Auto Level not compensate enough

[idzuwan]

Dea all,

On my mendelmax it seem the Auto Level not compensate enough when printing first layer I can always hear my extruder making clicking sound which indicate the nozzle is blocked and it can't extrude plastic on certain part of the bed, It was working fine few days ago now after changed a new batch of filament it started to have this problem.

Anyone know I can I debug this issues

[nicolas-rambaud]

Try to increase temperature, it happened to me also when i took filament from another reseller.

If the compensation worked before it should still work.

I also make the assumption that the cause is the firmware but it’s often due to something else we do.

Also Do you preheat your bed and hotend before starting your print? Because you need to make sure all the part have expended before doing level compensation or your hot end will be more into your head bed due to bed/glass/hotend expension.

De : idzuwan [mailto:notifications@github.com] Envoyé : jeudi 19 juin 2014 09:38 À : ErikZalm/Marlin Objet : [Marlin] Auto Level not compensate enough (#972)

Dea all,

On my mendelmax it seem the Auto Level not compensate enough when printing first layer I can always hear my extruder making clicking sound which indicate the nozzle is blocked and it can't extrude plastic on certain part of the bed, It was working fine few days ago now after changed a new batch of filament it started to have this problem.

Anyone know I can I debug this issues

— Reply to this email directly or view it on GitHub https://github.com/ErikZalm/Marlin/issues/972.[image: Image supprimée par l'expéditeur.]

[idzuwan]

Hi,

I'm currently printing with 220c with this PLA filament normally I only need 190c I'm now almost at ABS temperture :-/ at 220 I still get random blockages maybe this PLA spool can be trashed

[nicolas-rambaud]

You can also have blocage if the 0.1 tolerence on your filament is not respected, it can be stuck inside the head during print.

De : idzuwan [mailto:notifications@github.com] Envoyé : jeudi 19 juin 2014 16:41 À : ErikZalm/Marlin Cc : Nicolas Rambaud Objet : Re: [Marlin] Auto Level not compensate enough (#972)

Hi,

I'm currently printing with 220c with this PLA filament normally I only need 190c I'm now almost at ABS temperture :-/ at 220 I still get random blockages maybe this PLA spool can be trashed

— Reply to this email directly or view it on GitHub https://github.com/ErikZalm/Marlin/issues/972#issuecomment-46568695.[image: Image supprimée par l'expéditeur.]

[idzuwan]

So meaning that my filament is not actually 1.75mm in some part? Will need to check that, with previous batch I was printing perfectly fine now my calibration all gone, stringing are bad I will check this again later on, I'm printing with 220 now seem less blockages but printing quality are bad

[idzuwan]

I having a second thought about the filament diameter as the thing is the blockages awalys happen on my left side of the bed on the right side it was perfectly fine, when on the left side I can clear hear the sound of the nozzle dragging on the plastic surface when filling up layer but not on the right side

[nophead]

Perhaps your bed is not flat or running on Y bars that are not parallel, neither or which are compensated for in Marlin. It only compensates for a flat bed inclined relative to the Z plane.

[idzuwan]

I will check that! thanks!

[idzuwan]

Ok, I change a new filament no more jamming but my title still apply, I check my bed seem pretty leveled of course my front bed are a bit high then the back but it is flat and parallel, I noticed on the last 3 print the object on my right bed will always have problem with 2 layer, which after the first layer for some reason the edge does not have enough plastic, can see the gap between 1st and 3rd layer and seem fine after the 3rd layer, this also happen on the center object and the weird thing is the last object on the most left part of the bed get full solid layer with no gap.

[nophead]

Maybe backlash in your Z axis.

[nicolas-rambaud]

How many probing points are you taking ?

I never checked whether there is or not a difference between probing 9 or 4 points but at least there is 1 difference between 3 and 4 points.

On my side I’m always probing for 9 points as it seems more reliable to possible heatbed flatness issues.

Also, are the probing location well on the edge of your hotbed?

I remember having an issue when my probe points were too close from the homing end stops. It created me an offset on the center on my heatbed.

I don’t know if it’s solved but basically I try to take my first probing point 10 mm from the end stop to avoid that issue.

I know that I might not target your specific behavior but maybe it can help somehow.

About your 2nd layer not “having enough plastic”, I had similar issues on my slicer side. Somehow the geometry of my model was wrong, but it seemed fine when viewing the 3D model.

When looking at the gcode simulation (gcode viewer or something similar) I could see that the slicer was missing part of my geometry.

I corrected the model and everything was fine afterward.

If it happens exactly the same way on both of your objects it might correspond to this slicing issue.

Try to double check the generated gcode.

I hope this helps

De : Idzuwan [mailto:notifications@github.com] Envoyé : lundi 23 juin 2014 06:42 À : ErikZalm/Marlin Cc : Nicolas Rambaud Objet : Re: [Marlin] Auto Level not compensate enough (#972)

Ok, I change a new filament no more jamming but my title still apply, I check my bed seem pretty leveled of course my front bed are a bit high then the back but it is flat and parallel, I noticed on the last 3 print the object on my right bed will always have problem with 2 layer, which after the first layer for some reason the edge does not have enough plastic, can see the gap between 1st and 3rd layer and seem fine after the 3rd layer, this also happen on the center object and the weird thing is the last object on the most left part of the bed get full solid layer with no gap.

[image: Image supprimée par l'expéditeur. 20140623_123105] https://cloud.githubusercontent.com/assets/7881146/3353905/9a373128-fa90-11e3-803d-b112ab5d478c.jpg

— Reply to this email directly or view it on GitHub https://github.com/ErikZalm/Marlin/issues/972#issuecomment-46805846.[image: Image supprimée par l'expéditeur.]

[idzuwan]

Hi,

I'm using 4 probe point, my 4 point are on each corner of the bed, my Z home will be probed at the center of the bed, I will making it probe 9 point but not sure it will not crash my Y axis :), I sure model are find as the object is a pair of 2 so I printing 3 pair, the first 2 pair having the second layer issues (on the right side and center) while the left side pair printed perfectly fine without any flaws

Nophed, I tested for backslash does not seem to be the case, and does not explain it reproduce at the same layer every time when printing :).

I will try to re-slice the model or try do to a simple model on skecthup for test

update 1 :

I did a test print place 6 object on the bed, what I found is interesting, the left side was I thought perfect is actually not, I watched the print until it printing second layer, on the left side it actually plowing to the plastic to the extent blocking the nozzle, while on the right side first layer I can see plastic was curling on the bed indicating the nozzle was to high on initial layer and 2nd which make it extrude plastic on the air on layer two, not sure why does the firmware did not actually compensate the unevenness (well it actually it did but not enough)

to make thing even interesting, I re-flashed my firmware and disable the auto-leveling and level my bed normally and it printing as normal no problem at all, I will try to re-enable it back and make it probe with 9 point this time around see how it goes

Update 2 :

Enabled 9 point, I'm still having the same issues left is high (nozzle blockages) right is low (curling plastic, today my plan is to re-do my Y-Carriages and level everything again :-/, and re-do the Z-axis offset again

[idzuwan]

this problem getting weirder as it progress, I re-did my Y-Carriages and level everything, I even put back the spring to level my bed manually and once it leveled, I did test pring again with heat bed off to eliminated glass/pcb warp, it still doing the same thing, and a perfectly even bed it still lifting up the right side to high (curling) and flowing the the bed (to low) on the left side to the extent I have to do the bed adjustment during printing! which beat the purpose of having autoleveling in the firmware, I'm lost no idea why the firmware is doing this, now I'm printing with auto leveling turned of in the firmware until someone give me some idea how to actually debug this

[nicolas-rambaud]

What kind of coupler do you have on the Z axis ?

If one of your coupler is too loose on one side it can happen that one side of your z axis is lifted higher that the other side.

I have a Makerfarm printer and my couplers were just plastic tubes.

It was working fine for a long time now, but recently I had my left tube getting loose. Even by adding ties it was still slipping.

I had to print a proper plastic coupler to fix my issue.

Maybe what you see is the same?

De : Idzuwan [mailto:notifications@github.com] Envoyé : mercredi 25 juin 2014 16:58 À : ErikZalm/Marlin Cc : Nicolas Rambaud Objet : Re: [Marlin] Auto Level not compensate enough (#972)

this problem getting weirder as it progress, I re-did my Y-Carriages and level everything, I even put back the spring to level my bed manually and once it leveled, I did test pring again with heat bed off to eliminated glass/pcb warp, it still doing the same thing, and a perfectly even bed it still lifting up the right side to high (curling) and flowing the the bed (to low) on the left side to the extent I have to do the bed adjustment during printing!, I'm lost no idea why the firmware is doing this, now I'm printing with auto leveling turned of in the firmware until someone give me some idea how to actually debug this

— Reply to this email directly or view it on GitHub https://github.com/ErikZalm/Marlin/issues/972#issuecomment-47112262.[image: Image supprimée par l'expéditeur.]

[scotty1024]

Have you considered probe issues? Are the wires to the probe pulling on it on one side of the bed and lose on the other side?

Scotty

On Jun 25, 2014, at 7:57 AM, Idzuwan notifications@github.com wrote:

this problem getting weirder as it progress, I re-did my Y-Carriages and level everything, I even put back the spring to level my bed manually and once it leveled, I did test pring again with heat bed off to eliminated glass/pcb warp, it still doing the same thing, and a perfectly even bed it still lifting up the right side to high (curling) and flowing the the bed (to low) on the left side to the extent I have to do the bed adjustment during printing!, I'm lost no idea why the firmware is doing this, now I'm printing with auto leveling turned of in the firmware until someone give me some idea how to actually debug this

— Reply to this email directly or view it on GitHub.

[idzuwan]

Hi,

I'm using those metal coupler check them too seem fine and tight,

Scotty,

wire are loose enough no pulling the wire, but we never know.. I just ordered a opto endstop for the z probe will try my luck again with auto leveling when it arrive

[idzuwan]

I have some time and motivation to try this again, this time I'm printing with heatbed off, I noticed that the correction made by the firmware seem inverted? when it suppose to go up it actually go down and vise versa :-/, as I know my left bed is high and my right side is low, when it on the right side it flowing on the bed, I have actually pause the print and twist my z coupler to compesate the autoleveling correction,I'm gonna download a fresh copy of the firmware source re-do the config again just incase I mess up the current one

[idzuwan]

more update :),

I disabled autoleveling, and print as normally I do with manual bed leveling, it printed normal no issues, then I re-enabled the auto level same issues happen even with the bed actually leveled when the nozzle travel to left side the firmware raised the hotend to high for no particular reason to high which cause curling on the bed (skirt) and when on the right side it flow the bed

I try to download a new fresh copy but having another problem with Z feed rate are way to high! getting frustrated now lol, haven't done any actual printing for almost a week now, and I notice that it always plow the bed at Y40,X20 (which is my first point of the grid)

this is the actual log when it probing the bed

23:11:17.461 : Bed x: 20.00 y: 40.00 z: 18.71 23:11:29.045 : Bed x: 180.00 y: 40.00 z: 19.77 23:11:40.489 : Bed x: 180.00 y: 180.00 z: 20.31 23:11:53.228 : Bed x: 20.00 y: 180.00 z: 19.08 23:11:53.237 : Eqn coefficients: a: 0.01 b: 0.00 d: 18.39 23:11:53.237 : planeNormal x: -0.01 y: -0.00 z: 1.00 23:11:53.243 : echo:endstops hit: Z:19.08

984

[nophead]

You say your bed is level but the probe is returning a variation of 1.6mm! So it looks like your probe does not work accurately or there is something wrong with the Z axis.

[idzuwan]

Hi,

that why been bashing my head to figure out why im getting bad probe results, I guess no choice to wait for a new micro switch and do printing without auto leveling for now :) On 28 Jun 2014 23:33, "Chris" notifications@github.com wrote:

You say your bed is level but the probe is returning a variation of 1.6mm! So it looks like your probe does not work accurately or there is something wrong with the Z axis.

— Reply to this email directly or view it on GitHub https://github.com/ErikZalm/Marlin/issues/972#issuecomment-47430386.

[nophead]

Does it give consistent results if you run the levelling multiple times or is it random?

[idzuwan]

will need to check it again later it almost 1am now here need a break from this for now, I will post a few probe results in the morning :)

thanks On 29 Jun 2014 00:20, "Chris" notifications@github.com wrote:

Does it give consistent results if you run the levelling multiple times or is it random?

— Reply to this email directly or view it on GitHub https://github.com/ErikZalm/Marlin/issues/972#issuecomment-47431694.

[idzuwan]

Test 1

06:09:17.131 : Bed x: 20.00 y: 40.00 z: 18.71 06:09:28.675 : Bed x: 180.00 y: 40.00 z: 19.85 06:09:40.157 : Bed x: 180.00 y: 180.00 z: 20.30 06:09:52.892 : Bed x: 20.00 y: 180.00 z: 19.07 06:09:52.900 : Eqn coefficients: a: 0.01 b: 0.00 d: 18.42 06:09:52.900 : planeNormal x: -0.01 y: -0.00 z: 1.00 06:09:52.904 : echo:endstops hit: Z:19.07

Test 2

06:10:43.397 : Bed x: 20.00 y: 40.00 z: 19.77 06:10:54.916 : Bed x: 180.00 y: 40.00 z: 20.96 06:11:06.397 : Bed x: 180.00 y: 180.00 z: 21.43 06:11:19.116 : Bed x: 20.00 y: 180.00 z: 20.21 06:11:19.125 : Eqn coefficients: a: 0.01 b: 0.00 d: 19.49 06:11:19.125 : planeNormal x: -0.01 y: -0.00 z: 1.00 06:11:19.128 : echo:endstops hit: Z:20.21

Test 3

06:11:53.500 : Bed x: 20.00 y: 40.00 z: 19.87 06:12:05.063 : Bed x: 180.00 y: 40.00 z: 20.97 06:12:16.529 : Bed x: 180.00 y: 180.00 z: 21.47 06:12:29.300 : Bed x: 20.00 y: 180.00 z: 20.17 06:12:29.304 : Eqn coefficients: a: 0.01 b: 0.00 d: 19.56 06:12:29.308 : planeNormal x: -0.01 y: -0.00 z: 1.00 06:12:29.314 : echo:endstops hit: Z:20.17

Test 4

06:12:57.560 : Bed x: 20.00 y: 40.00 z: 19.89 06:13:09.111 : Bed x: 180.00 y: 40.00 z: 21.00 06:13:20.642 : Bed x: 180.00 y: 180.00 z: 21.35 06:13:33.316 : Bed x: 20.00 y: 180.00 z: 20.25 06:13:33.321 : Eqn coefficients: a: 0.01 b: 0.00 d: 19.65 06:13:33.323 : planeNormal x: -0.01 y: -0.00 z: 1.00 06:13:33.327 : echo:endstops hit: Z:20.25

Test 5

06:14:01.846 : Bed x: 20.00 y: 40.00 z: 19.70 06:14:13.380 : Bed x: 180.00 y: 40.00 z: 20.86 06:14:24.866 : Bed x: 180.00 y: 180.00 z: 21.30 06:14:37.609 : Bed x: 20.00 y: 180.00 z: 20.05 06:14:37.613 : Eqn coefficients: a: 0.01 b: 0.00 d: 19.42 06:14:37.617 : planeNormal x: -0.01 y: -0.00 z: 1.00 06:14:37.623 : echo:endstops hit: Z:20.05

I guess something really wrong with my probe :-1:

[Guilty73]

any news on this? I happen to have the same problem

[MozzaRozza]

I shall now ramble on, and give you some thoughts to chew on !!!!!!!!!!!!!!!!!

1..... My auto leveling works perfectly, except the first (base) layer is very thin on front left corner Machine is Prusa i3, with standard hotbed and 3mm glass clipped on

2.....I found that if I pressed on front left corner of glass, and its diagonal opposite, the glass rocked a little i.e. not in full contact with hotbed

3...... adjusted bed screws until glass was rock solid on bed, irrespective of the "inclination" of the bed

4.......Added 2nd piece of glass to give 6mm SO...........................................still printing thin on left front......sh*t !!!!

5.....applied logic........if the pressure of the probe hitting the glass was enough to "bend" it downwards by say, 0.2mm, the reading would be 0.2mm below the glass surface, which would spring back when the probe removes itself. Why would this happen....am suspicious of the table itself. On my design, there are two bushes on the right "Y" rod, but only one on centrally the left i.e. the Aluminium table blank is , sort of, triangular. if you press on the FRONT RIGHT corner, it is solid ----- but if you press on the FRONT LEFT, the entire table ---Aluminium, Hotbed, and glass moves a little.

6.......Thus, when the first layer is put down, the software now thinks the front left is 0.2mm lower than it really is

7....CHEAT.........when the first layer shows thin on the very first pass of the head, quickly tweak the front left bed adjuster one turn, to pull the corner down a bit

8.....This ain't the answer.......but I do then get a perfect base layer !!!!!!!!!

9....Am not giving up.................It's personal now.....It should work flawlessly You shouldn't have to cheat by twiddling knobs ...........................Bye Regards to Guilty73, who is also suffering

[MozzaRozza]

Sorry Guilty 73 is also suffering with thin corners on base layer....I have left comments !!!!!!!!!!!!!!!!!!!!!

Date: Thu, 17 Jul 2014 09:32:24 -0700 From: notifications@github.com To: Marlin@noreply.github.com CC: morriscoleman@hotmail.com Subject: Re: [Marlin] Auto Level not compensate enough (#972)

any news on this? I happen to have the same problem

— Reply to this email directly or view it on GitHub.

[Guilty73]

Ok, MozzaRozza, thanks to let me know, I have a completely different setup then yours and my bed is rock solid. So I am still not sure if is an hardware or software problem.. been struggling with it for more then a week now....

[MozzaRozza]

Well.....have been at it for 3 - 4 weeks........don't think is software, as zillions of people have got it to work OK

And the nice solid bed.....mine seems to pretty goo.... have checked with a ial guage to see how much probe "bends" the glass

It seems to be just .03 mm....not enough !!!

Keep trying..............😁

Date: Thu, 17 Jul 2014 12:31:51 -0700 From: notifications@github.com To: Marlin@noreply.github.com CC: morriscoleman@hotmail.com Subject: Re: [Marlin] Auto Level not compensate enough (#972)

Ok, MozzaRozza, thanks to let me know, I have a completely different setup then yours and my bed is rock solid. So I am still not sure if is an hardware or software problem.. been struggling with it for more then a week now....

— Reply to this email directly or view it on GitHub.

[Guilty73]

Hey MozzaRozza, I know what you are saying, the software should be working, but as it been tested on a core XY with M10 Lead screw and Reversed Bed movment? I am not sure since I could not find any information online, for my setup.

Meanwhile I keep on trying, Will let you know If I find a solution, or If I understand the problem better.

On 18 July 2014 00:29, MozzaRozza notifications@github.com wrote:

Well.....have been at it for 3 - 4 weeks........don't think is software, as zillions of people have got it to work OK

And the nice solid bed.....mine seems to pretty goo.... have checked with a ial guage to see how much probe "bends" the glass

It seems to be just .03 mm....not enough !!!

Keep trying..............😁

Date: Thu, 17 Jul 2014 12:31:51 -0700 From: notifications@github.com To: Marlin@noreply.github.com CC: morriscoleman@hotmail.com Subject: Re: [Marlin] Auto Level not compensate enough (#972)

Ok, MozzaRozza, thanks to let me know, I have a completely different setup then yours and my bed is rock solid. So I am still not sure if is an hardware or software problem.. been struggling with it for more then a week now....

— Reply to this email directly or view it on GitHub.

— Reply to this email directly or view it on GitHub https://github.com/ErikZalm/Marlin/issues/972#issuecomment-49375168.

[boelle]

is the issue still there?

my thought was that the probe offset is not correct...

[MozzaRozza]

I have now constructed a beefed up servo arm, and a better micro switch with roller lever, and generally the prints are brilliant, including the base layer. I have screwed the table down rock solid, and although It has a definite tilt on the X axis, I do 16 probe points, and the software seems to cope. Occasionally the offset goes out......but this is probably movement in the head mountings. If the head shifts by only, say, 0.2 mm, the effect on the base layer is dramatic.

Sent from my iPad

On 19 Dec 2014, at 21:19, Bo Herrmannsen notifications@github.com wrote:

is the issue still there?

my thought was that the probe offset is not correct...

— Reply to this email directly or view it on GitHub.re seems

[boelle]

so this seems more like a hardware problem than a software problem...?

[MozzaRozza]

I suspect hundreds of people have the auto levelling working very well. As we all use the same version, probably with identical Ramps and Arduino Mega 2560, it seems the software is proven. All you can do is reinstall the Marlin to make sure your PC is not sending a corruption in to the Arduino. HARDWARE I suspect is the weakness. For a start, on mine I know if I set bed temp at 100c, it has all sorts of temps over the area.... In one corner it is only 78c !!! This temp spread must cause variable warping in the glass. All you can do is bolt the bed down solid, and don't worry if it "tilts" in one direction. The Auto levelling can take care of that. The servo arm must be solid and repeatable....., despite forum suggestions I use 16 probes not 9.....works !!!!!

Sent from my iPhone

On 20 Dec 2014, at 09:10, Bo Herrmannsen notifications@github.com wrote:

so this seems more like a hardware problem than a software problem...?

— Reply to this email directly or view it on GitHub.

[boelle]

also try and let the bed heat for 5-10 min first so the heat is more even

Den lørdag den 20. december 2014 skrev MozzaRozza <notifications@github.com

:

I suspect hundreds of people have the auto levelling working very well. As we all use the same version, probably with identical Ramps and Arduino Mega 2560, it seems the software is proven. All you can do is reinstall the Marlin to make sure your PC is not sending a corruption in to the Arduino. HARDWARE I suspect is the weakness. For a start, on mine I know if I set bed temp at 100c, it has all sorts of temps over the area.... In one corner it is only 78c !!! This temp spread must cause variable warping in the glass. All you can do is bolt the bed down solid, and don't worry if it "tilts" in one direction. The Auto levelling can take care of that. The servo arm must be solid and repeatable....., despite forum suggestions I use 16 probes not 9.....works !!!!!

Sent from my iPhone

On 20 Dec 2014, at 09:10, Bo Herrmannsen <notifications@github.com javascript:_e(%7B%7D,'cvml','notifications@github.com');> wrote:

so this seems more like a hardware problem than a software problem...?

— Reply to this email directly or view it on GitHub.

— Reply to this email directly or view it on GitHub https://github.com/ErikZalm/Marlin/issues/972#issuecomment-67731377.

[MozzaRozza]

Probably good idea, and an enclosure would help

Sent from my iPhone

On 20 Dec 2014, at 12:12, Bo Herrmannsen notifications@github.com wrote:

also try and let the bed heat for 5-10 min first so the heat is more even

Den lørdag den 20. december 2014 skrev MozzaRozza <notifications@github.com

:

I suspect hundreds of people have the auto levelling working very well. As we all use the same version, probably with identical Ramps and Arduino Mega 2560, it seems the software is proven. All you can do is reinstall the Marlin to make sure your PC is not sending a corruption in to the Arduino. HARDWARE I suspect is the weakness. For a start, on mine I know if I set bed temp at 100c, it has all sorts of temps over the area.... In one corner it is only 78c !!! This temp spread must cause variable warping in the glass. All you can do is bolt the bed down solid, and don't worry if it "tilts" in one direction. The Auto levelling can take care of that. The servo arm must be solid and repeatable....., despite forum suggestions I use 16 probes not 9.....works !!!!!

Sent from my iPhone

On 20 Dec 2014, at 09:10, Bo Herrmannsen <notifications@github.com javascript:_e(%7B%7D,'cvml','notifications@github.com');> wrote:

so this seems more like a hardware problem than a software problem...?

— Reply to this email directly or view it on GitHub.

— Reply to this email directly or view it on GitHub https://github.com/ErikZalm/Marlin/issues/972#issuecomment-67731377.

— Reply to this email directly or view it on GitHub.

[boelle]

enclosure will not make the heat more even

Den lørdag den 20. december 2014 skrev MozzaRozza <notifications@github.com

:

Probably good idea, and an enclosure would help

Sent from my iPhone

On 20 Dec 2014, at 12:12, Bo Herrmannsen <notifications@github.com javascript:_e(%7B%7D,'cvml','notifications@github.com');> wrote:

also try and let the bed heat for 5-10 min first so the heat is more even

Den lørdag den 20. december 2014 skrev MozzaRozza < notifications@github.com javascript:_e(%7B%7D,'cvml','notifications@github.com');

:

I suspect hundreds of people have the auto levelling working very well. As we all use the same version, probably with identical Ramps and Arduino Mega 2560, it seems the software is proven. All you can do is reinstall the Marlin to make sure your PC is not sending a corruption in to the Arduino. HARDWARE I suspect is the weakness. For a start, on mine I know if I set bed temp at 100c, it has all sorts of temps over the area.... In one corner it is only 78c !!! This temp spread must cause variable warping in the glass. All you can do is bolt the bed down solid, and don't worry if it "tilts" in one direction. The Auto levelling can take care of that. The servo arm must be solid and repeatable....., despite forum suggestions I use 16 probes not 9.....works !!!!!

Sent from my iPhone

On 20 Dec 2014, at 09:10, Bo Herrmannsen <notifications@github.com javascript:_e(%7B%7D,'cvml','notifications@github.com'); <javascript:_e(%7B%7D,'cvml','notifications@github.com javascript:_e(%7B%7D,'cvml','notifications@github.com');');>> wrote:

so this seems more like a hardware problem than a software problem...?

— Reply to this email directly or view it on GitHub.

— Reply to this email directly or view it on GitHub https://github.com/ErikZalm/Marlin/issues/972#issuecomment-67731377.

— Reply to this email directly or view it on GitHub.

— Reply to this email directly or view it on GitHub https://github.com/ErikZalm/Marlin/issues/972#issuecomment-67733862.

[MozzaRozza]

Fair enough... But it is cold in Norway !!

Sent from my iPhone

On 20 Dec 2014, at 12:16, Bo Herrmannsen notifications@github.com wrote:

enclosure will not make the heat more even

Den lørdag den 20. december 2014 skrev MozzaRozza <notifications@github.com

:

Probably good idea, and an enclosure would help

Sent from my iPhone

On 20 Dec 2014, at 12:12, Bo Herrmannsen <notifications@github.com javascript:_e(%7B%7D,'cvml','notifications@github.com');> wrote:

also try and let the bed heat for 5-10 min first so the heat is more even

Den lørdag den 20. december 2014 skrev MozzaRozza < notifications@github.com javascript:_e(%7B%7D,'cvml','notifications@github.com');

:

I suspect hundreds of people have the auto levelling working very well. As we all use the same version, probably with identical Ramps and Arduino Mega 2560, it seems the software is proven. All you can do is reinstall the Marlin to make sure your PC is not sending a corruption in to the Arduino. HARDWARE I suspect is the weakness. For a start, on mine I know if I set bed temp at 100c, it has all sorts of temps over the area.... In one corner it is only 78c !!! This temp spread must cause variable warping in the glass. All you can do is bolt the bed down solid, and don't worry if it "tilts" in one direction. The Auto levelling can take care of that. The servo arm must be solid and repeatable....., despite forum suggestions I use 16 probes not 9.....works !!!!!

Sent from my iPhone

On 20 Dec 2014, at 09:10, Bo Herrmannsen <notifications@github.com javascript:_e(%7B%7D,'cvml','notifications@github.com'); <javascript:_e(%7B%7D,'cvml','notifications@github.com javascript:_e(%7B%7D,'cvml','notifications@github.com');');>> wrote:

so this seems more like a hardware problem than a software problem...?

— Reply to this email directly or view it on GitHub.

— Reply to this email directly or view it on GitHub https://github.com/ErikZalm/Marlin/issues/972#issuecomment-67731377.

— Reply to this email directly or view it on GitHub.

— Reply to this email directly or view it on GitHub https://github.com/ErikZalm/Marlin/issues/972#issuecomment-67733862.

— Reply to this email directly or view it on GitHub.

[boelle]

javist, så er det nok en god ide 😜

for the english speaking that means if very cold an enclosur could be a good idea

Den lørdag den 20. december 2014 skrev MozzaRozza <notifications@github.com

:

Fair enough... But it is cold in Norway !!

Sent from my iPhone

On 20 Dec 2014, at 12:16, Bo Herrmannsen <notifications@github.com javascript:_e(%7B%7D,'cvml','notifications@github.com');> wrote:

enclosure will not make the heat more even

Den lørdag den 20. december 2014 skrev MozzaRozza < notifications@github.com javascript:_e(%7B%7D,'cvml','notifications@github.com');

:

Probably good idea, and an enclosure would help

Sent from my iPhone

On 20 Dec 2014, at 12:12, Bo Herrmannsen <notifications@github.com javascript:_e(%7B%7D,'cvml','notifications@github.com'); <javascript:_e(%7B%7D,'cvml','notifications@github.com javascript:_e(%7B%7D,'cvml','notifications@github.com');');>> wrote:

also try and let the bed heat for 5-10 min first so the heat is more even

Den lørdag den 20. december 2014 skrev MozzaRozza < notifications@github.com javascript:_e(%7B%7D,'cvml','notifications@github.com'); <javascript:_e(%7B%7D,'cvml','notifications@github.com javascript:_e(%7B%7D,'cvml','notifications@github.com');');>

:

I suspect hundreds of people have the auto levelling working very well. As we all use the same version, probably with identical Ramps and Arduino Mega 2560, it seems the software is proven. All you can do is reinstall the Marlin to make sure your PC is not sending a corruption in to the Arduino. HARDWARE I suspect is the weakness. For a start, on mine I know if I set bed temp at 100c, it has all sorts of temps over the area.... In one corner it is only 78c !!! This temp spread must cause variable warping in the glass. All you can do is bolt the bed down solid, and don't worry if it "tilts" in one direction. The Auto levelling can take care of that. The servo arm must be solid and repeatable....., despite forum suggestions I use 16 probes not 9.....works !!!!!

Sent from my iPhone

On 20 Dec 2014, at 09:10, Bo Herrmannsen < notifications@github.com javascript:_e(%7B%7D,'cvml','notifications@github.com'); <javascript:_e(%7B%7D,'cvml','notifications@github.com javascript:_e(%7B%7D,'cvml','notifications@github.com');');> <javascript:_e(%7B%7D,'cvml','notifications@github.com javascript:_e(%7B%7D,'cvml','notifications@github.com'); <javascript:_e(%7B%7D,'cvml','notifications@github.com javascript:_e(%7B%7D,'cvml','notifications@github.com');');>');>> wrote:

so this seems more like a hardware problem than a software problem...?

— Reply to this email directly or view it on GitHub.

— Reply to this email directly or view it on GitHub < https://github.com/ErikZalm/Marlin/issues/972#issuecomment-67731377>.

— Reply to this email directly or view it on GitHub.

— Reply to this email directly or view it on GitHub https://github.com/ErikZalm/Marlin/issues/972#issuecomment-67733862.

— Reply to this email directly or view it on GitHub.

— Reply to this email directly or view it on GitHub https://github.com/ErikZalm/Marlin/issues/972#issuecomment-67733998.

[boelle]

did a enclosure solve the issue?

want to know if this was an hardware issue or software... i lean towards an hardware issue

[MozzaRozza]

A lot of building gone on since the posts !!!! I think the enclosure is only a minor contributor

I have now built TWO working Prusa i3 with auto-leveling (and water cooling).....and the leveling problems have disappeared on both. One has enclosure, one does not.

Pretty sure it is the hardware/physics of the table setup..........my simpler Prusa has a single hotbed, plus 3mm glass. The steps I took were to :-

1/ Make sure that when the servo deploys the micro-switch, the servo arm is vertical, and substantial. Any bend in the servo arm is catastrophic....it MUST be perfectly repeatable. I use servos with the little roller at the end of the arm.

2/ Beef up the springs on the four corner screws of the table until almost solid !! A little bit of springiness is OK, in case the head crashes in to the glass.........but it should be quite solid..... When solid, lay the glass on top, and see if it jiggles ---- you may well find it is "high" on 2 opposite corners....adjust until no jiggle in glass......and clamp glass. As I said in post, the table may even not be at all "level" to the head, but it doesn't matter --- the auto leveling will compensate for this.........THE PRIORITY IS ---- FLAT AND SOLID............being perfectly horizontal is not vital.

3/ My new Prusa build worked auto-levelling without any problems......but it does have TWO hotbeds sandwiched together. This serves two purposes ----- very quick heat-up.......110C in 2 minutes....and a very rigid table ---- maybe that explains. All in all, the new one, with my own water-jacketed head, prints like a demon !!!!!!!!!!!. You do need two power supplies though....one for the hotbeds (they take 18A) and one for all the rest......(takes about 5A). This stops the voltage pull-down when the hotbed heat kicks in causing instability in the electronics.

The software on both machines is identical, .......... Atmega 2560 + Ramps 1.4 (cheap Chinese clones) Run most prints on Cura, with modified start code to give the auto level, extrude 30mm starter filament in one corner, and then immediately in to print.

Fell free to come back again....if you think a few pictures would help....or whatever !!!

Date: Wed, 11 Mar 2015 01:53:36 -0700 From: notifications@github.com To: Marlin@noreply.github.com CC: morriscoleman@hotmail.com Subject: Re: [Marlin] Auto Level not compensate enough (#972)

did a enclosure solve the issue?

want to know if this was an hardware issue or software... i lean towards an hardware issue

— Reply to this email directly or view it on GitHub.

[boelle]

ok, i will then conclude that this was a hardware issue and not an issue with marlin as such

closing

can always reopen og open new issue later

[github-actions[bot]]

This issue has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.