Auto_Bed_Leveling: Enhanced G29 command

[DarkAlchemist]

Well, rightly or wrongly I was thinking that the amount of Z being off is at the point of the probe not where the hotend is at plus the offset from what I say is the X and Y difference and what actually is the difference only helps compound any issues.

edit: I like that it tells me to turn one full turn so how do I put that to use for me and my setup and which corner is that?

[Roxy]

Well, rightly or wrongly I was thinking that the amount of Z being off is at the point of the probe not where the hotend is at plus the offset from what I say is the X and Y difference and what actually is the difference only helps compound any issues.

edit: I like that it tells me to turn one full turn so how do I put that to use for me and my setup and which corner is that?

On the Z-Probe... You can measure the Z-Offset, but it is only a guess. You have to manually fine tune that number because every micro-switch is going to fire at a slightly different place. If it is close, it isn't too hard to watch the first layer to see if the nozzle needs to be slightly closer or further away. I move the Z_PROBE_OFFSET_FROM_EXTRUDER by .1mm at a time until I get it right for however things are screwed together on my extruder. (Every time I take it apart to clear a clogged up nozzle, the value changes... And I have to move it up or down a little to get the number right.) But once it is dialed in... Every print has the first layer go down nice and clean.

On the +.14mm per turn comment... Do you know how many threads you have per inch or per mm on your bed mounting? If so, you can just calculate the number and change the +.14 turns number to what ever you have. But if you have your bed mounted on springs in the corner, you could just take a micrometer and measure the thread spacing. That is where that number comes from. But as it turns out, that number can't be used directly. Here is why: If one corner is up 1mm from the mean, you probably only want to lower it .5 mm because the mean is going to move downward as you adjust things. (I realize it seems like you should be able to move it the full amount downward, but my experience is it is better to move it halfway of what is suggested and remeasure.) It is kind of an iterative process. With 3 or 4 adjustments you should be getting really close.

[RobH2]

Hey, there is a user here that I'm trying to help out. Any of you know if the autoleveling will work with a Flashforge machine where the bed moves for the 'z' motion instead of the hot end? It seems like it should, it's just reversed sort of.

[DarkAlchemist]

On the Z-Probe... You can measure the Z-Offset, but it is only a guess. You have to manually fine tune that number because every micro-switch is going to fire at a slightly different place. If it is close, it isn't too hard to watch the first layer to see if the nozzle needs to be slightly closer or further away. I move the Z_PROBE_OFFSET_FROM_EXTRUDER by .1mm at a time until I get it right for however things are screwed together on my extruder. (Every time I take it apart to clear a clogged up nozzle, the value changes... And I have to move it up or down a little to get the number right.) But once it is dialed in... Every print has the first layer go down nice and clean.

On the +.14mm per turn comment... Do you know how many threads you have per inch or per mm on your bed mounting? If so, you can just calculate the number and change the +.14 turns number to what ever you have. But if you have your bed mounted on springs in the corner, you could just take a micrometer and measure the thread spacing. That is where that number comes from. But as it turns out, that number can't be used directly. Here is why: If one corner is up 1mm from the mean, you probably only want to lower it .5 mm because the mean is going to move downward as you adjust things. (I realize it seems like you should be able to move it the full amount downward, but my experience is it is better to move it halfway of what is suggested and remeasure.) It is kind of an iterative process. With 3 or 4 adjustments you should be getting really close.

No, I completely understand because if you move one corner up they all change slightly and the opposite corner can change dramatically. My threaded rod is M5 at 0.8 pitch so 1 full revolution moves Z by 0.8mm and my bed is hard mounted so no springs.

So, knowing that what would I change in which file and when it is telling me to rotate something which corner, or what, is it telling me to turn?

[DarkAlchemist]

Hey, there is a user here that I'm trying to help out. Any of you know if the autoleveling will work with a Flashforge machine where the bed moves for the 'z' motion instead of the hot end? It seems like it should, it's just reversed sort of.

I don't see why not because it doesn't matter what is moving only that you can stick a probe next to the hotend to measure the Z differences along the bed.

[Roxy]

Hey, there is a user here that I'm trying to help out. Any of you know if the autoleveling will work with a Flashforge machine where the bed moves for the 'z' motion instead of the hot end? It seems like it should, it's just reversed sort of.

I don't see why not because it doesn't matter what is moving only that you can stick a probe next to the hotend to measure the Z differences along the bed.

Yes, that is correct. The Auto Bed Leveling is a software based system that maps coordinates from the physical domain to a new domain that is 'corrected' for the angle of the bed. It doesn't care how the physical movement of the machine happens. It just maps coordinates to the new coordinate system and tells the machine to go there instead of where it was supposed to go.

[DarkAlchemist]

So, there is no way to test for any correctness short of wasting precious filament many times until Z is dialed in with each partial print failure?

[Roxy]

So, there is no way to test for any correctness short of wasting precious filament many times until Z is dialed in with each partial print failure?

I'm sure with lots of care, effort and the right tools there is a way. But pretty much, you have to dial it in. It won't take long. I just leave the Arduino build system up with Configuration.h in the window at the right place. To make a change, recompile, and upload the firmware takes 90 seconds. I try to do it quickly because then I don't have to wait for the bed and nozzle to heat back up. But within 5 or 10 minutes, you should be golden. You will be impressed with how the first layer always goes down nicely.

And if you use this to help calibrate your Z-Offset, I bet you can get it calibrated for under $.03 or $.04 of filament. I know its a gut wrenching waste... But try to suck it up and live with it!

[AbuMaia]

So, there is no way to test for any correctness short of wasting precious filament many times until Z is dialed in with each partial print failure?

No, you could always do the paper test. Run G28 and G29 through pronterface, then, using pronterface, lower the Z axis to where it's at its zero, then try to slide a sheet of printer paper under the nozzle.

What's odd for me, if I do just a G28, then lower Z to its zero, it's always like 4-5mm high. It's only after a G29 that Z is at its appropriate height.

[Roxy]

No, you could always do the paper test. Run G28 and G29 through pronterface, then, using pronterface, lower the Z axis to where it's at its zero, then try to slide a sheet of printer paper under the nozzle.

Yeah... That gets you close. And I'm only speaking for me... But I have to see how it lays down the bead and how much it squishes out into the line of filament next to it. The good news is after you get things how you want them, just about every print gets its first layer put down nice until you change something.