Polar Printer Steps/Degree?

CNCModeller

I had some jitter on RRF 2 with my Duet ethernet, moving to RRF 3 seems to have cured it. But I've been rebuilding the printer since I upgraded the firmware so haven't done anything apart from test prints so far.

Phospherus

@cncmodeller
I am on 3.2.2 I tried using the latest beta also but polar seems broken in it. Would not really call it a jitter. Will check jerk settings.

dc42

@phospherus said in Polar Printer Steps/Degree?:

@cncmodeller
I am on 3.2.2 I tried using the latest beta also but polar seems broken in it. Would not really call it a jitter. Will check jerk settings.

In what way is 3.2.2 not working in polar mode?

Phospherus

@dc42

3.2.2 is working fine. I have a test print that is a 10mm width square that is almost the diameter of the bed. On 3.2.2 it prints fine. I tried upgrading to 3.3 beta 2 and it prints as a circle. Go back to 3.2.2 and works fine again.

dc42

@phospherus, please post your config.g file and your test print file, then I will try to reproduce the problem with 3.3beta.

Phospherus

@dc42

Here are the files you requested.

config.g

Test_Print.gcode

dc42

@phospherus, you have this in your config.g:

; Axis Limits
M208 X0 Y0 Z0 S1                              ; set axis minima
M208 X120 Y210 Z150 S0                        ; set axis maxima

but your GCode file contains this:

G1 X71.748 Y74.220 E9.6802
G1 X-71.748 Y74.220 E19.1301
G1 X-74.220 Y71.748 E19.3603

so you need to change your M208 S1 command to allow negative X and Y coordinates. M208 is always specified in Cartesian coordinates.

Phospherus

@dc42

Ok I understand what your saying. Question is though why did it work like that in 3.2.2 but not in the Beta firmware? And why would X ever go - value when 0 is center of bed? Perhaps I sliced it wrong after I updated to 3.3Beta2.

Another question more on setup of polar is my Y axis is setup for continuous rotation, No homing switch on Y. Does the Polar kinematics then ignore the 208 values? And if not what should it be?

dc42

@phospherus said in Polar Printer Steps/Degree?:

@dc42

Ok I understand what your saying. Question is though why did it work like that in 3.2.2 but not in the Beta firmware?

Perhaps the M208 limits were not to polar kinematics applied prior to 3.3beta.

And why would X ever go - value when 0 is center of bed? Perhaps I sliced it wrong after I updated to 3.3Beta2.

Please provide a test file that you have tested with 3.2.2beta.

Another question more on setup of polar is my Y axis is setup for continuous rotation, No homing switch on Y. Does the Polar kinematics then ignore the 208 values? And if not what should it be?

The M208 values are the Cartesian coordinate limits that you want to place on movement, not limits on the movement of the individual motors.

Phospherus

@dc42

Ok I went back to 3.3 beta2 for a sanity check. I will include my test print file that I just tried on 3.2.2 with the same config.

config.g

Test file is above limit for upload. https://drive.google.com/file/d/1TDfrflJFM2K79y0WpiGgJVcTUF-hkG9W/view?usp=sharing

First picture is behavior with 3.3 Beta2

Second is the same file printed on 3.2.2

As you you can see first picture it prints the circles then on infill does an arch is the same spot and just builds up until I stop it.

Second looks normal ofcourse. Not a complete print as I stopped it to do the 3.3 Beta test.

o_lampe

@phospherus
Do you use ArcWelder as post-processing script?

Phospherus

@o_lampe
No just sliced using Supersclicer.

Phospherus

@dc42

Did the above provide any insight?

shadi

@phospherus Hi can you please help me it seems i have the same issue my design is that y turntable plate and x is R should that work if i put Y value 2.5 is this what has been discussed in your subject ? and what about X value then ? M92 X?? Y2.5 is this how m92 should look like but i just would like to mention if that can work with SKR1.4 that has repaprfirmware on it ? I slice using Cura the output file from cura is linear coordinate and I assume the firmware should convert it to polar

Phospherus

@shadi
If you are using reprap on skr as far as I know the values should be the same but can not swear to that as I use OEM boards.

Your X axis is non rotational I assume so it would be figured by pulley teeth and step angle

So 200 steps = 1 revolution (360degress / 1.8 degrees per step). multiplied by micro stepping factor of 16 will = 3200

Pitch of the belt for GT2 is 2mm. Multiplied by the number of teeth on motor pulley. We will use 20. So 2 x 20 = 40

Then 3200/40 = 80 steps per mm for a 1.8 degree motor with GT2 belt and a 20 tooth pinion.

For the rotational bed we need to know what gear reduction you are using if any. The step angle of your stepper and the following should tell you what your value would be. If your not running 20:90 gear reduction then my value will not work for you.

The formula CNCModeller provided is.

1.8 Deg/step at the motor x 20 / 90 = 0.4 deg/step at the bed.

1/0.4 = 2.5 steps/Deg at the bed.

1.8 degree is the step angle of the stepper motor multiplied by my gear ratio.

20/90 is my gear ratio. Pinion or drive gear divided by my driven gear which is 90 which comes to 0.4

0.4 is the fraction of a degree at the bed through the gear reduction.

So we want to know how many steps to make a degree with the reduction. So you divide 1 degree by the steps at the bed which is the 2.5 steps to get 1 degree at the bed.

Then you must take the last value and multiply it by your step multiplier mine is 2.5 X 16 = 40. That is what is enter for your Y in the config.

I hope this helps.

And thank you CNCModeller for the original formula.

shadi

This post is deleted!

Phospherus

@shadi

Just tell me how many teeth are on the gear on the stepper motor and how many teeth are on the big gear driven by the stepper motor.

dc42

@phospherus, I don't have a polar printer, but I did some firmware tests in polar mode before I released 3.3beta3 and the motor movements appeared correct to me. Please can you re-test that print using the 3.3beta3 release.

shadi

@phospherus 16 teeth on the small gear and 258 teeth on the large gear, i just want highlight that the drawing is skewed with round edges,

I mean this machine should move Y= Theta, and X= R where are the SQRT (XX + YY)

is that correct ?

shadi

@phospherus 16teeth on the small gear and 258 teeth on the large one