>Thanks for the help on my zero motiion homing problem.  I am using 
>Jeff's workaround and all is working better than expected.

My current project is converting a Derbyshire watchmaker's lathe to 
CNC by using a couple of tiny (1 inch diameter by 4 inch long brush 
servos on the cross slide (x/y) and a CMC servo (2 inch dia by 5 inch 
long) driving the headstock The lathe is a traditional watchmaker 
type lathe with type D or 10 mm collet holding headstock.  I'm using 
Gecko servo drivers (step and direction from EMC2)

I mounted the tiny servos on the removable cross slide and was going 
to use it just to do profiling with no control over the spindle when 
I got the idea that if I used a servo for the headstock drive (with 
positve drive through an XL 5 pitch belt drive)
that I could set it up as an xyza configuration and do threading.

That's why I came up with the question about zeroing out the 
accumulated rotary motion on the A axis for doing multiple pass threading.

Well, tonight I got it all working and it is really neat to watch it 
do 10 passes threading a one inch 40 pitch thread and to have the 
thread start and stop at precisely the same point on each pass.

I am getting a following error on the rotary axis quite often and I'm 
having a problem getting rid of it.

I know that the number of steps per second is scale times velocity. 
In my case 13.3333 steps/degree x V where V is the number of degrees 
per second that I want to run the A axis.  I also am pretty sure that 
a following error occurs when the step generator can't create steps 
fast enough to drive the axis at the desired velocity.

What I am not sure about is how I take this information and my CPU 
speed or bus speed (whichever) and come up with the maximum practical 
speed to drive any axis

I recently saw a really nice discussion of how to calculate base 
period, servo period, task period, etc. based on cpu speed and 
maximum steps/sec required. I'm pretty sure I could fix the problem 
by trial and error but I would rather have the technical basis for 
what is going on.

I spent most of the evening trying to find the methodology but 
can't.  If someone could point me to it I promise I will add it to my 
"Must remember" file and will never bother you again.

Thanks,
Cecil 


-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Reply via email to