[cncvern]

I'm a newbie, so here we go! I put in a reground Y-axis ballscrew, everything went well. The Renishaw laser raw data said we needed 40 microns of comping (very linearly) in the 508mm (20") of travel. We put the numbers into the U10 control and we got down to 8-9 microns. I would have thought it would have been within 3 microns or less. Every time we relasered and adjusted comps different results were attained. Sometimes we would put large comp numbers in a position and the laser graph shows no change in that area but somewhere else 2-3 spots would move. We finally stopped after probably 50 lasers or more. The best we could do is 3 1/2 microns. Can someone tell us how the comps work in this U10 control and any tricks to putting in the values? Thanks

[Algirdas]

you stuck into control. First you need to make sure if mechanical condition is correct. No control trick to compensate mechanical chaos.
is tension of ball screw correct?
make several tests with time interval without any error compensation. Put all data into table and see if there is repeatability.

[cncvern]

Thanks for the reply! We ran 3 tests over 2 days with no comp in control and all 3 tests showed the same result, 40-42 microns in a linear pattern over the 508mm of travel. After each raw test we put in the laser comps from the Renishaw program and each time we ended up with about 8-9 microns left to comp in a less but somewhat linear pattern. The ballscrew was put in with .0032 stretch which was the recommendation from our Okuma service rep. Backlash result from laser switched on and off from 0 microns to 1 micron which was the same result I saw when I did a ballbar test on raw comped screw.

[Algirdas]

strange. Please, check your parameter settings. Maybe there is "in position" to high?
You don't use manual mode for accurate positioning I guess?

[OkumaWiz]

You have 120 "points" possible that you can use. Try to spread them out evenly over your 20" travel by setting pitch comp distance between points to work out to equal your 20". The more points you use, the more accurate you can be. going out to 21" would be better than stopping at 19" since your last inch of travel will not be compensated.

You may not see a "spike" that you put in intentionally because the points will average the numbers between your set points in order to cover every possible programmable position. This will also be a problem if you run out of points before you reach your 20" of travel since it will try to average back to zero comp...

I have seen unusual patterns when the ball screw nut is damaged because the balls are not tight and it takes a few revolutions in order to regroup the loose balls to all be pushing together again, but I don't think this is your problem since you say you have linear test results with it off and you said it was rebuilt. This also only seemed to show up with the axis reversal. This was similar to the backlash, but would suddenly change after about 4" of travel in the reverse direction. (Distance depends on nut damage)

Other tips = perform and set in metric if you aren't already doing that.
Make sure that the machine is anchored well. I had one creeping across the floor with every rapid move to the next laser point.
Make sure to compensate for any temp variations...sunshine hitting the machine will ruin your game in a hurry.
Check runout and straightness of ball screw. Typically this will show up as a waveform in your laser points equal to the pitch of the ball screw.

Best regards,

[cncvern]

Algirdas, we normally use auto mode for accurate position moves. I will check parameters but we bought machine new (10 years ago) and I can't think of why "in position" would have changed.

[cncvern]

OkumaWiz, thanks for joining! We did switch machine over to metric for laser tests. No sunshine or wind factors present. Temp is maintained around 73-74 deg F.
The 120 point scenario sounds like a great idea, I saw where this was mentioned in the manual this afternoon. ONE BIG PROBLEM- THE LASER AND OPERATOR WERE CONTRACTED FROM AN OUTSIDE SOURCE AND HAVE SINCE LEFT. We needed to get the machine up and running to get parts out, so we are running on last laser test comps we put in. I am pursueing this matter for the next laser test which hopefully will go better.
We can only get 508mm (20") of travel in Y, the hard limits are set just outside this range.
The rapid moves between test points are done at a fairly slow speed so machine shifting shouldn't be an issue. BUT--> The machine is sitting on a floating concrete pad 36" thick to soften vibrations from other machines. The laser was sitting on same pad when tests were ran. We do know pad does move when X or Y is moved to different locations due to moving the center point of weight.
I will try and recheck screw runout ASAP but machine is HOT at this time.
Thanks

[OkumaWiz]

For your next laser run, set your "point pitch" to 4.25mm. This will make your 120 points calculate out to 510mm which is just slightly more than you can travel. That way you will compensate over the entire travel distance. Even though you can't reach the last point, put it in as if you could in order to keep compensation more linear.

One other suggestion, if possible, mount your laser on the machine rather than on the floor. That should eliminate one more variable.

Best regards,

[cncvern]

Thanks OkumaWiz! --->I will do that.
After much discussion we decided to let machine run as is for a little while to let every thing from the ballscrew rebuild settle in. Then we will relaser the axis to finish the job. Inbetween time I will run some tests with our ballbar to verify nothing is drastically wrong. Looking at our final laser test, the machine is within specs, but not by a whole lot. I did do a comparison with some of our other verticle mills and the other mills have much better laser graphs and I just want this machine to be in the same chart area.
Thanks for all of the good advice. CNCVERN

[kawarta]

please also check the backlash parameter this should be on 0.010mm or less