Camsoft G76 threading

[Karl_T]

Last spring, I came across a difficult threading job that my Camsoft
controlled CHNC was just not able to do. On and <mostly> off over the
last several months I've working on improved threading.

<GLOAT> I finally made perfect threads today.


The attached file is a fully Fanuc compatible one or two line G76
thread cycle for Camsoft Pro. It should run with only minor modification
on most Camsoft lathes.

The macro requires a high speed Opto22 input that fires on a spindle
index mark. I used a slot sensor and disk. Camsoft also sells a
transistor module that amplifies the encoder Z pulse that should
work for most applications.

Karl

[davesnd]

Hi Karl - I am new to this forum and was just looking at your thread routine. I have written a couple of multipass thread routines for camsoft lathes, including one that will cut taper threads like NPT. It looks like my routine is a little simpler that yours - I don't know if that's good or bad? To get away from having to use fudge values (because they will be slightly different for each pitch) you need to make the ratio for x and z the same. This can be accomplished by gearing the encoder to the ballscrew and calculating a gear ratio to get the ratio setting in cncsetup the same for both axes.

Just curious - what is the diameter and pitch of the threads you are using this code for, and what RPM do you thread at?

[HuFlungDung]

I'm not 100% sure, but I don't think the fudge value is used to equalize X and Z, but rather is used to adjust the Camsoft feedrate variable itself.

In real time motion tests, you need to clock the duration of a feedrate move. It must end exactly when it should theoretically end, for example, a feed move of 10" at 10ipm must end exactly in one minute. If the system does not pass this test, then your thread leads will be out of wack.

The galil card that I used did not execute feedrates exactly 'on the clock' without a slight adjustment of the feedrate variable.

[Karl_T]

The fudge factor is due to the RPM not being perfect. camsoft sends a voltage to a VFD on my spindle and the RPM curve vs. voltage isn't perfect on my machine. I couldn't get the TRUERPM command to work well on my machine, I think its for servo drive spindles. I'm sure, with more effort, I could eliminate the fudge value.

The lathe has made 0-80 threads at 2500 RPM and quite a few .5"x20 tpi at 1000 RPM

The real goal of this exersize was to make it 100% compatible with the mastercam POST G76 thread cycle.

Karl