Y axis prints wrong length

General Discussion
1

hello everyone I finally managed to get my Anycubic Kobra Plus to run with Klipper,

BUT have I the following problem, which I can’t solve.
My parameters from the y-axis don’t agree. I only have a print length of 120mm on a distance of 150mm (now I’ve managed to get 141by changing the rotation_distance: from 40 to 34.)
I can’t change it anymore because the calibration fits exactly on my print bed.

Do you have an idea what I need to change, can?

[stepper_y]

step_pin: PC4

dir_pin: PA7

enable_pin: !PC3

microsteps: 32

rotation_distance: 34 <------

endstop_pin: !PC5

position_min: -22.5

position_endstop: -22.5

position_max: 305

homing_speed: 100

[tmc2209 stepper_y]

uart_pin:PA15

tx_pin:PA9

sense_resistor: 0.150

run_current: 0.9

uart_address: 1

stealthchop_threshold: 999999

BR Daniel

2

Looking at the printer it should have the default rotation_distance of 40. This setting depends on the hardware, e.g. pulley teeth count, belt pitch etc.

3

hi, yes normal is 40 right,
i changed it to 35! so that the printed length “becomes longer” but, as I have already written, I cannot reduce this value (see text above)

So I need more ideas what to do…

4

I don’t get it.
For me it sounds like you are messing with symptoms instead of finding the reason:

  • rotation_distance depends on the hardware
  • This is not a setting to mess around.
  • If your printer is as per the manufacturers standard and 40 is correct, then it is 40
  • If you have other issues, then it is not related to rotation_distance

Sorry, but I do not understand your problem.

6

sigh
Honestly, with this description, nobody can help you.
Post:

  • Maybe a picture
  • printer.cfg
  • klippy.log
  • sliced gcode file

Edit: If you find English troublesome, I suggest to use DeepL Translate: The world's most accurate translator

7

Actually I didn’t need any help in English but OK :slight_smile:
Topic is: My Anycubic Kobra +3D printer, after I flashed it with Klipper, prints in the y-axis to 300mm → 60mm shorter. that means my self-made STL. coordinate cross is correct in the x-axis but in the y-axis exactly (25%) of the length is missing. For this problem I was looking for a solution here, I assume that someone can help me.

8

I think first we have to find out if Kobra Plus’ motion system properties.

How many teeth are on the gears for X and Y. Are those gears genuine ore a mod.
Also: Are all screw tighten up.

Has the model accidentally been modified in size within the slicer.

Can you please share the zipped gcode file here in your next post?

9

What stepper motors are you using? are they 1.8* or 0.9*…
the math for rotation distance is the following:

1.8deg motor 200 full steps * 16 microsteps = 3200 steps per full rotation / 80 steps per mm (2mm belt & 20 teeth pully) = 40mm full rotation (rotation_distance)

image

FYI a good belt calculator to get steps per mm is included in repetier-host
https://www.klipper3d.org/Rotation_Distance.html?h=rotation_distance#belt-driven-axes
This is from Klipper docs…

10

I think you’re getting some excellent suggestions to resolve you problem.

One thing that jumps out at me is your use of “32” microsteps on the TMC2209. What happens when you change that to 16 (the “traditional” value that most people use)? I’m asking because as you increase the number of microsteps, the difference between them becomes less and less and the chance for slippage increases. I doubt that’s your problem, but it’s something that caught my eye.

11

Hello all together he is still many thanks for your solution approaches.

So I assume that: The Y-axis should actually have the same values as the x-axis of my Kobra. BECAUSE AS YOU CAN SEE IN THE PICTURES, THE X-AXIS IS PRINTED EXACTLY TO MM. (that’s why I think x and y have the seven number of teeth) regardless of the fact that I don’t know them (yet :wink:).

Attached both values from the CFG file once. As does the G code. From my little test print’s. I actually always use the Cura slicer, I didn’t change its data and settings after I flashed or clipped the Kobra.

I have the numbers in the cfg. file just made different to the x-axis to find out what is changing originally the y-axis had the same coordinates as the x-axis.

(INFO) I only printed half of the printer bed (150mm) the Kobra has a printer bed size of 300x 300mm

coordinate_cross_test_Kobra.gcode (489.9 KB)

Y axis printed measure

IMG_20230205_052144
IMG_20230205_052144432×640 341 KB

Sorry, there are only 17mm deviation not 20mm (was already late or early yesterday when I wrote it)

X axis printed measure
IMG_20230205_052225

these are the CFG data where x is correct and y is NOT

[stepper_x]
step_pin: PA5
dir_pin: PA4
enable_pin: !PC3
microsteps: 16
rotation_distance: 40
endstop_pin: !PA6
position_min: -6.2
position_endstop: -6.2
position_max: 304
homing_speed: 100

[tmc2209 stepper_x]
uart_pin:PA15
tx_pin:PA9
sense_resistor: 0.100
run_current: 0.9
uart_address: 3
stealthchop_threshold: 999999

[stepper_y]
step_pin: PC4
dir_pin: PA7
enable_pin: !PC3
microsteps: 16
rotation_distance: 40
endstop_pin: !PC5
position_min: -20.5
position_endstop: -20.5
position_max: 304
homing_speed: 100

[tmc2209 stepper_y]
uart_pin:PA15
tx_pin:PA9
sense_resistor: 0.100
run_current: 0.9
uart_address: 1
stealthchop_threshold: 999999

@mykepredko: I had that at first, of course, but now I just did a few tests that I found helpful. BGD

and @ EddyMI3D all screws are tight (that wouldn’t explain a 17mm difference either or. :wink: :yum:…)

:bangbang: Now you :sunglasses:… I can’t find the error :face_with_raised_eyebrow:

2 Likes
12

Have you checked the teeth on the gear on the Y-stepper?

13

I didn’t find it and I don’t want to take the printer apart BUT… the X axis prints correctly and has the same teeth and data…
Or what is your opinion

14

Just a question:

You bought the printer new or 2nd hand?

15

I’ve had this since last year (also printed for several hours) BUT just “clipped” and flashed last week… now the Y axis is 25% wrong.

and I can’t find the error…arrrggg

16

@Digedi
:frowning: have the EXACT same problem but not on klipper, when using the original 2.8.4 firmware all is good, but changing to Zombie v2 Full I have the same issue as you on the Y-Axis :frowning:

17

The Kobra Plus has different rotation_distances for the X and Y steppers. This can be seen in the Marlin configuration from the stock firmware:

#define DEFAULT_AXIS_STEPS_PER_UNIT   { 80, 100, 400, 405 }

Based on those values, the correct value for the Y rotation_distance is 32. This is different than on a Kobra, which uses the same value for X and Y.

Related – the endstop position values from your config appear to be for a Kobra, not a Kobra Plus; that’ll throw off the bed dimensions.

I used the Marlin configuration and built a Klipper config that seems to be working well. I need to finish up some of the cleanup before I submit a PR for it (checking for default values, getting the test added, etc.), but in the meantime here it is:

18

#EDACerton
Hello and first of all thank you for your message,
A.) I hadn’t found anywhere that x and y are different steppers have, but good to know.
B.) I actually thought I had tried everything and adjusted the stepper values to the 25% less distance…
However, after I did that, the calibration no longer worked.

You write below that another firmware must be installed, of course I have one on it that doesn’t work?

Otherwise I would just import your cfg file and see what happens.

19

and…unfortunately this cannot be opened here (page not found) :face_with_raised_eyebrow:

Bildschirm­foto 2023-03-06 um 01.37.09
Bildschirm­foto 2023-03-06 um 01.37.091294×882 77.4 KB

can you send me the CFG file again via PN? Please and thanks

20

Sorry about the link… I rebased that branch to get it ready for the PR, here’s a better link now:

If you’re already running Klipper, then you can just import the config file. The firmware comment is directed at folks that want to revert to Marlin — the modification changes the address of the X stepper, so to go back to Marlin you need a firmware with the address modified too.

1 Like
21

I’m guessing that this was a weird combination of two incorrect settings (rotation difference and endstop positions) resulting in a mostly-working config, but then breaking when one was fixed and the other wasn’t. The Kobra uses sensorless homing, the Plus uses endstop switches.

1 Like