Bed Auto Leveling.. check this out

Hello,

I'm having a bit of a problem with my autolevel. I have it wired correctly and everything is set up well in Marlin. However, in an attempt to gain more accuracy, I have enabled Probe Servo Dectivation Delay 300ms to keep my servo from twitching as it is taking measurements. When I do this, the hot end no longer lowers to the correct height. When I disable the feature, the hotened lowers, but since servo is twitching I get inconsistent measurements. Has anyone else had this problem? Thanks.

Are you using the latest version of marlin? There have been a lot of changes and bugs cropping up for a while now and this could be one of them. If you are it might be worth searching git to see if anyone else is experiencing it or open a new issue.

dc42

A simple solution to this would be to get some aluminium or copper tape and put little squares on the sheets of glass where the reads are taken.

then just a matter to work out the thickness of the tape and subtract from z prob offset.

Or keep an eye out for my next post which will be a completely different way to setup a probe

Edited 1 time(s). Last edit at 05/04/2015 05:54AM by SimCity.

HI Alex,

As i'm facing an issue, as I having an Prusa I3 from [www.3dprintersonlinestore.com] and it's terrible, as the system have the IR Sensor but the system don't have auto Leveling and can i know where can i get the auto leveling system,?

Hope you can help.

I have the latest version of Marlin. Built my P3Steel a month ago. Using Repetier Host and Slic3r.

Set up auto bed levelling and all works up to the point where the print starts. The hotend or z axis does not come down to the bed. It stays in the air where it stopped after G29 and starts to print. Then just keeps on printing in mid air.

I am attaching my config.h file for refference. Do not no where i missed something.

Riaan

Quote
RiaanLR
Hope you can help.

I have the latest version of Marlin. Built my P3Steel a month ago. Using Repetier Host and Slic3r.

Did you download the development branch of marlin? There is all sorts of bugs with the bed levelling at the moment. I think there is a more stable branch or download a previous version. There are discussions about it on git.

I am not sure. I think it is the stable one. Marlin 1.0.2

I will double check this.

Riaan

Here is a link to you tube of what is happening.

[www.youtube.com]

Riaan

Here is my g code from slic3r. Look at G1. It tells the hotend to start printing at .3 mm. Is the correct or should the value be .3 mm plus the offset value of the probe. Then it should be in the minus something.

; generated by Slic3r 1.1.7 on 2015-05-09 at 01:11:02

; perimeters extrusion width = 0.40mm
; infill extrusion width = 0.67mm
; solid infill extrusion width = 0.26mm
; top infill extrusion width = 0.26mm

G21 ; set units to millimeters
M107
M190 S100 ; wait for bed temperature to be reached
M104 S230 ; set temperature
G28 ; home all axis
G29
M106 S255 ; start fan speed to S
M109 S230 ; wait for temperature to be reached
G90 ; use absolute coordinates
G92 E0
M82 ; use absolute distances for extrusion
G1 Z0.300 F6000.000

Riaan

I have sorted my problem. The best way to learn in live is via trial and error.

In Marlin I had to change one of my speed settings from 5 to 3. The z axis could not get down fast enough and got stuck.

#define DEFAULT_MAX_FEEDRATE {500, 500, 3, 25} // (mm/sec)

Riaan

Here's a problem I can't seem to figure out or correct.

Machine is home built using latest Marlin on arduino and poluno. I use auto bed leveling on glass probing 9 points on the glass.

It does it's thing on startup, but then when the print starts the right side is always low and the left is always high. While it's printing the skirt I have to raise the right side z by a couple of clicks on the stepper and then lower the left side to get things right.

If I run another print, without reseting anything, I have to do the same again. After about 3 or 4 parts the right side is much higher starting out than the left and I have to re-level the x axis manually.

Really annoying. Any thoughts?? Need some help!!

Thanks a lot.

Bob Keyes

---

Not such a NEWBEE anymore
Thanks,

Bob Keyes

I'm trying to set up bed leveling as per Thomas Sanladerers video (https://youtu.be/EcGFLwj0pnA), using a LJ12A3-4-Z/BY sensor.
I've soldered the cables to the proper resistors, however, I can't get readings that make any sense from it.

I get readings fluctuating between 6v and 11v no matter if the sensor is opened or closed.
When I only connect power and ground without any resistors, I get 12v/0.5v, so it seems like the sensor is working.
Does anyone have any idea what could be wrong?

It's a bit hard to see everything in the picture, but that's what the connections looks like.

Quote
Akegata
I'm trying to set up bed leveling as per Thomas Sanladerers video (https://youtu.be/EcGFLwj0pnA), using a LJ12A3-4-Z/BY sensor.
I've soldered the cables to the proper resistors, however, I can't get readings that make any sense from it.

I get readings fluctuating between 6v and 11v no matter if the sensor is opened or closed.
When I only connect power and ground without any resistors, I get 12v/0.5v, so it seems like the sensor is working.
Does anyone have any idea what could be wrong?

It's a bit hard to see everything in the picture, but that's what the connections looks like.

Never mind this. I picked up an LM7805. Works perfectly.

I also commented out raise after probing along with

G1 X0 Y0 F4000
G1 Z1

this got me printing perfect on the bed, if I enable raise after probing, it print at the height you defined in the firmware after probing, I can't yet find another way around it

Not sure whether anybody can help, but I'm having trouble getting auto-levelling to work on a Taz 5 (Rambo 1.3) using a capacitive sensor.

Using Marlin v1.0.2 and although G28 followed by G29 seem to be doing the right things, the actual print doesn't seem to be taking it in to account. Here's my start G-Code:

G91 ; switch to relative positioning
G1 Z10 ; safe raise of z axis to ensure probe doesn't hit bed clamp
G90 ; switch back to absolute positioning
G28 ; home all axes
G29 ; level print bed

I'm checking 9 points and the Z offsets looks sensible (ranging from -0.09 to 0.62), and they are repeatable if I run G28 followed by G29 multiple times (+- 0.01).

I'm currently just trying to print a calibration piece (see attached) but this is where the problem starts, as the left hand side of the print is too close, and is resulting in a smear of plastic, the right hand side of the print is a millimetre or more too high so the filament is just being dropped on.

Anybody had any similar issues? I have tried a couple of different firmware versions in case it's a bug in the Z offset map, but they all seem to produce the exact same results. Can anybody recommend a firmware they know works?

Whilst printing, I can see the Z-Axis slowly rotating back-and-forth, so it obviously thinks it's doing what it should.

Any help or suggestions gratefully received. I've attached my configuration.h in case it's any use.

Thanks!!

Have you checked the latest bugs on GitHub? There has been an issue with bed levelling since last year when the developers started trying to make a stable and cleaner version. I'm still using a version from a couple of years ago that has the bed levelling before the bugs started. I've been told to stick with it until they confirm the latest version is bug free. I haven't been able to keep up with all the problems and changes they've made but thinkyhead always responds to issues and can probably help you bug fix.

Thanks for reply Brainscan.

I assume your auto-levelling works a treat then? Which version of Marlin are you using then? Any link to the GitHub download fork you used...there seems to be hundreds of versions now. I don't mind if it's older...I just want something stable with auto-bed levelling working.

I'm pretty sure the hardware side of what I've done is working OK, so it's either firmware or simply configuration.

Edited 1 time(s). Last edit at 08/11/2015 07:34AM by Astek777.

I'm not even sure it has a version number (sorry I'm at work so can't check until later) but it's an old modified version given to me by Roxy at 3dprintboard forum. I think the files are still available on the forum. I think if you search google for G29 advanced it should still pop up. If you don't have any luck with that let me know and I'll share my copy and you can just alter the config files to suit your machine.

Once it's set correctly it does work beautifully though, I get consistent first layers every time. Before that I was lucky if half of my prints worked first time. I use a capacitance sensor too and found it much more reliable than servos.

Thanks for the advice. I managed to find the post on 3dprintboard, but the two links were dead. Reading through the posts (65 pages now) there was a link to a version of Marlin with the updates discussed already installed.

I updated the config and installed. The advanced G29 is much better with the feedback it can provide.

I also managed to make a quick change to Marlin_main.cpp to raise the Z axis when 'G28' is called to clear the print bed brackets before homing x and y. If anyone is interested I added:


if(home_all_axis)
{
do_blocking_move_relative(0, 0, Z_RAISE_BEFORE_HOMING);
destination[Z_AXIS] += Z_RAISE_BEFORE_HOMING;
}


...directly after the line...


home_all_axis = !((code_seen(axis_codes[X_AXIS])) || (code_seen(axis_codes[Y_AXIS])) || (code_seen(axis_codes[Z_AXIS])));


However, the new firmware doesn't seem to have made any different to my print quality.

My bed was pretty level before, maybe a mm out across the whole bed, which is why I thought it wasn't working at all. I've tried to make my print bed particularly wonky now, so I'm probably 3mm out in one corner just to see what would happen. It does appear the z distortion map is working to a degree, it's just not as perfect as I would hope considering the readings from the capacitive sensor are so accurate and repeatable. I'm basically getting the same results regardless of how off my print bed is. The extruder is roughly following the level of the print bed, but it seems to be high and low at various places. Maybe I'm just expecting too much?

I'm also getting a strange clicking (every couple of seconds) and slight shuddering on the two z-axis steppers now as they slowly try and adjust.

Hi all, I have built a gMax 1.5 Printer, it is using RAMPS 1.4 & Marlin and it has autobed level setup on it, but since installing the upgrade, not once has it provided a perfect first layer. The adjustments are measured using a 3 x 3 grid, when probing the results seem to be accurate to the actual print bed. However when the printer starts printing the first layer, EVER time, the first layer is no where near even, i always have 1 side too far from the bed, and the other too close. It's like the autolevel is compensating the wrong way around, where it should be going closer to the bed it gets further, and vice versa. I have scoured the net and come across a few similar instances of this, but there has never really ever been a fix for it, other than making sure the bed is as close to level physically as possible; which kind of defeats the purpose of autoleveling... im using 4mm toughened glass as my print surface, it is completely flat with no uneven spots.

Any advice is greatly appreciated.

Thanks

Alex

Edited 1 time(s). Last edit at 08/17/2015 09:19AM by xelad.

Hi Alex

Welcome to my world! :-)

Just out of interest, which corners are high and which are low? Currently my front left (nearest X0Y0) is too low and the right front is too far away. My readings are also spot on and very repeatable.

Depending upon which version of Marlin you have installed you can get a topography map using the G29 command, but you need to have 'Advanced G29' installed, or a newer version of Marlin.

Try 'G28', followed by 'G29 V4 T n5'. This should scan a 5x5 square and give you the offsets that the 'auto levelling' uses to offset it's measurements. You can repeat this a few times (make sure you start with 'G28' each time to reset the previous map), and see how consistent your results are.

Are you using a capacitive sensor or other method?

Just so you know my printer is still off. Just about to try a simply 3 point plane and then start taking some of my X and Y construction apart to see if I can find anything that is warped.

As far as I was aware, and from what I've seen with some purposely wonky print beds, the auto bed levelling, if working correct and having scanned enough points, should be able to compensate for this, but I'm not getting it...yet.

Cheers

Sean.

Edited 1 time(s). Last edit at 08/17/2015 10:04AM by Astek777.

Hi Sean,

Thanks for the feedback, when i get back from work i will gather the data you have mentioned and report back. I am also unsure of what version of marlin i am using, to be sure i am going to download the latest version, apply my settings and install onto the printer tonight.

I will update as i progress.

Thanks

Alex

There is no real 'latest version' of Marlin, just lots of development at the moment. There are also a lot of known issues with G29 auto-levelling in the current development forks, but they are slowly getting sorted.

If you go here:
[3dprintboard.com]

There's a lot of information about auto levelling. On page 48(ish) there's a download of Marlin which has all the Advanced G29 stuff already installed. It seems to be a quite well tested version and Roxy, the person who developed it, provides some fantastic support, as you'll see from my recent posts on there.

My issues haven't been fixed yet, but your problems aren't necessarily the same as mine, even though they sound similar, but it might be useful to rule out the software side of things.

Also, ensure your 'probe offsets' are correct. Obviously the position of your probe isn't where your extruder actually is, so these need to be correct.

Good luck!

Well i have tried Roxy's version of the firmware, and although maybe slightly better, the problem still persists, also dont seem to be able to specify how many grid points to probe, or the Z offset from extruder, no matter what i change in the firmware, the values remain the same.

I am using a mechanical endstop, the probe provides consistent readings. Running out of options to try, i can remember the last time i had a successful print any bigger than 1" square.

Alex

EDIT: I have since read that post fully and have also responded in there.

Edited 1 time(s). Last edit at 08/18/2015 09:08AM by xelad.

Have you got eeprom enabled? If so you probably haven't been reloading the settings after you change them in the firmware. It's a common thing people forget to do when setting up ABL. It's been so long since I've had to change anything I can't remember the exact M codes but it's well documented, something like M502 then M500. That seems the most likely thing if you say changing the values has no effect

Edited 1 time(s). Last edit at 08/18/2015 10:11AM by Brainscan.

I have searched the web and this forum for hours trying to figure out if simple 3 point "leveling" is still supported in any version of Marlin. I want to update my Taz 4 Cartesian printer with a simple and basic auto level using 3 points on a large triangle. I have FSRs mounted to support my build plate - 2 at each corner for the square plate and the 3rd at the middle of the far edge. I only need to probe these 3 points, my bed is as flat as I need it and the mechanics of this printer are very good with no X or Y sag (running a Bowden so there is very little moving mass on the carriage.)

This thread started off with exactly what I'd like to do but I've found references that this was taken out of Marlin, that it's still in Marlin, etc. Does anyone have a definitive answer on whether 3 point is supported in any branch or recent version? If not recent, what is the best working earlier version?

thanks!
Michael

Hi Michael

I think most of the newer Marlin versions still feature it, the one I was using (1.0.2) had it, but it's not the default method, so I did need to uncomment/comment a few lines to get it working in configuration.h.

I didn't have much luck with it, but my problems were caused by un-level Y rails which was causing a 'twist' effect and stopping the levelling algorithm from working.

I did also try manually bed levelling which worked better, even with my un-level setup, and if you have EEPROM enabled you can save the probing back to memory so it doesn't need to be performed every time.

I did spend a few hours getting everything straight and level to quite a high accuracy and I'm using auto bed levelling with a 3x3 grid and getting consistent prints every time.

Cheers

Sean.

Astek777, thank you for the response. On further investigation the implementation that did (and may still) exist is not exactly what I'm looking for. I was expecting to be able to specify the X and Y coordinates for each of the three points explicitly. This way, the points could be located over the bed suspension points in a 3 point attachment system or over the FSRs. It appears that the points are derived from 4 numbers specified in the config file that are combined into coordinates to describe a right triangle rather than the preferred (at least to me) isosceles triangle.

I'll try out the 3x3 grid and be done with it if it works.

thanks,
Michael

Hello
I have build a Delta kossel and nearly everything is working.
I have problems with the autoleveling when i run a G30 it works fine, but when I use the G29 it al goes to ****
here is the code I get from my 27 point (know it is a lot )
Bed X: 28.000 Y: -74.000 Z: 9.995
Bed X: 22.000 Y: -74.000 Z: 19.985
Bed X: 16.000 Y: -74.000 Z: 29.975
Bed X: 10.000 Y: -74.000 Z: 39.965
Bed X: 4.000 Y: -74.000 Z: 49.955
Bed X: -2.000 Y: -74.000 Z: 59.945
Bed X: -8.000 Y: -74.000 Z: 69.940
Bed X: -14.000 Y: -74.000 Z: 79.935
Bed X: -20.000 Y: -74.000 Z: 89.930
Bed X: -26.000 Y: -74.000 Z: 99.925
Bed X: -38.000 Y: -68.000 Z: 109.915
Bed X: -32.000 Y: -68.000 Z: 119.905
Bed X: -26.000 Y: -68.000 Z: 129.895
Bed X: -20.000 Y: -68.000 Z: 139.885
Bed X: -14.000 Y: -68.000 Z: 149.875
Bed X: -8.000 Y: -68.000 Z: 159.865
Bed X: -2.000 Y: -68.000 Z: 169.855
Bed X: 4.000 Y: -68.000 Z: 179.850
Bed X: 10.000 Y: -68.000 Z: 189.845
Bed X: 16.000 Y: -68.000 Z: 199.840
Bed X: 22.000 Y: -68.000 Z: 209.835
Bed X: 28.000 Y: -68.000 Z: 219.830
Bed X: 34.000 Y: -68.000 Z: 229.825
Bed X: 40.000 Y: -68.000 Z: 239.820
Bed X: 46.000 Y: -62.000 Z: 249.810
Bed X: 40.000 Y: -62.000 Z: 259.800
Bed X: 34.000 Y: -62.000 Z: 269.790
Bed X: 28.000 Y: -62.000 Z: 279.780
Bed X: 22.000 Y: -62.000 Z: 289.770
Bed X: 16.000 Y: -62.000 Z: 299.760
Bed X: 10.000 Y: -62.000 Z: 309.750
Bed X: 4.000 Y: -62.000 Z: 319.740
Bed X: -2.000 Y: -62.000 Z: 329.730
Bed X: -8.000 Y: -62.000 Z: 339.725
Bed X: -14.000 Y: -62.000 Z: 349.720
Bed X: -20.000 Y: -62.000 Z: 359.715
Bed X: -26.000 Y: -62.000 Z: 369.710
Bed X: -32.000 Y: -62.000 Z: 379.705

I also attach my Configuration and my marlin main
I really hope you can help me.

You don't need that many check points. If your print surface is that uneven, replace it. It's most likely running out of memory. Does it have the problem with three?