Auto Bed Levelling Glitches

[orionthunter]

I had a problem with my offsets not working correctly because I calibrated incorrectly the first time, and wrote them to the EPROM, then I fixed it in the firmware, but forgot to clear the EPROM. So it kept using the old value. This is probably not your issue, but I thought I would mention it just in case.

[Roxy]

I had a problem with my offsets not working correctly because I calibrated incorrectly the first time, and wrote them to the EPROM, then I fixed it in the firmware, but forgot to clear the EPROM. So it kept using the old value. This is probably not your issue, but I thought I would mention it just in case.

A lot of people are falling into this trap. And that is why there is an Invalidate EEPROM thread to give your firmware an extra command:

http://3dprintboard.com/showthread.p...lidate-Command

[MiniMadRyan]

Hmm thanks guys I will try the updated firmware tonight. I originally loaded Dacb's firmware shortly after building the printer, and honestly never checked much else into it.... Will have to get a third printer just to write and learn custom firmware on

[MiniMadRyan]

So i've flashed the updated script you included Roxy, and switched over to Pronterface/Printrun, and the problem still exists, like it was previously. I've included the output of a cold start, M114, G28, M114, G28, M114, G29, M114 below.

Here's an interesting tid-bit, as you're aware, from a cold start, with the head at a random position on the bed, the first G28 does not centre for the z probe. However, I power cycled the printer, leaving the head at 0,0 (with the end stops triggered) and the first G28 then actually returned to the middle of the bed to probe the Z...it's almost as if there's an if statement or something somewhere in the G28 that is stating "if x=0, y=0 then return to centre" but that it's only being called when the command is first initiated, and not after the triggers are actually triggered...

Hmm, here's something weird. I just ran a G28 after a power cycle, the G28 homed to 0,0. For giggles, I ran a G28 Z:0, and it probed and ran like it does normally. I then moved my x and y axis, and manually homed each one, then ran another G28 Z, and got the error that the probe was out of area...interesting!


Power on, connect and M114 before doing anything else:

Connecting...
start
Printer is now online.
echo: External Reset
Marlin1.0.0
echo: Last Updated: Nov 8 2014 17:55:19 | Author: (MakerFarm Inc, 10" i3v Prusa, dacb)
Compiled: Nov 8 2014
echo: Free Memory: 3519 PlannerBufferBytes: 1232
echo:Stored settings retrieved
echo:Steps per unit:
echo: M92 X80.00 Y80.00 Z4000.00 E859.00
echo:Maximum feedrates (mm/s):
echo: M203 X250.00 Y250.00 Z2.00 E22.00
echo:Maximum Acceleration (mm/s2):
echo: M201 X1000 Y1000 Z5 E1000
echo:Acceleration: S=acceleration, T=retract acceleration
echo: M204 S500.00 T500.00
echo:Advanced variables: S=Min feedrate (mm/s), T=Min travel feedrate (mm/s), B=minimum segment time (ms), X=maximum XY jerk (mm/s), Z=maximum Z jerk (mm/s), E=maximum E jerk (mm/s)
echo: M205 S0.00 T0.00 B20000 X20.00 Z0.40 E5.00
echo:Home offset (mm):
echo: M206 X0.00 Y0.00 Z0.00
echo:PID settings:
echo: M301 P22.20 I1.08 D114.00
echo:SD init fail
>>>M114
SENDING:M114
X:0.00 Y:0.00 Z:0.00 E:0.00 Count X: 0.00 Y:0.00 Z:0.00

Initial G28 and M114, the head is at 0,0 and the Z is probed from there as well

>>>G28
SENDING:G28
Starting G28
1 G28 home_all_axis
G28 homing X axis.
0.00 = position of X axis.
G28 homing Y axis.
0.00 = position of Y axis.
G28 doing home of Z axis towards bed.
0.00 0.00
G28 doing Z Home towards bed.
0.00 0.00
G28 done Z Home towards bed.
93.00 131.00
G28 fixing Z offset.
93.00 131.00 10.00
G28 position set to.
93.00 131.00 10.00
>>>M114
SENDING:M114
X:93.00 Y:131.00 Z:10.00 E:0.00 Count X: 93.00 Y:131.00 Z:10.00

2nd G28, Homes to 0,0 then returns to centre and probes Z

>>>G28
SENDING:G28
Starting G28
1 G28 home_all_axis
G28 homing X axis.
0.00 = position of X axis.
G28 homing Y axis.
0.00 = position of Y axis.
G28 doing home of Z axis towards bed.
0.00 0.00
G28 doing Z Home towards bed.
0.00 0.00
G28 done Z Home towards bed.
93.00 131.00
G28 fixing Z offset.
93.00 131.00 10.00
G28 position set to.
93.00 131.00 10.00
>>>M114
SENDING:M114
X:93.00 Y:131.00 Z:10.00 E:0.00 Count X: 93.00 Y:131.00 Z:10.00

G29 after second G28, all probes are done around the middle of the bed

>>>G29
SENDING:G29
Eqn coefficients: a: -0.00 b: -0.00 d: 1.33
planeNormal x: 0.00 y: 0.00 z: 1.00


Bed Level Correction Matrix:
0.999997 0.000000 -0.002513
-0.000001 1.000000 -0.000342
0.002513 0.000342 0.999997
echo:endstops hit: Z:0.94
>>>m114
SENDING:M114
X:118.00 Y:156.00 Z:4.92 E:0.00 Count X: 118.01 Y:156.00 Z:4.57

[Roxy]

I'm on the way out the door... Can you edit your post and high light which G28 is good. And which is bad? Or if you don't have both a good one and a bad one, go generate the output of a good and bad one? I'll look through them in the morning and compare the code paths that are being taken.

[MiniMadRyan]

I've updated it Roxy, I hope that makes sense and is what you're looking for, even weirder, the output of the first two G28's are identical.......

[Fri]

Roxy,
I inserted your marlin main with debug code into my fw. I acts the same way. I had to comment out the G28 matrix deletion though to simulate my issue.
What should I be looking for?

[AbuMaia]

Wait, why are you telling it to not reset the bed level matrix after a G28? If you have to do that in order to recreate your issue, then that's where your problem seems to be coming from.

[Fri]

This is how I had it setup before it started to act strange. It worked great. One abl per reset or whenever you think it is necessary. If I am uncommenting the G28 plan bed leveling, the z stop is not causing x and y to think they are home.
And yes, this is where it is coming from or is related to, without G29, all works fine.

[Roxy]

Roxy,
I inserted your marlin main with debug code into my fw. I acts the same way. I had to comment out the G28 matrix deletion though to simulate my issue.
What should I be looking for?

The initial line of the G28 to reset the bed level correction matrix to unity should be there unless you are saving the matrix to EEPROM. Are you saving (and restoring) your matrix to (and from) the EEPROM?