I have a removable CNC'd 4th axis rotary table for a standard 3 axis mill. I normally run EMC2 and the 4th axis gcode is commonly generated with deskproto (which is great btw). Anyway in the past, I kinda 'eyeball' it in order to zero the part properly. Results are great. Never really going for ultraprecision anyway.
Im wondering what is the proper procedure you recommend for adjusting the axis? For example, imagine there is a cylindrical extrusion in a chuck that was faced off with a lathe and then the outer diameter had a pass run over also. The cylindrical stock also never leaves the jaw chuck, and the jaw chuck is merely transferred and assembled onto the 4th axis on the mill, taken directly from the lathe.
There are 4 axis now, X Y Z A. Normally as we all know, only 3 axis are used, X Z A. I normally program the gcode so the top of the part is the zero although you can make it the center of the stock also. Anyway, all the same how does one precision zero the tool relative to the cylindrical part?
First one needs to make sure the non-moving axis, Y in this case, is set properly so that the cutting tool axis is directly intersecting the axis of the rotary table A. In other words, the tool should be aligned in Y so that it touches the top of the apex of the cylindrical extrusion. In the past I eyeball this. I draw a straight line down the center of the cylinder, rotate it on the 4th axis so it points upwards on the mill along the Z direction, and move the y until the tool look parallel to the line. This is eyeball work.
I'm thinking maybe a more extensive way is to get an indicator somehow perfectly aligned on the spindle axis, and then have the indicator measuring on the cylindrical stock and simply jog the y and zero it when the indicator shows the highest point indicating the apex, and thus center for Y.
The X axis should indicate the very edge of the part in most of my case, although a slight offset wouldnt be a deal breaker since no features will be missed but it may leave extra stock material on that end that is not cut. I eyeball this now too.
I eyeball this one too. The zero i normally program to the outer diameter of the stock. So I always lathe it first, measure and then adjust the program to indicate stock diameter then I cut down to the part. Ideally the Z should just touch the outer diameter of the stock. I guess I can do the paper shim method also.
Doesnt matter since its often cylindrical stock. Zero blindly.
So what is the proper way of doing this?
Another more elaborate way may involve some permament datum fixture that i always measure and zero from that has a known offset to the rotary table. Problem is the Axis gets moved around a lot and cannot assure repeatability and also duno if this is a good method so I ll let you tell me if that is how it should be done.