EMC Feb 29 & Stepper speed

["Tim Goldstein" <tgoldste-at-mlh-services.com>]

I loaded the Feb-29 build of EMC and it seems to work great using freqmot. I
think I found the problem that was causing my reduction in speed. When I had
first set up freqmot I had started with the P value real low and kept
raising it until I didn't get any follow errors. I ended up with a value of
200. When I loaded the Feb-25 build I had to edit the generic run file to
get it to work as my previous run and ini files didn't do the trick. Well,
some where along the way I missed changing the P value and they are now set
to 1000 in generic.ini. While this allowed the machine to run it was causing
me to occasionally miss steps at high rates of feed.

I have upgraded my system again and now have EMC on a AMD K6 2D 500 with 64
mb RAM. I tried compiling with a variety of period settings in emcfreqmot.c
and the lowest I can use is 10. I have the axis CYCLE_TIME set to .0003 and
my P is now 150. This seems to give me very solid motion at 60 ipm on X and
Y (Z is a problem, but it is mechanical not EMC) with no backlash. If I set
in .005 backlash I find I get a stall on an axis very occasionally at 60 ipm
and never at 48 ipm. It looks like the backlash is throwing the acceleration
off somehow. This was VERY obvious when I had the P set to 1000.

In spite of everyone's best efforts I am still clueless as to what P,
CYCLE_TIME, and period are really doing. OK, I think I understand period as
determining the base frequency that is generated to issue the stepper
pulses. But, I still don't understand the P and CYCLE_TIME entries. Is it
good to set then to the lowest values possible?


Tim
[Denver, CO]

timg-at-ktmarketing.com <timg-at-ktmarketing.com>
http://www.ktmarketing.com