for my bumpy bed, doing a 11x11 bed levelling is increasing the first layer quality dramatically! I am approaching a perfect first layer, but I’m not fully there. Reason is: The new heated bed calibration feature of the firmware does not seem to apply to the “manual” calibration I use for 11x11. So here’s what I do:
Use Luban to have the bed at 60° and wait a while so it stabilizes
Run the following commands:
G1029P11
G1029A
After the matrix is done, use touchscreen to adjust Z height (It is OK to disconnect from Luban inbetween!)
Run GCode:
G1029S
G1029D0
As I said, it improves my first layer by an order of magnitude, but the bed cools down during the lengthy 11x11 run. Any idea how to keep it hot?
Just another tipp to share here: I still had issues with bed adhesion - turns out I was laying down the first layer too fast. I tried it at 50 mm/s - slowing down to 30 mm/s did the trick.
I ran it and it takes around 18 minutes but is otherwise very straightforward. Also there was no need to disconnect from Snapmaker during the process of the final Z. Maybe they updated something…
The issue with the bed temperature dropping during the process is still there, although I tried resending the bed temperature via LUBAN, however by then it had already done 121 probes so had finished anyway. I did raise the temperature to 60 before doing the final Z though and that works.
That said, someone did solve the bed temperature issue and I am going to quote the post here for what it is worth, but it does involve messing with firmware and is beyond my skill level. NOTE: I did not manage to capture the whole post so there is some missing code at the end.
Here’s an update to using 11x11 pattern with my bumpy bed: I did a “really cold” calibration, i.e. the bed was at 18°C or so. The resulting matrix was rather consistent with my older, 60°C calibrations with the cooling-down bed. So I got bold and started a large print with the cold calibration data. The print had the bed at 60°C, and apart from adjusting the Z offset to +0.2 mm the first layer was really good without further ado. So for my bed the thermal expansion is consistent all over the bed and it seems I do not need to bother with heated bed calibration - which is good news!
The print covers ~2/3 of the surface - in the pipeline is one that in one dimension goes to the full size - will report if I have more difficulties there.
The developers recently accepted a pull request that updates the G1029 command to heat the bed, which will hopefully fix this when the new firmware version is released. Ideally they would add the larger mesh sizes to the GUI as well, but this is a good first step.
For quite some time, many of us have known that most printers across the board do not have a noticeable difference in geometry between a cold bed and a hot bed. While the temperature is changing (heating up or cooling down) there is a difference, as different parts are at different temperatures. However, once the temperatures have stabilized, there is very little difference. Most beds (platforms) stabilize within 30 minutes of reaching their set temperature. I am not saying that this is true for all printers, but for a very large majority, the difference is negligible.
