The coefficients of the extruder.

droftarts

Trying to interpret from his previous thread, I actually think this is for a remote drive extruder, powered using a belt. I think a picture or diagram would really help!

Ian

Pro100sanek23

@droftarts
yes, that topic has a direct connection to this question. if it is not forbidden, then I will leave a link to the video where it is clearly demonstrated what I would like to achieve.
https://youtu.be/TQYYHrxIG90
the printer on the video runs on the marlin firmware, I have the same RRF and I would also like to use this extruder on the RRF firmware.

Pro100sanek23

@fcwilt
here you can see the video above. that is, it is not just a fantasy but an already implemented prototype.

Pro100sanek23

@Rushmere3D thanks for the answer. but I had a little different.

Pro100sanek23

@droftarts
my friend who created the prototype from the video is now working on the implementation of this kinematics in the Klipper. a friend contacted the creator of the Klipper and was interested in this idea, but so far there is no implementation of this kinematics in the Klipper. I volunteered to help him implement the RRF in the firmware. if dear @dc42 helps, I will be very happy and grateful. I wrote my vision of how this can be implemented in the topic that you asked. if you give access to the extruder to the m669 team, then the implementation would not be very difficult. I can be wrong to the best of my inexperience.

fcwilt

@Pro100sanek23 said in The coefficients of the extruder.:

@fcwilt
here you can see the video above. that is, it is not just a fantasy but an already implemented prototype.

The video makes it clear what you are asking about.

Thanks.

Frederick

droftarts

@Pro100sanek23 looking at the video, it seems the extruder is working using a Markforge belt layout. This is supported in RRF, but is generally used for adding an additional carriage. See https://duet3d.dozuki.com/Wiki/Gcode#Section_Parameters_for_Cartesian_CoreXY_CoreXZ_CoreXYU_CoreXYUV_MarkForged_kinematics_RRF_2_03_and_later_only

However, I don’t think M669 currently applies to extruder axes. I’ll ask @dc42.

Ian

Pro100sanek23

@droftarts
I agree, we need a 4x4 matrix where the fourth coordinate will be the coefficients of the extruder. as a result, the matrix will look something like this x1:1:0: 1 y1: -1 :0:1 z0:0:1: 0 e0:0:0: 1

droftarts

@Pro100sanek23 reply from @dc42 :

That's not currently supported. We would need to add additional rows in the movement matrix for extruders, and route the extruder motion planning through the kinematics module along with the axes.

So this is quite complicated, and unlikely we would add this soon, though perhaps it helps you to know where to look. The difficulty is that all XY movements need to be limited by how fast the extruder can move, bearing in mind that the steps per mm is different.

It may be simpler to compute the extruder position using a Gcode post-processor, rather than doing it in firmware. Then it would work on stock RRF, Marlin and Klipper firmware.

Ian

zapta

I wonder if some kind of a smart kinematic mixer would work here.

Imaging an electronic unit with three stepper inputs (step, direction) for cartesians channels X, Y, E, and three stepper outputs for the kinematic steppers A, B C, such that pulses the input kinematics are mapped to pulses of output kinematic. Do thinks like this exist?

Somehow related to the electronic gear idea in this video (though it's a single channel) https://www.youtube.com/watch?v=FTs9GygRQ-U&list=PLDlWKv7KIIr90ZZ7Zqt-ge5nVVdS3WVgg