search

knutwurst / Marlin-2-0-x-Anycubic-i3-MEGA-S

Marlin 2.0.x Version for Anycubic i3 MEGA M/S/P/X/CHIRON and 4MAX with Anycubic TFT or the "new" DGUS Clone TFT - Now also with BLTouch!
GNU General Public License v3.0
788 stars 187 forks source link

[BUG] ABL not correct #103

Closed serioustk closed 3 years ago

serioustk commented 3 years ago

Bug Description

When running ABL, leveling is not done correct. Grid seems prone, test pattern does not look the same on the whole base. Please find pictures below.

My Configurations

MEGA_S_DGUS_TMC_BLT_10_v1.1.9.hex

Steps to Reproduce

  1. Install knutwurst marlin firmware
  2. Install original BLTouch according to this tutorial.
  3. Run auto bed leveling from special menu
  4. Print this leveling test GCODE bed_leveling_test.zip

Expected behavior: [What you expect to happen] Test squares look identical in each corner of bed and in center.

Actual behavior: [What actually happens] Test squares are not identcal. On the left side, test squares look fine while on the right side nozzle was too far away from bed while printing.

Additional Information

I use Anycubic i3 Mega S brought in early 2021 with blue/yellow display and mainboard without gorilla logo. I experienced issues with an uneven ultrabase. Manually, I was able to level the corners correctly but the center was too low so I decided to add an ABL sensor.

If you look at the test print, the leveling map of the ABL seems to be prone to the right side. As you can see on the pictures, test squares on the left side (front and back) look as expected, center is not so well and the right side is far from fine.

What I have tried / checked so far:

I really can't figure why the test squares does not look identical. I'm glad to provide further information or do some testing if required.

Pictures

Overview

overview

Front left

front left

Center

center

Front right

front right

v-a-d-e-r commented 3 years ago

Same crap here with orig. BLTouch... :-(

20210312183852141

haupthakan commented 3 years ago

Same for me but I did not print these squares. But on the right side the nozzle is always way lower than on the left side. I already printed some parts but I always put my parts on the front left and now the big part where is realized that. I tried the same as you. Base higher and playing with Z offset but nothing worked.

IMG_1494 IMG_1493

v-a-d-e-r commented 3 years ago

Ok, 3(+) people can not be stupid in the same way. ;-) I assume that the mesh values are not part of the Z-offset calculation (just ignored)!? I get by far more better print results with manual leveling and a piece of paper!!! That was not the goal of buying an (expensive) orig. BLTouch sensor....

haupthakan commented 3 years ago

Yeah so I tried so things and got a bit better but it's still unusable. Im not able zu use this printer till there is a solution. I'll try some other firmwares maybe these will fix this issue. If it fixes I'll post it in here

v-a-d-e-r commented 3 years ago

Ok, this is my 2nd Anycubic Mega printer (the 1st one told me after two weeks the print bed has -14°C! I didn't believe that...g --> warranty case...). With manual leveling I get useable prints. But it is annoying to do leveling with a piece of paper. So I hope the issue will be fixed soon...

vergie commented 3 years ago

Please check the correct vertical frame alligment on the left and right side (using a spirit level could be enough). Then also check the vertical allignemnt of both X rods. If there is a differnce between the left and right side you have a problem with the BL Touch. Responsibe for the differenz in this case is the Y-offset of BL-Touch. This Problem was described in detail here (in german) [https://makerware.thingiverse.com/thing:2230882]

haupthakan commented 3 years ago

Please check the correct vertical frame alligment on the left and right side (using a spirit level could be enough). Then also check the vertical allignemnt of both X rods.

If there is a differnce between the left and right side you have a problem with the BL Touch. Responsibe for the differenz in this case is the Y-offset of BL-Touch. This Problem was described in detail here (in german) [https://makerware.thingiverse.com/thing:2230882]

I've read it and his idea is right If your carriage would sag a bit the bl would sit lower so maybe my x axis is twisted. But I've tried another firmware and it went better as with this one Maybe I'll install the manual meshbed version for now and print the mk4 xcarriage where the y is almost 0 I think. Thanks for your answer. I would've never think of this :)

haupthakan commented 3 years ago

Please check the correct vertical frame alligment on the left and right side (using a spirit level could be enough). Then also check the vertical allignemnt of both X rods.

If there is a differnce between the left and right side you have a problem with the BL Touch. Responsibe for the differenz in this case is the Y-offset of BL-Touch. This Problem was described in detail here (in german) [https://makerware.thingiverse.com/thing:2230882]

I've checked the x rod alignment and they seam really the same so this doesn't seam to be the problem. :/ Even if there would be I wouldn't have found a way to adjust the x rods

haupthakan commented 3 years ago

If someone is interested I'll put my mesh points in here. This is different almost everytime. Between those I didn't touch anything. I just ran it couple of times

Recv: 0 1 2 3 4 Recv: 0 +0.565 +0.532 +0.490 +0.415 +0.357 Recv: 1 +0.500 +0.642 +0.622 +0.480 +0.387 Recv: 2 +0.578 +0.625 +0.598 +0.520 +0.395 Recv: 3 +0.612 +0.678 +0.617 +0.472 +0.420 Recv: 4 +0.675 +0.650 +0.573 +0.448 +0.307

Recv: Bilinear Leveling Grid: Recv: 0 1 2 3 4 Recv: 0 +0.347 +0.522 +0.482 +0.377 +0.312 Recv: 1 +0.522 +0.622 +0.622 +0.457 +0.360 Recv: 2 +0.600 +0.673 +0.607 +0.525 +0.425 Recv: 3 +0.710 +0.717 +0.610 +0.515 +0.420 Recv: 4 +0.740 +0.725 +0.605 +0.490 +0.327

Recv: Bilinear Leveling Grid: Recv: 0 1 2 3 4 Recv: 0 +0.382 +0.617 +0.558 +0.445 +0.365 Recv: 1 +0.652 +0.735 +0.698 +0.568 +0.430 Recv: 2 +0.710 +0.767 +0.707 +0.588 +0.462 Recv: 3 +0.770 +0.787 +0.707 +0.575 +0.477 Recv: 4 +0.835 +0.820 +0.673 +0.517 +0.325

Recv: Bilinear Leveling Grid: Recv: 0 1 2 3 4 Recv: 0 +0.565 +0.665 +0.630 +0.507 +0.418 Recv: 1 +0.698 +0.800 +0.757 +0.607 +0.475 Recv: 2 +0.773 +0.820 +0.740 +0.632 +0.495 Recv: 3 +0.830 +0.855 +0.755 +0.617 +0.510 Recv: 4 +0.905 +0.852 +0.715 +0.550 +0.372

v-a-d-e-r commented 3 years ago

I also made 3 tests in a row, but the values are very close to each other....

< Bilinear Leveling Grid: < 0 1 2 3 4 < 0 +0.987 +1.130 +1.231 +1.263 +1.231 < 1 +0.997 +1.113 +1.221 +1.256 +1.228 < 2 +0.955 +1.110 +1.253 +1.327 +1.347 < 3 +1.031 +1.142 +1.241 +1.285 +1.265 < 4 +1.066 +1.184 +1.263 +1.275 +1.251

< Bilinear Leveling Grid: < 0 1 2 3 4 < 0 +0.985 +1.123 +1.226 +1.256 +1.231 < 1 +0.987 +1.103 +1.206 +1.251 +1.224 < 2 +0.945 +1.103 +1.251 +1.327 +1.354 < 3 +1.029 +1.137 +1.236 +1.283 +1.260 < 4 +1.059 +1.177 +1.258 +1.273 +1.253

< Bilinear Leveling Grid: < 0 1 2 3 4 < 0 +0.980 +1.108 +1.214 +1.243 +1.219 < 1 +0.987 +1.100 +1.206 +1.248 +1.221 < 2 +0.940 +1.098 +1.241 +1.322 +1.354 < 3 +1.022 +1.132 +1.224 +1.268 +1.253 < 4 +1.054 +1.172 +1.248 +1.263 +1.236

haupthakan commented 3 years ago

How did you get those values? Did you change something since your first post? Are you still getting a bad first layer? yeah I would be more than fine with these values :D but I'm not getting close to really good values :/ I haven't even really used my printer today because I'm not even close to a solution to this :( I've found a new mount for the BL-Touch where the probe is on the same Y as the nozzle. Maybe this would help for the beginning.

v-a-d-e-r commented 3 years ago

Nope, I changed nothing....

haupthakan commented 3 years ago

I ran the probe accuracy test and the probe should be alright. I think it said something like 0.017. I'm really frustrated because manual mesh leveling works fine but it's very time consuming.

v-a-d-e-r commented 3 years ago

Yeah, it is. But the print test I used is really extreme. It can also be that the temperature in the edges is too low.

serioustk commented 3 years ago

Please check the correct vertical frame alligment on the left and right side (using a spirit level could be enough). Then also check the vertical allignemnt of both X rods. If there is a differnce between the left and right side you have a problem with the BL Touch. Responsibe for the differenz in this case is the Y-offset of BL-Touch. This Problem was described in detail here (in german) [https://makerware.thingiverse.com/thing:2230882]

I've read it and his idea is right If your carriage would sag a bit the bl would sit lower so maybe my x axis is twisted. But I've tried another firmware and it went better as with this one Maybe I'll install the manual meshbed version for now and print the mk4 xcarriage where the y is almost 0 I think. Thanks for your answer. I would've never think of this :)

This was the hint I needed, thank you!

For all who do not understand german here is a quick translation of the referenced post: The shafts of the X-Axis are not perfectly perpendicular to each other compared to the base. This varies depending on the X-position of the extruder unit. This leads to a different tilt of the extruder unit depending on the X-position. As the probe has a offset in Y direction compared to the extroder, the probe is sometimes more, sometimes less far away from the bed leading to wrong results.

I was able to adcust the X-shafts with a small spirit level loosing the holder screws of the shafts and turning them around a little. With a littie trying around, the spiril level now only shows a very small offset over the whole X-Axis leading to a very good looking test pattern using BLtouch calibration.

Thank you! This issue is solved and can now be closed.

github-actions[bot] commented 3 years ago

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.