I have incremental bed leveling mostly working in Smoothieware.

Controllers Smoothie
,
1

I have incremental bed leveling mostly working in Smoothieware.

Originally shared by Nicholas Seward
https://www.youtube.com/attribution_link?a=lbQAJy4gM6c&u=/watch?v%3DT8sIGOBbTHg%26feature%3Dplayer_embedded

2

Really nice :slight_smile:

3

Excellent!!!

4

I look forward to reading the code updates you made to Smoothie, when might you be posting it to GitHub?

5

I will publish in this next week. I will need someone that actually knows what they are doing to get ready for the main branch.

6

When you make your GitHub available I will have a (quick) look at it :slight_smile:

7

To be honest I really do not think we need yet another bed leveling method, so i recommend keeping it in a fork and we can list it in the third party branches. We do not have unlimited flash space unfortunately.

8

looks pretty much like the grid leveling that was recently added that allows you to set arbitrary rectangles to level.

9

@Wolfmanjm I don’t really care if it is included or not but for the record it is a much more generic approach than arbitrary rectangle probing. It uses about the same or less code to accomplish a more generic result.

You could do spiral probing which would be great for polar/delta printers. You could do rectangle probing which is good for most printers. You could do incremental probing for any odd shaped beds to avoid clips and other things. My point is that probing every point in a grid is a waste of time and limits you to a rectangular probing area. You could even do strange/useful things such as chain the probing of multiple rectangles and circles together to define a shape useful to you. (You could have an option to turn off re-probing within a certain distance of an existing probe point.)

I would be glad to work with you guys to provide a generic (as backwards compatible as makes sense) leveling module. I think making one module handle arbitrary rectangles, circles, and incremental probing would be worthwhile. There is really no reason to have >3 point based leveling reside in multiple files (CartGridStrategy.cpp and DeltaGridStrategy.cpp and Morgan has yet another one). I am also happy just doing my own thing and scratching my own itch.

10

Here is some really gross code. https://github.com/NicholasSeward/proto_RepRap_HELIOS/blob/master/Smootieware/src/modules/tools/zprobe/IncrementalStrategy.cpp

It is only a proof of concept. It currently, uses all the points for calculating your adjusted Z for every little move. I have it provisioned to fill in a grid after every new probe but I have some little bug somewhere. This way you could do quick bilinear interpolation.

Let me know if there is any interest in including this and I will correctly fork the repo and create some more thoughtful code.