Questions about How Auto Bed Leveling Works

[gmay3]

Thanks very much Roxy, it's cool to see how the code creates this matrix of measurements. The amount of work to implement auto bed leveling in the Marlin Firmware must have been extraordinary!

Are the measurements taken every time a print is starting up as part of the normal printing routine or are they done only when you want?

If it is automatic/included in every print start-up sequence, does it measure while the bed is hot?

[Roxy]

Thanks very much Roxy, it's cool to see how the code creates this matrix of measurements. The amount of work to implement auto bed leveling in the Marlin Firmware must have been extraordinary!

Are the measurements taken every time a print is starting up as part of the normal printing routine or are they done only when you want?

If it is automatic/included in every print start-up sequence, does it measure while the bed is hot?

Whoever did the original work was a truly gifted individual. It was a lot of work.

There was somebody here exploring the possibility of saving the bed correction matrix in the EEPROM so it doesn't have to be constructed at the start of each print. I don't know where in the process that work is. Or if it will see the light of day. It really isn't that hard to do. All the support routines for the EEPROM and default settings is in place to do it.

So right now... You just do a G28 and a G29 at the start of each print and the correction matrix is generated and used during the entire print. I have my Slicer Start Up GCode wait until the bed is up to temperature before doing the probing. That seems to help accuracy (at least for me). And I have the nozzle coming up to temperature during the probing. That works out pretty good for me.

[gmay3]

There was somebody here exploring the possibility of saving the bed correction matrix in the EEPROM so it doesn't have to be constructed at the start of each print. I don't know where in the process that work is. Or if it will see the light of day. It really isn't that hard to do. All the support routines for the EEPROM and default settings is in place to do it.

Interesting. I was under the impression that the EEPROM has a more limited number of writes/rewrites then typical memory and that it should only be used when necessary for permanent value changes, but I could be totally wrong! I don't even know if the RAMPS has another type of memory.

So right now... You just do a G28 and a G29 at the start of each print and the correction matrix is generated and used during the entire print. I have my Slicer Start Up GCode wait until the bed is up to temperature before doing the probing. That seems to help accuracy (at least for me). And I have the nozzle coming up to temperature during the probing. That works out pretty good for me.

So could one bake the G28 and G29 into the start up GCode after the bed has reached its desired temp?

I'm guessing this is a non-issue but does the probe switch ever show signs of melting with the heat from the bed?

[Roxy]

ooops... I started this post before lunch... And then just finished it. I didn't see that everybody else had jumped in!

Interesting. I was under the impression that the EEPROM has a more limited number of writes/rewrites then typical memory and that it should only be used when necessary for permanent value changes, but I could be totally wrong! I don't even know if the RAMPS has another type of memory.

It is limited in the number of erase cycles that it can do... But it is guaranteed to be up around 10,000 or so. Probably way more. And the whole intent of doing this is so you don't have to do the probing at the start of each print. So until your bed gets too far out of kilter, you would not re-probe and save the new settings. You would just keep using the previous bed level correction matrix.

So could one bake the G28 and G29 into the start up GCode after the bed has reached its desired temp?

Yes. That is what is done. This is what my Slicer Start GCode is:

G1 Z15 F2000 ; lift nozzle
G28
G29 v4 n3; Do bed leveling probe
G1 Z15 F2000 ; lift nozzle

I'm guessing this is a non-issue but does the probe switch ever show signs of melting with the heat from the bed?

No... The plastic roller I have on my super cheap micro-switch is just fine. I suspect the plastic doesn't melt until a much higher temperature. But it is only touching for an instant. And the bed is at a much lower temperature than the nozzle. I do get a little bit of a coating of hair spray on it however!

[gmay3]

Yes. That is what is done. This is what my Slicer Start GCode is:

G1 Z15 F2000 ; lift nozzle
G28
G29 v4 n3; Do bed leveling probe
G1 Z15 F2000 ; lift nozzle

Ok perfect! I guess this automatically runs when the printer is ready to start reading the gcode after it's heated up. That makes things easier, I thought I was going to have get my stop watch out and put time delays for how long it takes for the bed and nozzle to heat up in the startup gcode or some silliness. Makes sense!