Auto_Bed_Leveling: Enhanced G29 command

[DarkAlchemist]

Well, I don't use repetier as a host and have none of these issues but I think someone said they tried pronterface, which is what I use, and had the same issues so I am stumped as to what is going on. As far as the timeout I have mine left at 300.

[paulnd]

Well, I don't use repetier as a host and have none of these issues but I think someone said they tried pronterface, which is what I use, and had the same issues so I am stumped as to what is going on. As far as the timeout I have mine left at 300.

I use repetier host but had the same problem on pronterface

[squirrel]

Well, I don't use repetier as a host and have none of these issues but I think someone said they tried pronterface, which is what I use, and had the same issues so I am stumped as to what is going on. As far as the timeout I have mine left at 300.

I believe Tyson uses pronterface... I also left the servo time out at 300, but extended stepper time out from 60 to 180 which didn't seem to make a difference. Id like to say I'm stumped but to be honest I've had very little clue from the start

[tweellt]

Hello guys,

I've changed G29 command and probe_pt code like Roxy explained on his first post. My glass doesn't seem to be totally flat, but (correct me if I'm wrong) the enhanced G29 should compensate for that, right? No?
When I print, the nozzle is closer to the bed on the right side, so I have less adhesion on the left and too much on the right :-/

Here are the outputted values after G29 n 4 T 4:

Bed x: 40.00 y: 50.00 z: 4.09
Bed x: 86.00 y: 50.00 z: 4.03
Bed x: 132.00 y: 50.00 z: 3.92
Bed x: 178.00 y: 50.00 z: 3.78
Bed x: 40.00 y: 80.00 z: 4.35
Bed x: 86.00 y: 80.00 z: 4.22
Bed x: 132.00 y: 80.00 z: 4.08
Bed x: 178.00 y: 80.00 z: 3.93
Bed x: 40.00 y: 110.00 z: 4.64
Bed x: 86.00 y: 110.00 z: 4.53
Bed x: 132.00 y: 110.00 z: 4.24
Bed x: 178.00 y: 110.00 z: 4.05
Bed x: 40.00 y: 140.00 z: 4.88
Bed x: 86.00 y: 140.00 z: 4.70
Bed x: 132.00 y: 140.00 z: 4.47
Bed x: 178.00 y: 140.00 z: 4.11
Eqn coefficients: a: -0.00 b: 0.01 d: 4.04
Mean of sampled points: 4.250922


Bed Height Topography:
+0.62408 +0.44858 +0.21983 --0.13842
+0.39233 +0.27833 --0.00667 --0.19992
+0.10383 --0.03267 --0.17217 --0.32142
--0.16342 --0.22417 --0.33292 --0.47517

Bed Level Correction Matrix:
0.999993 0.000000 -0.003825
0.000025 0.999978 0.006613
0.003824 -0.006613 0.999971


Any ideas?

Cheers

[tstockgl]

Hello guys,

I've changed G29 command and probe_pt code like Roxy explained on his first post. My glass doesn't seem to be totally flat, but (correct me if I'm wrong) the enhanced G29 should compensate for that, right? No?
When I print, the nozzle is closer to the bed on the right side, so I have less adhesion on the left and too much on the right :-/

Here are the outputted values after G29 n 4 T 4:

Bed x: 40.00 y: 50.00 z: 4.09
Bed x: 86.00 y: 50.00 z: 4.03
Bed x: 132.00 y: 50.00 z: 3.92
Bed x: 178.00 y: 50.00 z: 3.78
Bed x: 40.00 y: 80.00 z: 4.35
Bed x: 86.00 y: 80.00 z: 4.22
Bed x: 132.00 y: 80.00 z: 4.08
Bed x: 178.00 y: 80.00 z: 3.93
Bed x: 40.00 y: 110.00 z: 4.64
Bed x: 86.00 y: 110.00 z: 4.53
Bed x: 132.00 y: 110.00 z: 4.24
Bed x: 178.00 y: 110.00 z: 4.05
Bed x: 40.00 y: 140.00 z: 4.88
Bed x: 86.00 y: 140.00 z: 4.70
Bed x: 132.00 y: 140.00 z: 4.47
Bed x: 178.00 y: 140.00 z: 4.11
Eqn coefficients: a: -0.00 b: 0.01 d: 4.04
Mean of sampled points: 4.250922


Bed Height Topography:
+0.62408 +0.44858 +0.21983 --0.13842
+0.39233 +0.27833 --0.00667 --0.19992
+0.10383 --0.03267 --0.17217 --0.32142
--0.16342 --0.22417 --0.33292 --0.47517

Bed Level Correction Matrix:
0.999993 0.000000 -0.003825
0.000025 0.999978 0.006613
0.003824 -0.006613 0.999971


Any ideas?

Cheers

I created one of these for myself just to help visualize things. I plugged in your probe points for this one. Thought I'd share. I've scaled things a bit proportionately so you can see things better:

https://drive.google.com/file/d/0B2R...ew?usp=sharing

Hope that helps!

Tyson

[tweellt]

Thank you Tyson, that helped a lot!!!

Your "plot" showed me that I had the wrong origin point, it should be ORIGIN_BACK_LEFT instead of ORIGIN_FRONT_LEFT.
I did some tweaks and I think I'm getting somewhere.

Latest output:

Bed x: 40.00 y: 50.00 z: 4.28
Bed x: 86.00 y: 50.00 z: 4.21
Bed x: 132.00 y: 50.00 z: 4.02
Bed x: 178.00 y: 50.00 z: 3.85
Bed x: 40.00 y: 80.00 z: 4.48
Bed x: 86.00 y: 80.00 z: 4.37
Bed x: 132.00 y: 80.00 z: 4.25
Bed x: 178.00 y: 80.00 z: 3.95
Bed x: 40.00 y: 110.00 z: 4.68
Bed x: 86.00 y: 110.00 z: 4.53
Bed x: 132.00 y: 110.00 z: 4.33
Bed x: 178.00 y: 110.00 z: 4.00
Bed x: 40.00 y: 140.00 z: 4.89
Bed x: 86.00 y: 140.00 z: 4.65
Bed x: 132.00 y: 140.00 z: 4.39
Bed x: 178.00 y: 140.00 z: 4.04
Eqn coefficients: a: -0.00 b: 0.00 d: 4.37
Mean of sampled points: 4.306094


Bed Height Topography:
--0.02584 +0.16891 +0.37491 +0.58691
--0.10109 +0.06466 +0.21891 +0.33941
--0.28134 --0.06084 +0.02466 +0.07941
--0.45434 --0.35959 --0.30784 --0.26684


Bed Level Correction Matrix:
0.999990 0.000000 -0.004478
0.000020 0.999990 0.004418
0.004478 -0.004418 0.999980

[paulnd]

I created one of these for myself just to help visualize things. I plugged in your probe points for this one. Thought I'd share. I've scaled things a bit proportionately so you can see things better:

https://drive.google.com/file/d/0B2R...ew?usp=sharing

Hope that helps!

Tyson

Hi Tyson,

This is a lot more meaningful to me than rows of numbers. Is it created with commercial software or something that you're able to share please? I'll be very grateful as I tried to plot the probes in Excel which is the only software which I have that can remotely produce a visual plot but I eventually gave up.

Paul

[CptanPanic]

Any plans to submit this to Marlin baseline?

[tstockgl]

Hi Tyson,

This is a lot more meaningful to me than rows of numbers. Is it created with commercial software or something that you're able to share please? I'll be very grateful as I tried to plot the probes in Excel which is the only software which I have that can remotely produce a visual plot but I eventually gave up.

Paul

Hey Paul,

I just manually mapped the probe points to a polygon plane in Maya and the text was added in Photoshop. You could probably do something similar in any CAD/3D software. It would be nice to have a program that did it automatically though!

Tyson

[Roxy]

Yup, same thing happens to me Tyson, on about 4 out of 10 G29s i send. Totally stumped and cant figure out what its related to or triggered by. Happens like this: it does 8 probes without a hitch, starts doing the last, touches once, goes back up, starts lowering slowly and as soon as probe triggers (or there about) the LCD goes blank, and the whole thing seems to freeze up. Any ideas? Other than that been loving the enhanced ABL and the fact the probe doesnt no longer scrapes the bed on retracts, thanks Roxy!!!!

I'll have to hunt down some information from another thread... But the fix for this might be known.

Glad I'm not the only one experiencing this. I'm still trying to figure it out. I have determined that I don't need to do a full power off, power on to fix it, I just need to disconnect from Pronterface and then reconnect and I can do one probe matrix without any problems.

Tyson

A number of bugs have been fixed in the main line Marlin code base. We need to get the better G29 merged into that. Probably a lot of these issues go away. Alternatively, we need to change this thread so it is making changes to the current code base.

I created one of these for myself just to help visualize things. I plugged in your probe points for this one. Thought I'd share. I've scaled things a bit proportionately so you can see things better:

https://drive.google.com/file/d/0B2R...ew?usp=sharing

Hope that helps!

Tyson

Tyson: How did you do this? Can you start a thread and document what you did? This is very nice! It would be nice if this could be automatically done by the firmware!