Does anyone know if there is a solution for Marlin and a remote driven extruder together with a CoreXY system?
I’ve seen solutions with a traditional cartesian XY system and a remote driven extruder, but never with a CoreXY system.
Take a look at reprap-xy on github
Thanks
But please correct me if im wrong, but this seems to be just a standard CoreXY with bowden and/or direct drive extruder.
I cant see options for a remote driven extruder.
Ah thought you meant a bowden feed when you said remote. What is a remote driven extruder?
On a remote driven extruder the motor is fixed either on the frame or on the y axis. The torque for the extruder will be transferred via belt or square profiled extrusion to the x carriage.
This way you’ll have a direct driven extruder without the need to move the mass of the motor. At least if the motor for the extruder is fixed on the frame.
Something like this for example :
http://www.thingiverse.com/thing:398324
But I would like to fix the motor on the frame with a corexy and guide the belt to the xy carriage.
This makes it necessary to compensate the movement of the carriage from the extruder steps.
@Nicholas_Seward suggested an approach in this post some time back.
https://plus.google.com/+NicholasSeward/posts/iwYWRu4oZBr
The problem with the way I proposed is that extruding will be the same as pushing on the side of the effector. This will cause the effector to deflect based on how hard you are extruding. You can minimize that with larger pulleys on the effector but it will be there and I imagine with long string/belt runs that are present in the CoreXY that you will have artifacts.
The square drive will have a similar problem. The square drive work well when the force it experience is in the z opposing the screws. In this case the force will be opposed by string/belts.
There is the flexible dremel-like shaft approach. However, some math shows that the angular flex in that shaft is about what you would expect with a bowden.
Bottomline: we have solutions but you have to pay for them.
@Richard_Earl
Thanks, this concept is exactly what I had in my brain 
@Nicholas_Seward
Hm, now that I’ve seen your port and thought about it I also think you could be right.
But what would be if I find a way to compensate this behaviour?
Would you change the firmware to test it? 
To be honest, hardware is no problem for me, but coding is
Can’t you control deflection by just moving the point where the belt/cable applies force as close to in between the bearings/bushings as possible? A simple redesign could put it in between the linear rods. Just don’t give it a lever arm.
@Matt_Maier . Even if the line of action is dead center the force will be bigger on one side of the effector vs the other side. This has to be true or you can’t apply force to the filament. This delta F will cause a deflection in the direction of the force. That isn’t really in question. The real question is how big will this deflection be. It might be imperceptible in a well engineered system.
@Sven_Eric_Nielsen If you build it I will write you a gcode preprocessor for testing. Using standard CoreXY firmware will only require changing the E values of the gcode. The E max velocity, acceleration, and jerk values will have to be set correctly in firmware.
@Nicholas_Seward hmmm, okay, there does have to be a torque to push the filament. A counter torque would neutralize the force. Bowden tubes don’t cause the print head to deflect because they push and pull equally.
@Matt_Maier Yes. I played with several solutions to providing counter torque and the only one that worked in my head was to have two separately driven belt loops. The filament would be pushed from both sides. Obviously, this adds a level of complexity. Could be worth it.
@Nicholas_Seward
The concept of my actual printer project is almost finished and it’s probably not very easy to implement this system in this printer without changing almost everything. But I think I’ll work on this in a side project as soon as my printer works.
But I would like to show an idea that is in my head. Can I use your picture of the CorEXY as a comparison to make it more clear?
@Sven_Eric_Nielsen Of course.