G29S1 not loading during print and G32 not leveling properly

moe.adams9093

that's what I have on my gcode file

g28;
g32;
g29s1;

engikeneer

G29S1

OR

G29 S1

I think you need the space in between

If you're still having trouble, try manually sending each line from the start of a print one at a time and checking if the mesh compensation is applied/disabled via DWC

fcwilt

@jay_s_uk said in G29S1 not loading during print and G32 not leveling properly:

Heightmaps get unloaded whenever G28 is ran.

Why do you think that?

I have not found that to be the general case.

Frederick

fcwilt

Hi,

Your current bed.g has a single G30 before the four G30 Pn commands.

The G30 should occur AFTER the four G30 Pn commands.

And since bed.g is part of the bed leveling process normally you don't load your height map in it.

M561 ; clear any bed transform
M671 X50.0:450.0:450.0:50.0 Y50.0:50.0:450.0:450.0 S2   	; ZBelts Near Locations  
G30;  
G30 P0 X50 Y50 Z-99999;
G30 P1 X450 Y50 Z-99999; 
G30 P2 X450 Y450 Z-99999; probe near a leadscrew, half way along Y axis
G30 P3 X50 Y450 Z-99999 S4; probe near a leadscrew and calibrate 2 motors
G29 S1;

moe.adams9093

I'm still having an issue where when I send G32 my 4 corners are not compensated correctly it only moves the front right side, does the have to do because 3 of my Z steppers are connected to the duex 5 and the 4th one is connected to the mainboard.

Veti

post the output of the G32 command

moe.adams9093

here I ran it 3 times

3/9/2021, 11:54:19 PM: Connected to 192.168.1.252
3/9/2021, 11:55:57 PM: Upload of bed.g successful after 0s
3/9/2021, 11:57:50 PM: G32: Leadscrew adjustments made: 0.621 1.020 1.241 0.561, points used 4, (mean, deviation) before (0.861, 0.282) after (-0.000, 0.000)
3/9/2021, 11:58:47 PM: G32: Leadscrew adjustments made: -0.043 -0.128 0.225 -0.002, points used 4, (mean, deviation) before (0.013, 0.131) after (0.000, 0.000)
3/10/2021, 12:09:35 AM: G32: Leadscrew adjustments made: -0.152 -0.343 0.219 0.043, points used 4, (mean, deviation) before (-0.058, 0.210) after (-0.000, 0.000)

moe.adams9093

and here's my updated bed.g file

; bed.g
; called to perform automatic bed compensation via G32
;
; generated by RepRapFirmware Configuration Tool v2.1.8 on Thu Apr 30 2020 23:57:44 GMT-0500 (Central Daylight Time)
M561 ; clear any bed transform
M671 X50.0:450.0:450.0:50.0 Y50.0:50.0:450.0:450.0 S2   	; ZBelts Near Locations  
G30 P0 X50 Y50 Z-99999;
G30 P1 X450 Y50 Z-99999; 
G30 P2 X450 Y450 Z-99999; probe near a leadscrew, half way along Y axis
G30 P3 X50 Y450 Z-99999 S4; probe near a leadscrew and calibrate 2 motors
G30;

moe.adams9093

so I did some modification to my M558 code I added new parameters to see if I can get some more accuracy

NEW CODE:

; Z-Probe
M558 P1 C"zprobe.in" H5 R.02 A11 S2 F120 T6000                   	; set Z probe type to unmodulated and the dive height + speeds
G31 P500 X5 Y29 Z0.77                                 	; set Z probe trigger value, offset and trigger height
M557 X50:450 Y50:450 S40                                	; define mesh grid

and here's a screenshot of my new heightmap

Screenshot 2021-03-10 035751.png

3/10/2021, 3:47:24 AM	G29
121 points probed, min error -0.071, max error 1.820, mean 0.505, deviation 0.402
Height map saved to file 0:/sys/heightmap.csv

RepRapFirmware height map file v2 generated at 2021-03-10 03:47, min error -0.071, max error 1.820, mean 0.505, deviation 0.402
xmin,xmax,ymin,ymax,radius,xspacing,yspacing,xnum,ynum
50.00,450.00,50.00,450.00,-1.00,40.00,40.00,11,11
  0.609,  0.512,  0.468,  0.510,  0.401,  0.363,  0.488,  0.612,  0.703,  0.933,  1.184
  0.446,  0.322,  0.258,  0.262,  0.228,  0.032,  0.397,  0.529,  0.573,  0.867,  1.015
  0.361,  0.173,  0.143,  0.090,  0.066, -0.071,  0.336,  0.439,  0.575,  0.769,  1.029
  0.306,  0.158,  0.027,  0.071,  0.042,  0.043,  0.248,  0.152,  0.499,  0.752,  1.046
  0.215,  0.066,  0.037,  0.036, -0.064,  0.029,  0.232,  0.340,  0.570,  0.796,  1.032
  0.188,  0.077,  0.005,  0.048,  0.101,  0.139,  0.313,  0.465,  0.567,  0.813,  1.082
  0.252,  0.082,  0.097,  0.082,  0.022,  0.163,  0.350,  0.462,  0.618,  0.874,  1.157
  0.343,  0.140,  0.152,  0.086,  0.141,  0.269,  0.408,  0.516,  0.711,  1.016,  1.309
  0.384,  0.291,  0.229,  0.293,  0.258,  0.355,  0.559,  0.674,  0.884,  1.150,  1.468
  0.468,  0.446,  0.411,  0.413,  0.439,  0.492,  0.675,  0.885,  1.018,  1.376,  1.611
  0.842,  0.669,  0.605,  0.626,  0.705,  0.745,  0.965,  1.118,  1.340,  1.582,  1.820

I'm not sure how exactly M558 A & S Parameters should work on the Gcode Dictionary it doesn't explain a whole lot if somebody can share how these 2 parameters can make significant changes. Thanks in advance.

fcwilt

@moe-adams9093 said in [G29S1 not loading during print and G32 not leveling properly]

I'm not sure how exactly M558 A & S Parameters should work on the Gcode Dictionary it doesn't explain a whole lot if somebody can share how these 2 parameters can make significant changes. Thanks in advance.

Perhaps an example will help.

With A=10 and S=0.01 the firmware will probe each point up to 10 times trying to get two consecutive readings that are within 0.01.

If after 10 times two such readings do not occur then all 10 readings will be averaged.

I use those values for A and S and they work well.

Frederick

moe.adams9093

Thank you for explaining it. I do have some questions changing the P500 parameter to something else may be like 525 will it help to get it lot more accurate. As you see my map it’s not leveled at all

fcwilt

@moe-adams9093 said in G29S1 not loading during print and G32 not leveling properly:

Thank you for explaining it. I do have some questions changing the P500 parameter to something else may be like 525 will it help to get it lot more accurate. As you see my map it’s not leveled at all

Changing the P parameter is not likely to make much difference.

It could be that your bed is just not very flat.

What is it made of?

Frederick

moe.adams9093

Aluminum 6 mm thick I know the bed is not 100% I would say about 70% I have some twist In some spots

fcwilt

@moe-adams9093 said in G29S1 not loading during print and G32 not leveling properly:

Aluminum 6 mm thick I know the bed is not 100% I would say about 70% I have some twist In some spots

Is it cast aluminum tool plate?

Frederick

moe.adams9093

No it’s aluminum 6061

fcwilt

@moe-adams9093 said in G29S1 not loading during print and G32 not leveling properly:

No it’s aluminum 6061

If you have the resources you may want to upgrade.

This is from one of my suppliers:

Cast Aluminum Plates have a precision machined surface for superior flatness, flat within .015'', and +/- .005'' thickness tolerance,

Frederick

Phaedrux

If you have the means you might be able to bash it into flatness, or at least take out the worst of the peaks and valleys.

fcwilt

@Phaedrux said in G29S1 not loading during print and G32 not leveling properly:

bash it into flatness

I think machinists might cringe at that.

Frederick

Phaedrux

@fcwilt At least I didn't suggest parking a car on it. ..Yet

fcwilt

@Phaedrux said in G29S1 not loading during print and G32 not leveling properly:

@fcwilt At least I didn't suggest parking a car on it. ..Yet

I hadn't thought about that approach - good idea.

Maybe something with treads - like a tank.

Frederick