bit like soldering: Frustrating  and exasperating to master the first few times, but something you quickly develop a feel for.

You indicated in an earlier post that you're printing with PLA. If you're comfortable sticking with lower-temp materials like PLA, a BuildTak surface works very well. It really should be cleaned and will deteriorate and need replacement periodically, but it's an alternative if you're not happy with PEI. I tried the textured vinyl on a lark and was surprised it adhered amazingly well for PLA. Hotter filaments will destroy these surfaces however.

My notes and disclaimers on 3D printing

and miscellaneous other tech projects
He is intelligent, but not experienced. His pattern indicates two dimensional thinking. -- Spock in Star Trek: The Wrath of Khan

Posted : 15/03/2020 9:54 pm
bobstro
(@bobstro)
Illustrious Member
RE: help with lost adhesion please?

Quick update:

  • Not all resellers carry BuildTak in the Prusa format. You can find the Mk3 version directly on their website. They also carry PEI and Nylon sheets cut similarly.
  • PrintedSolid also carries a very nice spring steel sheet if you don't want to mess up a Prusa original. It's a bit thinner (and springier) so will require a separate one-time Live-Z calibration.

Unfortunately, I'm not up on resellers in France, but I know there are suppliers for other alternate surfaces in Germany.

 

My notes and disclaimers on 3D printing

and miscellaneous other tech projects
He is intelligent, but not experienced. His pattern indicates two dimensional thinking. -- Spock in Star Trek: The Wrath of Khan

Posted : 15/03/2020 11:05 pm
jpadie
(@jpadie)
Eminent Member
Topic starter answered:
RE: help with lost adhesion please?

i'm isolating (covid-19) as a member of my household came back from his father with a temp of 105F last night. so i have a little more time to look at this.

starting with the print bed as the root cause makes little sense to me.  it's just not likely and cleansing, scrubbing etc has been done.  the flip side of the PEI sheet shows identical behaviour and had never been used.  

an Occam's razor approach suggests that the heatbed is not the place to continue researching the issue.  

which leaves either the frame or the hot-end.  

prusa's firmware suggests that the frame is squared, the z-axis perpendicular.  so not much more to do there.  i cannot perceive any misalignment with my (elderly) vision nor a steel ruler; and i'm not seeing any of the awful skew that dogged the ender 3.

however the calibration prints that have worked do show that the nozzle is scraping the print in some places and not in others.  the front left hand corner being a particular hot spot.  the scraping then lifts the extruded material and the filament starts bunching against the nozzle.  at that point the print is dead.  

so EITHER the heatbed is very out of level AND the mesh bed levelling is not being taken into account (this was a suspicion of mine with marlin 2 on the ender 3 as the z correction values it produced looked very odd; but I didn't have time to chase the code all the way through and the machine is junked now); OR something else is causing the misalignment.  

two things appear awry on the extruder assembly to me:

1. the hot end freely rotates.  there is no seepage from the heatbrake or anywhere other than the nozzle itself, but the heater block freely spins.  there is nothing to secure or retain it.  Is this normal?  

2. the entire heat sink assembly inside the printed plastic parts has some play in it.  a few mm on prodding.  however all the plastic parts are snug.  is this normal?

thanks

Posted : 16/03/2020 10:10 am
jpadie
(@jpadie)
Eminent Member
Topic starter answered:
RE: help with lost adhesion please?

I made an error above.  I have now disassembled the extruder and find that it was not the heater assembly that was freely rotating but the entire hot end as a single piece.  

Apologies for misleading any reader

Posted : 16/03/2020 10:32 am
rmm200
(@rmm200)
Noble Member
RE: help with lost adhesion please?

I personally think Occam needs a new strop.

Adhesion issues are always the print surface. It is not the frame - it is not the hot end.

If your car slides off the road on an icy day - it is an adhesion issue. You can replace the carb - but it is unlikely to help.

Please print a calibration square (about 2 inches, center of the build plate) from this link and post a picture.

https://forum.prusa3d.com/forum/original-prusa-i3-mk3s-mk3-assembly-and-first-prints-troubleshooting/life-adjust-z-my-way/

We can tell more from the picture than your descriptions.

Posted : 16/03/2020 4:21 pm
jpadie
(@jpadie)
Eminent Member
Topic starter answered:
RE: help with lost adhesion please?

There is a photo in the thread.  

 

But since that time there has been no print successful.  Just filament wafting across the build plate and being bundled into the nozzle.  Irrespective of z height. 

Posted : 16/03/2020 4:28 pm
rmm200
(@rmm200)
Noble Member
RE: help with lost adhesion please?

Lets start over.

What filament are you using, what temperature is the hot end and what temperature is the bed?

A bed significantly under temp could cause this. A slipped PINDA could also if the nozzle is "way up in the air".

I would at least check the PINDA spacing. One zip tie distance above the nozzle when touching a piece of paper.

Posted : 16/03/2020 4:36 pm
bobstro
(@bobstro)
Illustrious Member
RE: help with lost adhesion please?
Posted by: @justin-adie

i'm isolating (covid-19) as a member of my household came back from his father with a temp of 105F last night. so i have a little more time to look at this.

Best of luck to you. We're hunkering down and limiting activities. This will be a good time for idle pursuits to keep us busy.

starting with the print bed as the root cause makes little sense to me.  it's just not likely and cleansing, scrubbing etc has been done.  the flip side of the PEI sheet shows identical behaviour and had never been used.  

It is however part of the process of elimination. You do the basic things to eliminate them from the equation, not because you "know" they are the cause of the problem. How do you remove parts from the bed? Do you lift it or handle it at all? If so, and you're using isopropyl alcohol wipes, you're likely just moving around broken-down grease. A wash rinses it away due to the sheer volume involved.

an Occam's razor approach suggests that the heatbed is not the place to continue researching the issue.  

Not sure how you got there. Occam's Razor suggests that the simplest answer is usually correct one. It's hard to get simpler than a surface that is physically touched and manipulated regularly might be a problem. I really don't understand your reluctance to give the sheet a wash.

[...] however the calibration prints that have worked do show that the nozzle is scraping the print in some places and not in others.  the front left hand corner being a particular hot spot.  the scraping then lifts the extruded material and the filament starts bunching against the nozzle.  at that point the print is dead.  

Have you enabled 7x7 mesh bed leveling? That can help with a slightly warped bed.

[...] 1. the hot end freely rotates.  there is no seepage from the heatbrake or anywhere other than the nozzle itself, but the heater block freely spins.  there is nothing to secure or retain it.  Is this normal?  

2. the entire heat sink assembly inside the printed plastic parts has some play in it.  a few mm on prodding.  however all the plastic parts are snug.  is this normal?

It should not be wiggling around, but it's not held in by bolts. I'd pop off the extruder cover and give it all a good look.

 

My notes and disclaimers on 3D printing

and miscellaneous other tech projects
He is intelligent, but not experienced. His pattern indicates two dimensional thinking. -- Spock in Star Trek: The Wrath of Khan

Posted : 16/03/2020 5:33 pm
rmm200
(@rmm200)
Noble Member
RE: help with lost adhesion please?

Also, nothing secures the hot end in one position. It just a round object sitting in a recess.

It is like a straw filament passes through. If the straw rotates a bit - so what. It won't catch on the object being printed.

The nozzle is symmetric, and orientation does not matter.

Posted : 16/03/2020 5:42 pm
jpadie
(@jpadie)
Eminent Member
Topic starter answered:
RE: help with lost adhesion please?

@robert-rmm200

thanks for the reply Robert.

What filament are you using, what temperature is the hot end and what temperature is the bed?

I'm primarily using PLA. I have tried with two reels and also with a reel of PETG. last week, when the prints were working beautifully, I had printed a good number of parts in PETG and a good number in PLA.  It was all good.

the hot-end and bed-temp are either whatever is pre-set in the gcode of the sample prints or whatever comes out of prusa-slicer natively. I've not fiddled with those.  typically looks like 215/60 for PLA.

A bed significantly under temp could cause this. A slipped PINDA could also if the nozzle is "way up in the air".

bed temp seems fine.  too hot to touch for a sustained period.  PINDA did occur to me.  but i have reseated the probe and redone the wizard four times now.   the distance from the bed is important, I assume, to calculate the level of inductance that should trigger the probe.  so I would think that as long as the probe calibration is done that's enough.  the rest is down to z-offset.  In any event, I stripped the extruder assembly right down this morning as that was the last thing left for me to try, so when I assemble it i will match it to the zip tie again. 

The nozzle is symmetric, and orientation does not matter

I don't think that is the case here.  It seems that on rotation the cables from the hot end press against the side of the housing and physically _move_ the nozzle out of orientation.  this is made worse with the build up of the bundle of filament as that then forces against the bed and increases the lateral moment.  The extruder had millimetres of lateral play in the forward direction (towards the front of the heat bed), I'd guess, not just rotational.

@bobstro

This will be a good time for idle pursuits to keep us busy.

true indeed!  I have a bedroom to paint, a heat pump to repair and a 15 macs in my repair queue at the moment. Fun and games for all the family.  France (where I am) is on curfew from tomorrow, so the rumour goes.  All non-essential travel is to be banned.

You do the basic things to eliminate them from the equation, not because you "know" they are the cause of the problem.

agreed.  But I have already done everything possible to the steel sheet short of putting glue on it - and it still does not explain why from one moment to another the whole print fails when there is no external intervening change - that feels much more like an alignment issue to me.  Also to note, as I said, I've tried the underside of the sheet too, to zero avail. 

So I'd like to consider these things eliminated from the matrix.

I really don't understand your reluctance to give the sheet a wash.

 

I really don't understand why my posts are not appearing for other people.  I can read them!  I posted on 13/3 at 8h51 that I had followed your advice and scrubbed the board with fairy liquid.

Have you enabled 7x7 mesh bed leveling? That can help with a slightly warped bed.

yes.  i have enabled this, but again all prints were perfect a few days ago.  I find it difficult to see how such significant warping can occur overnight when there was such a minor change in ambient temperature and no other external intervening factor. 

It should not be wiggling around, but it's not held in by bolts. I'd pop off the extruder cover and give it all a good look.

I have stripped the extruder assembly completely and rebuilt it.  the heat block and heat sink no longer rotate and there is no general play.  One of the long bolts had sheared, improbably.  I do wonder whether the lack of rigidity in the extruder is a blame-factor here.  

Posted : 16/03/2020 6:20 pm
rmm200
(@rmm200)
Noble Member
RE: help with lost adhesion please?

First layer "printing in air" bothers me.

Working properly, the first layer is "squished" (technical term) against the plate, and would be flat - not round.

There should be NO movement of the extruder in the Y axis. And a sheared screw does not sound good...

This is plastic - don't tighten the screws enough to shear. Could part of the extruder be cracked?

See if you can locate where the play is coming from. If it is in the area of the bearings - something is wrong. Mounts or something.

This post was modified 4 years ago by rmm200
Posted : 16/03/2020 6:32 pm
Page 2 / 4
Share: