Zum Inhalt springen

CNC-Fräse: Unterschied zwischen den Versionen

Bkubicek (Diskussion | Beiträge)
Bkubicek (Diskussion | Beiträge)
Zeile 2: Zeile 2:
A CNC mill like the dear Geil-O-Mat has three axis that can be moved independently. A spindle with a [http://de.wikipedia.org/wiki/Fr%C3%A4swerkzeug mill cutter] typically removes material.
A CNC mill like the dear Geil-O-Mat has three axis that can be moved independently. A spindle with a [http://de.wikipedia.org/wiki/Fr%C3%A4swerkzeug mill cutter] typically removes material.


Each axis is driven by one(Y and Z) or two (X) stepper motors. Basically, theses motors can only rotate in 1.8 degree steps, and hence no secondary encoder is needed for the machine knowing its current location. By a trick called "microstepping", currently the resolution is increased to 1/8 of 1.8 degrees.
Each axis is driven by one (Y and Z) or two (X) stepper motors. Basically, theses motors can only rotate in 1.8 degree steps, and hence no secondary encoder is needed for the machine knowing its current location. By a trick called "microstepping", currently the resolution is increased to 1/8 of 1.8 degrees.
As the stepper leads in to a "Zwillings-Trapezgewindespindel", the rotation is transformed into linear motion, one revolution= 6mm.
As the stepper leads in to a "Zwillings-Trapezgewindespindel", the rotation is transformed into linear motion, one revolution= 6mm.
Warning: If there is too much force for the motor to move one step, it skips the step, typically failing also in subsequent movements, resulting in an ugly noise, and shift in the positioning.
Warning: If there is too much force for the motor to move one step, it skips the step, typically failing also in subsequent movements, resulting in an ugly noise, and shift in the positioning.


The Motors are connected to a driver device, which sits on top of the whole machine. It creates the strong currents for the stepper drivers, out of signals from the PCs parallel port.
The motors are connected to a driver device, which sits on top of the whole machine. It creates the strong currents for the stepper drivers, out of signals from the PCs parallel port.


Hence, the computer has to send signals telling which axis should move one step forward or backward at a given instance. This can/is done by a software called "EMC". It is open source, and the thread doing this parallel port communications is using the patched real time Linux-kernel.
Hence, the computer has to send signals telling which axis should move one step forward or backward at a given instance. This can/is done by a software called "EMC". It is open source, and the thread doing this parallel port communications is using the patched real time Linux-kernel.
Zeile 28: Zeile 28:
If you have a 2d-cad file (e.g. dxf) and specify depths for some areas that should be milled away, one speaks of 2.5D-CAM. This is performed e.g. by CamBam, or Camexpert.
If you have a 2d-cad file (e.g. dxf) and specify depths for some areas that should be milled away, one speaks of 2.5D-CAM. This is performed e.g. by CamBam, or Camexpert.


The biggest trouble is the radius compensation of the milling heads. You always have to cut on a path half a diameter outside of the actual position.  For this, there are various workflows. In Gcode, one can tell specify if one wants to cut left, right or directly on the actual path. This is called cutter radius compensations. Its a pain in the ass. Hence, often one uses no such thing, but either draws directly the offsetted lines in CAD, or has some software like cambam, which does this compensation and outputs already compensated "paths". The drawback of both methods is that one can not change the diameter of the cutter without recreating the gcode.  
The biggest trouble is the '''radius compensation''' of the milling heads. You always have to cut on a path half a diameter outside of the actual position.  For this, there are various workflows. In Gcode, one can tell specify if one wants to cut left, right or directly on the actual path. This is called cutter radius compensations. Its a pain in the ass. Hence, often one uses no such thing, but either draws directly the offsetted lines in CAD, or has some software like cambam, which does this compensation and outputs already compensated "paths". The drawback of both methods is that one can not change the diameter of the cutter without recreating the gcode.  


In the Gcode compensation, a "tool table" is used. Each cutter is index by a number, and in the table the diameters and other things are specified.
In the Gcode compensation, a "tool table" is used. Each cutter is index by a number, and in the table the diameters and other things are specified.
The current machine's state  is given by a coordinate triple, and some states (emergency swich, spindle speed, ...). The machine '''coordinate system''' is defined by the now automated homing routine. Thereafter, one can move the machine to a location, and "touch off", giving explicit coordinates for this location. Thereby the working coordinate system is defined. This is also the system displayed in Axis, However, in the background, the machine still works in the home-coordinate system for checking its movement limits. Clever!


== Current Status==
== Current Status==