Grid bed compensation

dc42 · 24 Nov 2016, 23:32

David, being able to probe 441 points is one thing, I have not dug into your grid compensation code but does having that number of (different) points in the grid have an effect on the gcode->step generation speed? Does that limit the step rate or once the curve is fitted then the process required is the same?

It won't have any significant effect. The reason I chose 441 is so that you can probe a 400x400 bed with a grid spacing of 20mm.

Sniffle · 25 Nov 2016, 01:49

@dc42:

What I meant was: did you enter the M557 and G29 commands on the PanelDue keyboard, or by some other means?

The M557 command I endtered through the web interface, and then I moved to the paneldue keyboard to initiate the G29 command because I had interference concerns and wanted to witness the first row so I could make a proper adjustment should it be needed.

Jackal · 25 Nov 2016, 11:59

I am using the H offset in auto calibration, does grid levelling in the current state help to improve it?

dc42 · 25 Nov 2016, 19:17

@Jackal:

I am using the H offset in auto calibration, does grid levelling in the current state help to improve it?

No. If you need values of the H parameter that are greater than the residual height errors, then grid compensation will probably make things worse.

Russ Gries · 26 Nov 2016, 23:18

@dc42:

The autocalibration takes account of the H offsets in bed.g but the grid bed compensation doesn't.

Ah yes, i found out the hard way on this one. everything is set "longer" so i know the probe will make the bed, the everything offsets "shorter"

so getting everything just right was a bit of a trick for me, i think i got it but ran out of time last night. will try again tonight BUT,

can some one please point out the differences between theses H values from these codes. i think I'm still a bit unclear even after a lot of reading. im not sure why but i would be grateful for some one to clear this up for me.

 M558 P1 X0 Y0 Z16 H3 F100 T10000

what is the Z value here?

 G31 X0 Y0 Z20 P500

This Z value is the off set, so if i put G31 Z20 this means that when the probe is triggered it thinks the hot end is 20mm from where it was triggered right?

so why don't we use the probe off set for X,Y ? how dose it know where the probe is VS the hot end? for bed compensation this would be needed right?

Thanks in advance and Thanks for bed compensation!!!! its needed me thinks even with a .140 deviation. i think i need more ( I've been cutting PCB's with .1mm depth)

~Russ Gries

dc42 · 27 Nov 2016, 08:05

In the M558 command, Z16 is the same as Z1 and means that the Z probe is used for Z homing. H3 is the dive height, which means that when you use G30 with defined XY coordinates to probe the bed (and also G29 in the 1.17 dev build), probing starts 3mm + the probe trigger height above where the bed is expected to be.

In the G31 command, the Z parameter defines the trigger height of the Z probe.

The documentation for how gcodes behave in RRF is the reprap.org wiki at http://reprap.org/wiki/G-code. Let me know if you think particular parts need to be clarified.

HTH David

Dougal1957 · 27 Nov 2016, 15:50

These are my results bit concerned by one reading noticed whilst it was probing that the LED on the IR Board lit but it kept on moving down (did this just once towards the end of the third row)

David would appreciate it if you could cast you eyes over this and give any advice that you may have.

RepRapFirmware height map file v1
xmin,xmax,ymin,ymax,radius,spacing,xnum,ynum
-140.00,140.10,-140.00,140.10,150.00,20.00,15,15
0,0,0,0,0,-0.123,-0.137,-0.189,-0.191,-0.205,0,0,0,0,0
0,0,0,-0.051,-0.052,-0.086,-0.113,-0.152,-0.211,-0.237,-0.263,-0.285,0,0,0
0,0,0.053,0.026,-0.016,-0.051,-0.051,-0.123,-0.175,-0.203,-0.254,-0.264,-3.224,0,0
0,0.140,0.072,0.038,0.001,-0.061,-0.051,-0.039,-0.174,-0.252,-0.265,-0.235,-0.237,-0.318,0
0,0.177,0.129,0.086,0.112,-0.089,-0.062,-0.128,-0.148,-0.213,-0.202,-0.202,-0.286,-0.288,0
0.304,0.226,0.178,0.123,0.124,0.042,-0.087,-0.422,-0.153,-0.212,-0.180,-0.223,-0.213,-0.347,-0.332
0.308,0.249,0.366,0.125,0.174,0.077,-0.073,0.174,-0.091,-0.128,-0.087,-0.161,-0.172,-0.203,-0.173
0.360,0.313,0.303,0.186,0.190,0.101,0.004,0.158,-0.071,-0.111,-0.148,-0.190,-0.175,-0.154,-0.225
0.338,0.299,0.250,0.213,0.173,0.139,0.063,0.162,0.012,-0.074,-0.125,-0.185,-0.161,-0.163,-0.187
0.400,0.375,0.296,0.274,0.236,0.197,0.110,-0.001,0.026,-0.062,-0.075,-0.124,-0.142,-0.113,-0.115
0,0.234,0.180,0.214,0.216,0.191,0.127,0.086,0.061,0.005,-0.025,-0.024,-0.071,-0.061,0
0,0.438,0.389,0.312,0.313,0.236,0.125,0.127,0.102,0.061,0.000,-0.121,-0.050,-0.040,0
0,0,0.371,0.300,0.264,0.211,0.215,0.237,0.127,0.121,0.102,-0.046,-0.026,0,0
0,0,0,0.387,0.324,0.339,0.226,0.226,0.150,0.153,0.100,0.049,0,0,0
0,0,0,0,0,0.302,0.275,0.229,0.175,0.114,0,0,0,0,0

Looking at the figures it would suggest that the bed is not flat however it is 6mm thick Ecocast Tooling plate with Printbite on it Think I will have to get a straight edge on it and check it?

Doug

PS not always repeatable run again I gat the following

RepRapFirmware height map file v1
xmin,xmax,ymin,ymax,radius,spacing,xnum,ynum
-140.00,140.10,-140.00,140.10,150.00,20.00,15,15
0,0,0,0,0,-0.139,-0.188,-0.139,-0.202,-0.224,0,0,0,0,0
0,0,0,-0.058,-0.066,-0.109,-0.141,-0.129,-0.186,-0.198,-0.191,-0.176,0,0,0
0,0,0.013,-0.008,-0.053,-0.071,-0.087,-0.113,-0.162,-0.190,-0.199,-0.267,-0.237,0,0
0,0.124,0.076,0.025,-0.026,-0.054,-0.078,-0.137,-0.127,-0.165,-0.201,-0.189,-0.227,-0.226,0
0,0.198,0.120,0.047,0.089,-0.074,-0.097,-0.153,-0.188,-0.477,-0.190,-0.199,-0.237,-0.211,0
0.312,0.229,0.198,0.098,0.097,0.004,-0.089,-0.516,-0.150,-0.209,-0.197,-0.183,-0.216,-0.296,-0.250
0.287,0.263,0.292,0.100,0.190,0.015,-0.102,-0.039,-0.125,-0.149,-0.137,-0.198,-0.188,-0.220,-0.192
0.378,0.289,0.328,0.172,0.133,0.078,-0.086,0.134,-0.100,-0.150,-0.176,-0.234,-0.187,-0.199,-0.221
0.360,0.291,0.260,0.185,0.111,0.108,0.024,0.073,-0.024,-0.116,-0.187,-0.252,-0.201,-0.215,-0.187
0.447,0.397,0.336,0.276,0.180,0.164,0.073,-0.050,-0.049,-0.109,-0.151,-0.172,-0.211,-0.175,-0.161
0,0.337,0.289,0.227,0.179,0.127,0.086,0.034,-0.039,-0.060,-0.113,-0.108,-0.171,-0.153,0
0,0.478,0.397,0.374,0.270,0.141,0.085,0.074,0.037,-0.048,-0.080,-0.187,-0.126,-0.175,0
0,0,0.373,0.364,0.265,0.161,0.139,0.212,0.040,0.046,-0.008,-0.149,-0.115,0,0
0,0,0,0.346,0.295,0.273,0.148,0.136,0.084,0.024,-0.055,-0.078,0,0,0
0,0,0,0,0,0.240,0.178,0.084,0.090,0.004,0,0,0,0,0

Which to my untrained Eye looks much better Think I may have a wiring Crosstalk Issue to be honest Need to maybe re-wire the Hot End area using Screened Cable?

dc42 · 27 Nov 2016, 19:09

Doug, I think you may have a bad connection between the IR sensor output and the Duet. I can't think why else the head would continue to move down after the LED has lit. I have a long unshielded cable between the hot end and the Duet too and it doesn't cause me any problems.

Dougal1957 · 27 Nov 2016, 20:02

Thats what I am starting to think Myself (and with the PT100 as Well) will get some Shielded cable tomorrow and re-do those connections (If I have to re-do them I may as well use Shielded stuff)

Doug

Adamfilip · 27 Nov 2016, 21:26

So Do I still need to run G32? after or before G29
or

Just a G29
then A G28 to home then print?

Adamfilip · 27 Nov 2016, 21:28

Does the grid fade after a certain Z height or is it maintained all the way up?
How can you tell if the Grid compensation is active?

Adamfilip · 27 Nov 2016, 21:30

Here is my first grid result. had a error near the end when it probed on top of a clip holding down the glass

RepRapFirmware height map file v1 generated at 0116-10-27 16:15
xmin,xmax,ymin,ymax,radius,spacing,xnum,ynum
20.00,280.00,20.00,280.00,-1.00,28.00,10,10
0.002,0.080,0.121,0.163,0.171,0.142,0.100,-0.004,-0.135,-0.278
0.026,0.113,0.197,0.275,0.271,0.262,0.221,0.118,0.019,-0.161
0.039,0.141,0.225,0.283,0.314,0.280,0.257,0.180,0.103,-0.061
0.019,0.165,0.236,0.257,0.284,0.306,0.275,0.216,0.114,-0.016
0.019,0.139,0.220,0.277,0.295,0.296,0.276,0.235,0.134,0.001
-0.019,0.081,0.179,0.224,0.264,0.261,0.279,0.216,0.141,0.016
-0.095,0.031,0.105,0.179,0.223,0.243,0.216,0.194,0.120,0.015
-0.214,-0.039,0.027,0.098,0.173,0.197,0.183,0.119,0.065,-0.010
-0.279,-0.178,-0.094,0.027,0.070,0.134,0.125,0.046,0.005,-0.104
-0.395,-0.299,-0.212,-0.136,-0.054,-0.034,-1.277,-1.320,-1.651,-1.608

Adamfilip · 27 Nov 2016, 22:31

You mentioned that you cant load a grid yet.
so if i restart the Duet Wifi G29 S1 wont load current csv?
G29 will just recreate a new mesh for now.

dc42 · 28 Nov 2016, 08:04

@Adamfilip:

You mentioned that you cant load a grid yet.
so if i restart the Duet Wifi G29 S1 wont load current csv?
G29 will just recreate a new mesh for now.

That's correct. I should get time to implement loading the height map from file later this week.

jmjcoke2 · 28 Nov 2016, 16:23

Works like a charm, had to change dive height to 5mm.

Adamfilip · 28 Nov 2016, 18:14

Where do you change this?

dc42 · 28 Nov 2016, 21:33

@Adamfilip:

Where do you change this?

If you mean the dive height, see http://reprap.org/wiki/G-code#M558:_Set_Z_probe_type.

ronnien59 · 30 Nov 2016, 00:15

I am trying to run the grid bed compensation, however when i run the G29 I get:

7:10:20 PM G29 Error: No valid grid defined for G29 bed probing
7:04:32 PM G29 Error: No valid grid defined for G29 bed probing

I have placed M557 X10:250 Y40:260 S40 in my config.g.
Would like to get this working but do not know enough about it

ronnien59 · 30 Nov 2016, 02:47

@ronnien59:

I am trying to run the grid bed compensation, however when i run the G29 I get:

7:10:20 PM G29 Error: No valid grid defined for G29 bed probing
7:04:32 PM G29 Error: No valid grid defined for G29 bed probing

I have placed M557 X10:250 Y40:260 S40 in my config.g.
Would like to get this working but do not know enough about it

I was able to run it but had to use the PanelDue to run it it would not run over the wifi??

Adamfilip · 30 Nov 2016, 04:53

I had no issue running over wifi