I’m pretty sure the ball screw on my recently purchased 770 is bad. When using Pathpilot with .001" steps, I don’t get any table movement until about .010".
I have a hardcopy procedure for checking the cause of motion lost (I don’t recall where or who provided it to me). First step is checking the angular bears. The angular bears were recently replaced and they only have a few .0001" of movement. The second is to check the ball nut and ball screw. The procedure says to move the table move shift the table by hand, but I can’t get it to budge. The third is check the gibs. I’ve adjusted the Y gibs until it almost locks up the table if I use too high of Rapids and the .010" still remains.
If I put a dial gauge on the ball nut housing and step PathPilot by .001" , from .001 - .009 I get a few .0001 of movement, until I get to .010" via PathPilot then I get movement of .001 for .001.
Thoughts?
In a situation with a complete lack of motion you will see a bit of lag on the indicator. When the machine is in full motion you will find that the lag is non existent. Use the machine. Cut a few interpolated circles and then measure the circles. I can understand wanting things perfect but it is a motion system not a static system.
Sounds like thrust bearings may need adjustment/replacement.
You mentioned it’s a new mill, so I’m assuming not used. In that case, it sounds like backlash compensation might help. There’s a couple videos on YT for Tormach Pathpilot Backlash Compensation. I watched those videos and put in a few tenths of compensation and was able to get 0.001 movement for 0.001 change in Pathpilot for my X and Y.
Alternatively, NYC CNC has a video on YT on how to adjust the gibs and angular contact bearings on a Tormach scientifically.
Hope this helps.
@SETH_arseneau Do you mean the angular bearings?
@Joe_Thompson This is a used mill that was in storage in reasonable shape. I replaced the angular bearings on both the X & Y axis. The X backlash is .001" The Y is unchanged after the bearing replacement. When adjusting the Y angular bearings the torque required to turn the ballscrew was much great than the X ballscrew. I attributed the Y ballscrew hand rotational force to the additional weight of the Y and X saddle and bed assembly, is this a correct assumption?.
There shouldn’t be much difference in the resistance between the X and Y. If it was in storage for a long time, you may want to pull the gibs and make sure they’re cleaned well. With .010" backlash, you’d be able to see movement at the angular contact bearings. It sounds like you got those done properly. Perhaps the ball screw nut itself is not secured properly.
@SETH_arseneau I checked for possible movement of the Y ball nut carrier. I put a dial indicator on the Y on ball nut carrier and I’m seeing the same ~.009 lost motion. This is why I suspect the Y Ball nut, but I’m not sure how much the gib and angular bearing adjustment when measuring on the ball nut carrier. I think this weekend I just might remove the Y ball screw and see how the table moves, inspect the ball nut. I just want to be sure before buy a new one.
Marty,
Just to make sure, the lost motion testing corresponds to this page, correct?
https://tormach.atlassian.net/wiki/spaces/IMK/pages/3010330636/Lost+Motion+Testing
If so, the two things I would recommend checking are the bolts holding the ballnut to the ballnut casting and the bolts holding the ballnut casting to the saddle casting are tight. If they are though, then yes, your lost motion is coming from the ballscrew itself.
The gibs and angular contact bearings are not components of the movement you are seeing. The lost motion of the angular contact bearings was isolated from Test 1 and any effect from the gibs is isolated from the tests (unless they are too tight causing stiction, not the case here).
Thank you,
Norman
1 Like
@nkowalczyk Thanks for providing the link. I only have a hardcopy of this procedure and I couldn’t remember where I got it from.
I checked the bolts holding the ballnut to the ballnut casting and the bolts holding the ballnut casting to the saddle casting. All of these are tight. So I started on removing the ball screw assembly.
Once I had the motor, angular bearings, bearing cap and everything else, I expect I should be able to move the table by hand, but I couldn’t. I loosened the gib a good full turn and tapped the center end to move it out. Still no table movement. I unscrewed the front gib adjuster out by as least 6 -10 turns out, pushed the gib forward again. Then the table was really easy to move with little effort.
I think the gib was way too tight. The there was torsion twist in the ballscrew shaft (.010") and once the Y motor provide sufficient torque to overcome the gib, the table moved. This also explains why I was seeing a few tens or thousands of movement leading up the movement of the table.
Does sound right? I’m going to put the everything back together with the gib loose and see what happens.
1 Like
Adjusting the Y gib resolved the issue. Now I have .002” of lost motion on the y and x axis. Good to go.
1 Like
Thank you for sharing such a detailed explanation and resolution of this issue. There’s no doubt it will be helpful to someone in the future. This is a great community for these cool machines!
1 Like