RE: help with lost adhesion please?
@robert-rmm200
do you meant the flat calibration square by Jeff Jordan? if so, I can do that!
RE: help with lost adhesion please?
@robert-rmm200
do you meant the flat calibration square by Jeff Jordan? if so, I can do that!
That is my all-time favorite. But pretty much anything is better than the built-in one.
RE: help with lost adhesion please?
Whatever you did with your pictures - those last ones came through well.
I can't see 'em! 😕
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
RE: help with lost adhesion please?
Could someone summarize how to post pictures again? I am sure Joan did - but I can't find it.
JPG, magic with the media button, etc.
RE: help with lost adhesion please?
all the photos are fully visible to me.
here is a link to a google album: https://photos.app.goo.gl/5UFBUwi4a9CnQXAF8 that shows the calibration square.
RE: help with lost adhesion please?
Could someone summarize how to post pictures again? I am sure Joan did - but I can't find it.
The quick & dirty version is to drag an image file into the edit box when composing a post. The file will upload, then you can press the Insert into post button. It's a little funky but it works. I think Tim-m30 had another process but I haven't bothered given the odd state of the forums lately.
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
RE: help with lost adhesion please?
bedtime in my world. Just before retiring I printed another calibration square. Again _nothing external intervened_ since the last one.
here is the result. I clearly still have a serious issue somewhere.
I tried to print a more complex first layer calibration too (before). this was it: https://www.thingiverse.com/thing:4096848. Completely failed. no usable lines of filament anywhere on the build surface.
RE: help with lost adhesion please?
I really truly wish you had access to a different print surface to test...
Either filament is not adhering to your surface or your nozzle has decided to drift way high.
I hate glue sticks - but that may be a painless thing to try for a test?
RE: help with lost adhesion please?
i have a glass heat bed for the ender 3 and the mag sheet for the ender 3. neither will fit exactly but calibration prints could be adjusted to fit within their boundaries.
my key question is always what can passively change (without any active external intervention) between prints that can cause such wildly different results? i'm wondering whether it can be environmental given the never ending issues I also had with the ender 3 (which had new extruders, new nozzles, new hot ends, new y gantry, new ABL sensors etc in the pursuit of a solution to this class of issue).
I live in the south west of france in a timber framed house. there is nothing particular about the environment that I know of other than the sometimes warmer days and the sometimes cooler nights.
I am really loathe to use glue as I am not sure what it really tells me experimentally as the PEI _should_ adhere just fine. i cannot see how both coated sides could fail at the same time after the same print. it's too coincidental! but I'm the first to admit that I may not be thinking straight about this.
My best guess is that there must be some skew in the z-axis irrespective of calibration. i cannot see any way of testing this, however.
the pinda probe does not move (although the mount is very poorly designed imo - it should be held rigid in more than one plane).
another potential gremlin is the y-gantry. that has deliberate, built in play which would be removed by linear rails. but that itself would result in a very unforgiving geometry and the mk3s is designed, it seems, to have forgiving geometry and allow/cater for such play. the last potential gremlin is simply that ABL does not work well in this firmware. but that would prejudice every user, not just me. so is very low on the likelihood spectrum.
RE: help with lost adhesion please?
There have been instances of plain - defective - PEI plates. Not many. You may just be lucky.
Use the Ender 3 mag sheet if you can. The thickness will be different that Prusa, but you are printing a calibration square anyway.
Anything to get your current sheet out of the equation. Either we vindicate it - or prove it is the source of the problem.
The calibration square is only a 2 inch square (roughly). Should fit in the center of the plate nicely.
RE: help with lost adhesion please?
Use the Ender 3 mag sheet if you can.
yes. it's currently sitting in a box in my trailer ready for the recycling centre; which is now closed for a couple of months. silver linings!
will give it a whirl in the morning.
thanks
justin
RE: help with lost adhesion please?
apologies for radio silence. I seem to have broken my ankle so pain and sense kept me horizontal for a day.
since my last post I ran perhaps 50 calibration prints. I was not able to use the magnetic bed from the ender3 as ABL kept failing on the Prusa.
I got closer and closer to a good z-offset and adhesion magically returned. Some photos in this album: Thursday calibration prints.
After printing the calibration squares and getting closer and closer to what felt 'ideal' i changed to a full bed calibration print and at the end of that felt that I was 'close enough'. Photos of those are in the album too.
Feeling quite confident at the end (even given the artefacts) I went for a torture test of the easter egg by Angus @ makers muse. The first didn't separate neatly so I lifted the z-height a bit and printed it again. Unfortunately the second did not separate either. A skill knife sorted both out however.
I felt confident then that the issues were behind me. And left a shortish print going overnight. Literally no intervening change other than to put the gcode on to the sd card. The print was back to being something that would be useful for me (a jig for sonoff minis). Unfortunately I came down in the morning to find this:
I do understand the variability of temperature, grease and filament composition. But given that there was perhaps only 25 minutes of intervening time between the two prints (if that), the variability must surely come from the geometry; but I cannot see where and z-calibrations show the axes hitting the top of the gantry at the same time.
I tried the print again, going back to the live-z that the first easter egg was printed at (0.1 difference). same again.
honest opinion please: is this the moment to give up and go back to Prusa saying 'something is arbitrarily wrong'?
RE: help with lost adhesion please?
hmm. the inline photos are very squished in the view page but are fine when uploaded and in preview mode. site glitch? I have added an album here too: https://photos.app.goo.gl/WGWBBnFqwvwZWg398