Welcome the world of problems we wouldn’t have without Bambu printers !
Ok, jokes aside, let me explain the problem at hand:
The base has a certain thickness and infill.
Once the top layer of the base is done only the walls and their infill remain.
A significant drop in layer time.
In theory this should not matter and does not matter with most printers and slicers out there - at least not to this degree.
The filament heats up and the flow rate and print speed affect the time it spends in the melting chamber.
Have a glue gun and sticks ? Great !
Check what happens once the heating light goes off.
The thing is ready for use now - but using it NOW is much harder than a few minutes later.
And if you need a lot of glue you notice how the gun might struggle to heat it up fast enough.
You face the opposite problem.
From having to pump at a fixed rate the hotend has to do it now at a different speed, affecting the filament properties.
Sadly Bambu did not stop here and somehow managed to affect the actual path of the extrusion as well.
The slicer/printer can ‘compensate’ in two ways.
- by slowing things down do match the flow/pressure changes so they won’t have a severe effect - not a favour for fast printers…
- by slightly adjusting the extrusion path for the walls.
And that seems what Bambu is in favour of, not sure is just slicer based or in the firmware.
If you check the slicer preview you will find corresponding speed/flow rate changes in the affected layers.
In Studio 1.9.5 you can try the discontinuity feature in order to mitigate the issue a little bit.
In reality though only adjusting the speeds so the flow rate changes and layer time changes are less severe helps.
I tried to fix the issue through proper filament calibration but it is impossible to get rid of the effect for good.
On models with flat walls you can nicely observe how the affect changes from the start to the end of the extrusion line for the wall.
So really no chance to calibrate things that keep changing on the go…
We can only hope Bambu addresses this issue in a future bug fix as it is going on for a very long time now…
For me the striking part is that in many cases the defects are far too severe to be possible without a change for the actual path the nozzle takes.
I had models with a difference of over 1mm - and with clear stepping indicating the path changed to a larger or smaller circumference…