John, There are two pictures on www.mpm1.com:8080/cinci one is the positive capture - the one that jumps - positive-jumps.png
one is both the positive and negative capture - pos-neg.png the positive on the left and the negative on the right I moved the axis .0001 positive and then .0001 negative they both look the same it must be the machine - SCR amp? I will try it on the Y and Z axes later. thanks Stuart On Wed, Oct 8, 2008 at 11:12 AM, John Kasunich <[EMAIL PROTECTED]> wrote: > Stuart Stevenson wrote: >> Gentlemen, >> I have the lead screw compensation working - finally. >> It seems as if the compensation is applied differently depending on >> which side of the nominal value the machine is on. >> >> I have comp in my machine as follows: >> >> in the X axis comp file I have >> >> -0.5 -0.502 -0.4972 >> 0.0 0.000 0.0043 >> 0.5 0.5005 0.5052 >> >> all the rest of the numbers from -61.5 to 61.5 have the triplets equal >> - no compensation >> >> I am working around the point X zero >> >> there is about .0042/.0046 backlash in the X axis ballscrew. I have >> been told by a longtime maintenance man that a ballscrew this size >> will be built to have .003/.004 backlash. I don't like it but that is >> another story. Maybe my experience of rebuilding a ballscrew of this >> size was not the dismal failure I thought it was. >> >> set the machine to jog the X axis .0001 or .001 >> position the X axis to -.0002 or more negative >> using the + or - buttons in AXIS or a handwheel - jog the axis >> positive or negative one unit of motion >> the axis motion is smooth and accurate - very nice >> >> set the machine to jog the X axis .0001 or .001 >> position the X axis to -.0001 or more positive >> per above - jog the axis in the negative direction one unit and then >> one unit positive >> >> the compensation is immediate and harsh - the table is moved (bounced) >> farther than the unit - jogging the axis farther positive reveals the >> ballscrew catches up with the table and the compensation is accurate. >> On axis reversal you can feel the 'hit' of the compensation. >> >> per above - jog the axis in the positive direction one or more units >> and then one unit negative - the table is moved smooth and accurate to >> the indicated position - no 'hit' or impulse is felt during axis >> reversal in the negative direction. >> >> when the X axis is -.0001 or more positive - oscillating the jog one >> unit will reveal a 'hit' for every jog (axis reversal) positive and >> smooth motion for every jog (axis reversal) negative. >> >> the compensation magnitude seems to be accurate for all motion >> directions and positions >> >> the compensation seems to implement the tuning for all negative jog motions >> >> the compensation seems to be immediate and harsh (not per the tuning >> of the axis) for positive jog motions on the positive side of the >> nominal compensation location >> >> further testing reveals the exact same results when using MDI >> >> I will add another triplet on each side (-1.000 and 1.000) and check >> the motion at -.500 and .500 >> thanks >> Stuart >> > > This calls for halscope. > > There are three items of interest: > > axis.0.backlash-corr : this is the total amount of correction to be > applied. It will instantly change when you reverse direction (if the > forward and reverse comp values differ, which they do in your case). > > axis.0.backlash-file : this is the above value, after being subjected to > velocity and acceleration limits. This is the value that is actually > added to the position and goes to the motor. It should never jump > abruptly, if it does it points to a bug in the accel limiting code. > > axis.0.backlash-vel : I believe this is an internal variable of the > accel limiting code, and would be usefull for debugging. > > Since axis.0.backlash-corr should jump abruptly on a direction reversal, > it is probably the easiest thing to trigger on. > > If you can capture scope screenshots of a "good" and "bad" direction > reversal, and post them maybe we can figure out what is going on. > > Regards, > > John Kasunich > > ------------------------------------------------------------------------- > This SF.Net email is sponsored by the Moblin Your Move Developer's challenge > Build the coolest Linux based applications with Moblin SDK & win great prizes > Grand prize is a trip for two to an Open Source event anywhere in the world > http://moblin-contest.org/redirect.php?banner_id=100&url=/ > _______________________________________________ > Emc-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/emc-users > > ------------------------------------------------------------------------- This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/ _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
