binary works - compilation not smoothly???

Hi,
ok, seems I am misunderstanding something here….

My idea was to use XYZ+A for simulanous axis stuff (like milling a cylinder along the a axis, where the centerline of the cylinder is not the centerline of the a-axis)

And to use UVW+A for indexed machining (driving to A0 then do operations in uvw, then driving to A90 and so on).

I found other possibilities like switching between rotated coordinates (YZ depending on A) and trival coordinates (YZ not depending on A). See M128. Do I understand correctly that by switching between TCP and trivial coordinates (M128) I am switching the kinematics module between trival coordinates and rotated coordinates, too? How do I setup this???? What happens to the actual position when changing the coordinates?

Other point is the WCS? I think the only way to setup a WCS is to use the center line of the A axis as 0 for Y and Z, right?

I'm feeling free to use XYZ+A+UVW or XYZ+A + M128. Depending what you tell me, what is the best choice. I have to adapt the postprocessor anyhow


Georg

schorsch wrote: Hi,
ok, seems I am misunderstanding something here….

My idea was to use XYZ+A for simulanous axis stuff (like milling a cylinder along the a axis, where the centerline of the cylinder is not the centerline of the a-axis)

And to use UVW+A for indexed machining (driving to A0 then do operations in uvw, then driving to A90 and so on).

I don't think that is even necessary. Why can't you do both with XYZA?

Hi,
ok... to be honest.... please don't laugh… I do not how to handle this in the postprocessor… but maybe you are right and it is easier to spend more time on this than to change the kinematics

Georg

Shit it really took me only 30min to check the pptions of the postprocessor…. damn…


seems I really to not need UVW any longer

Thanks a lot for your hints!
Georg

PS for everyone else who is using Inventor HSM:
optimizeMachineAngles2(0) means TCP mode -> rotated coordinates needed
optimizeMachineAngles2(1) means non-TCP mode -> trivial coordinates needed

Thanks again

Hi, Andy et al,
I managed again to run with 4 axis and 4 joints. And the postprocessor seems to produce good code, too. But finally I have problems to jog the A axis when using the trivial coordinates.

When multiply pressing on the jog button for the A axis very fast this occurs
It happens faster when using higher velocities...

I managed to hang the complete PC with this error. Really full stop. keyboard was blinking, couln't even move the mouse....

I do not have the same problem on any other axis


Is this a latency problem? I never had problems here

Thanks in advance
Georg

Following error is likely to be that the A axis simply can't move as fast as commanded.

But it really shouldn't hang the machine. The whole point of LinuxCNC os to be rock-solid

Hi, thanks a lot for your quick answer,

can't move that fast because of what?
I get drive it to much higher speeds, but when Pressing the button multiple times quickly after each other (like double-click speed) it crashed?? Which parameter do I have to change?


I do not understand this, as I can drive even higher stepper speeds at the linear axis without problems

Schorsch

schorsch wrote: can't move that fast because of what?

Some limit or other....

What step rate are you commanding? How does that compare to the actual step-pulse + step-space time?

What is the velocity limit on the stepgen? Accel limit? How does this compare to the axis limit?

Great, seems to be fixed.

MAX_VELOCITY was larger than STEPGEN_MAXVEL, same with acceleration

couln't reproduce the error anymore, since I changed!

Thanks a lot

Georg

Hm....
Unfortunately, I'm running into a similar error when running in automatic mode

Most important stuff of my configuration are:

hm... I'm just realizing that I can run higher speeds in MDI than in the manual mode....why? Here it si crashing..maybe because of too high speed.... why it is like this...where can I set the total general limit?

thank you for any hint

Georg