Auto_Bed_Leveling: Enhanced G29 command

[DarkAlchemist]

Well, I came in here and noticed my m5 nut had come out of the Z on the right side so no wonder those numbers were that bad (this has been an issue with this printer from day one even with brand new parts printed for it),

Here is my real matrix (more like what I am used to seeing too):

Bed Height Topography:
+1.27089 +0.60914 --0.06336 --0.69936
+1.00064 +0.39539 --0.22661 --0.81211
+0.74964 +0.20364 --0.37411 --0.91611
+0.46789 --0.03111 --0.55411 --1.02036

[Roxy]

Those numbers are still higher than you want them to be... But the Auto Bed Leveling can correct those just fine. The problem mostly comes in when you are trying to set the nozzle height precisely for the first layer. The less that the Auto Bed Leveling has to correct, the better....

What happens when you try to print? Where is the nozzle?

[DarkAlchemist]

Well, is there a way I can do g29 and move Z to 0 and see where the nozzle is over printing? I have never had an issue with printing only with it being smushed down to the glass.

[Roxy]

If you use PronterFace to give it the G28 and the G29... It should tell you how high it thinks the nozzle is above the bed. You can click the Z Axis move buttons and move it down .1mm at a time until you get close to 0.0 You should be able to see which way you have to adjust the Z_PROBE_OFFSET_FROM_EXTRUDER number.

[DarkAlchemist]

>>> m114
SENDING:M114
X:192.02 Y:172.00 Z:7.79 E:0.00 Count X: 192.08 Y:171.98 Z:6.53
>>> m114
SENDING:M114
X:192.02 Y:172.00 Z:0.19 E:0.00 Count X: 192.02 Y:172.00 Z:-1.07

Notice 1.6 difference until it hit my card. Now 0.0 means it would be smack up against the glass which is not how I adjusted it with the 1.6.

[Roxy]

Well... Bump the Z_PROBE_OFFSET_FROM_EXTRUDER to be a bigger negative number. Try adding .25mm to it and see if the nozzle ends up closer to the right place.

UPDATE: I'm off to the library to study. But we can pick this up tomorrow.

[DarkAlchemist]

>>> m114
SENDING:M114
X:192.00 Y:170.00 Z:7.82 E:0.00 Count X: 191.99 Y:169.99 Z:8.10
>>> m114
SENDING:M114
X:192.00 Y:170.00 Z:0.52 E:0.00 Count X: 192.00 Y:170.00 Z:0.80

#define Z_PROBE_OFFSET_FROM_EXTRUDER -1.85

Now shouldn't that be zero when it hits my card?

Bed Height Topography:
+0.08786 +0.07386 +0.06711 +0.10086
--0.05364 --0.01064 +0.03186 +0.11586
--0.17089 --0.07164 +0.00086 +0.12311
--0.26139 --0.13414 --0.03164 +0.13261

edit: Interesting but not by making the offset larger but by making it smaller does the gap increase. That is contrary to what I thought it should be.

So, the closer to 0 that offset is the wider the gap from the bed becomes until the gap is at its widest with the offset set at 0. Honestly this makes no sense unless there is something hidden with a static value or something.

[Roxy]

Bed Height Topography:
+0.08786 +0.07386 +0.06711 +0.10086
--0.05364 --0.01064 +0.03186 +0.11586
--0.17089 --0.07164 +0.00086 +0.12311
--0.26139 --0.13414 --0.03164 +0.13261

That looks better! From that set of numbers it looks like your bed is very flat except for the very left side.

edit: Interesting but not by making the offset larger but by making it smaller does the gap increase. That is contrary to what I thought it should be.

So, the closer to 0 that offset is the wider the gap from the bed becomes until the gap is at its widest with the offset set at 0. Honestly this makes no sense unless there is something hidden with a static value or something.

Are you using a switch? Or a Hall Effect sensor? Because with a switch, you have to press the button until it triggers. And that gets included (Or subtracted from) your Offset. But either way, don't get hung up on that number. Just adjust it up or down until you get the right separation for the first layer of the print to go down nicely.

There is some history here that is causing problems for you. The G28 was written, debugged and used for a long time. Then the G29 came along. And both of them measure or probe with a different technique. I'm sure that will get resolved and both will do it the same way eventually. But if you look at the code for each of them, they do their thing different ways. That is causing some differences in the Zero numbers. Just ignore it. Set your offsets so things work well after the G29.

[DarkAlchemist]

Yep, I noticed that straight away and man does it get confusing when they both use the same numbers in different ways and have different meanings.

Yes, I am using a switch and I don't see the not flatness on the left side as you do but I also have it over constrained and that I can physically see. I was just trying to get those numbers down but I can tell you, as I do everyone, I can't get this bed level. I wish I could because I don't actually like G29 as it is wearing out my M5 leadscrews (you can physically see it as well as get the blackish residue off after each print). If I could get this bed level and never have to G29 again I would be so happy. G28 I love with the dynamic Z endstop but G29 I really wish I could eliminate.

[Roxy]

Yep, I noticed that straight away and man does it get confusing when they both use the same numbers in different ways and have different meanings.

Yes, I am using a switch and I don't see the not flatness on the left side as you do but I also have it over constrained and that I can physically see.

I was wrong... I didn't look at it long enough. The left column is fine. I made a mistake as I was calculating the distance between each number. I wish my bed was as flat as yours!

+0.08786 +0.07386 +0.06711 +0.10086
--0.05364 --0.01064 +0.03186 +0.11586
--0.17089 --0.07164 +0.00086 +0.12311
--0.26139 --0.13414 --0.03164 +0.13261

I was just trying to get those numbers down but I can tell you, as I do everyone, I can't get this bed level.

I think this is very common. Hence the pull towards Auto Bed Leveling!

I wish I could because I don't actually like G29 as it is wearing out my M5 leadscrews (you can physically see it as well as get the blackish residue off after each print). If I could get this bed level and never have to G29 again I would be so happy. G28 I love with the dynamic Z endstop but G29 I really wish I could eliminate.

I know what you mean about the blackish residue. But I don't think the lead-screws or the nuts on them are actually wearing. I have a light coating of very light oil on them. And I think that is what attracts the blackish residue. But my Z Axis have been going back and forth, back and forth for over a year now. And I don't feel any slop because of worn lead-screws or nuts. And even if they did wear out over a year's worth of time, they are easy to replace. My guess is you won't have to replace them. And I don't think you need to clean the blackish residue every print either!