Nozzle suddenly scrapes the prints, then rips PEI off the build plate

Hello everyone, I’ve been baffled by a strange behaviour from my printer for the past few days:

Strange bug. Can you assist in debugging?

First I’ve had an hyper-adhesion between printed PETG (HF-PETG by Jayo) and the textured PEI buildplate, and the nozzle scraped the first layers causing some. I attributed the issue to either over-extrusion, poor retration, too high of a temperature or a Z-offset issue. I attempted to solve this by lowering the bed temperature and nozzle temperature, which improved the issue for one day, but now the nozzle scrapes even more into the bed, even after recalibrating Z-offset, redoing the bed leveling with the screws and Z-leveling. Testing one more time led me to the nozzle ripping a couple beads of PEI off the bed

I’m using a SV07+ which I acquired last year, with no specific issues at the time of purchase, with Orcaslicer 2.4.2 with Christian Vick’s SV07+ presets, and I have the Klipper printer additions on the printer itself. https://www.printables.com/model/585691-klipper-printer-additions-30b91-more-than-boring-s

Right now I’m wondering if I need to disassemble the printing head to replace the Z-probe somehow or if I need to do something else entirely to repair my printer

Welcome to the forum.

First thing, enter PROBE_ACCURACY in the console and see what you get. Arrow around and test 4 or 5 spots.

Next

The CV extensions maintain a “library” of bed meshes for different bed temperatures. Works fantastic EXCEPT if you touch the bed leveling knobs ALL stored meshes must be deleted or regenerated. I bet CV is loading an outdated mesh for your current bed temp.

Read through the documentation PDFs in the CV .zip

Thanks a lot, I’ve managed to renew the meshes without further incidents and I’m going to run a test print today

Just ran a first test with adaptive mesh mode, scraped another line off my bed, then with multiple mesh mode, scraped another line. I’ve had to input a 0,9mm Z-offset to stop the scratching, with poor surface results anyway.

I’m puzzled by this problem…

The bed meshes might be the source of the probing error

After cleaning the underside of the bed and manually refreshing all of the bed meshes, the nozzle keeps scraping the PEI unless i add a Z-offset in Orca.

I’ve also noticed that during the bed probing sequences, the nozzle bumps into the PEI…

I think I’m looking at a probe issue

Did you do this test?

From PROBE_ACCURACY I get :

probe accuracy results: maximum 0.654000, minimum 0.646500, range 0.007500, average 0.651250, median 0.651500, standard deviation 0.001750

From what I read it doesnt explain my issue, and while running the command it seems like the nozzle still bumps the buildplate

Well for some reason the silicon socket and nozzle seemed to be hanging too low, and after trying to check the adjustment with a spanner, it somehow slid upwards, and doesnt seem to bump into the bed anymore. Very strange

Remove the silicone sock. Use your phone or a small mirror to look straight up at the heat block.

You’ll find 2 deeply recessed screws that are supposed to hold the heat block up against the heatsink. It is possible the holes have filament residue you’ll need to extract to get access to the screws.

Interesting, a screw fell out the moment I pulled it off

I’ve followed your advice and replaced a damaged screw in the heat block, put a wee bit of threadlocker for good measure, ran the Z-level calibration and refreshed the bed meshes, and could print a benchy without scraping the bed

Thank you so much!