Take some scrap metal and try a more aggressive cut, faster feed, etc.
Check the X-axis ballscrew mounts. Both where the ball nut connects to the saddle as well as the ends of the screw where it's mounted to the table. Preload on the ball screw bearings is also suspect, but that would be the last mechanical thing I'd check.
They may be snug but not reefed down. If they're snug, then they might be stable enough to move the table appropriately while checking backlash but will shift or resonate when a cutting load is applied to the table in the X-axis.
Lost motion only along the X-axis either means the Y-gib is loose, or the X-ballscrew (and or ballnut) is loose. Remember, when you cut in Y direction, it's the X motor & ballscrew assembly that's holding the table steady (along with the Y gibs).
Sounds like you've figured out one of the contributing issues and maybe it's time to check the second.
As other folks have noted, flexing (column, head, too, holder, etc.) won't leave patterns like that. It looks like it's cutting properly with no chatter - only the table is shifting along the X during the cut. Remember earlier in this thread when someone speculated that this was G-code movement, and you responded in the negative?
Those patterns are not chatter or flex. That's 'precision' shifting along the X axis.
And another thing...
If this isn't mechanical, another thing to check is to watch (with a mirror maybe) the X-axis stepper to ballscrew coupling while taking the cut. If it rotates while you're cutting only in Y, then you've got electrical noise or some other source is commanding the X-motor to turn when not desired. Frankly, those patterns are spaced too far apart to be something vibrating.
Dennis
No easy access to X-axis ball screw mount that I could find, I think it would require disassembling stuff from motor side since from opposite side I do not see any bolts. I really don't want to spend more time on this stuff.
Problem is still there, i.e. not solved :-(
I really hate wasting time doing this stuff.
Dennis
The motor side should have the ball screw locating bearings. The opposite side should have a simple roller bearing that lets the screw end slide freely without whipping.
If it's the ball nut, that's a pain to get to. It's under the table and screwed to the saddle.
I am starting to think that problem might not be strictly with ways and ball nut, the pattern is too regular. It creates equally wide stripes up and down in material and on the axis which is not moving while cutting is going on. If X motor is turning when it should not why it would turn in regularly spaced intervals and turn back and forth? Why would loading up the cut reduce the pattern?
Dennis
He should go back and do as I suggested in post #31.
"have you tried roughing like you did and then doing a less DOC finish pass in the opposite direction?
It seems to me like doing this should still see the same pattern if it is in the machine. If you don't see it in the same locations, then I might suspect something else. Do the finish pass a little higher so you can see the original pattern."
Thinking about it more, I am not sure what the results would mean exactly, but might help to narrow down how much of a pattern it actually is.
Lee
I will try that, I have tried doing finish pass in opposite direction and pattern remained. It would be interesting to see though whether marks align.
Yesterday I made couple of parts and pattern is visible even on rough cuts done with Shearhog.
Last edited by coffeetek; 12-21-2017 at 10:28 AM.
Dennis
Hey Dennis. Have you tried running a dial indicator along an edge in Y while doing that same cut in air? (not actually cutting material, but spinning the spindle while simulating that same cut above the stock) If the X motor is picking up noise, the dial indicator will show that.
Just trying to separate electrical control issues from mechanical issues.
Dennis
You were just jogging, correct?
If so, that is probably mechanical. Too much precision for E noise.
Are you certain that Gib is getting lubricated?
Lee
Not jogging, running the G1 command from start of the video...
Yes, I have checked lubrication on gibs and ways.
Dennis
Why not try it in jog mode?
Lee
Just tried jogging, exactly the same up/down pattern.
Dennis
Now that eliminates the model and CAM.
Lee
An obvious observation. Firstly I will assume that your units are mm rather than inches for obvious reasons. Secondly, I count about 16 cycles of movement on the indicator
and that is over a distance of 80 mm. That comes out very close to 5 mm, which in turn is also *very* close to the leadscrew pitch (5.08 mm). Hence, it appears that you have a bent Y-axis leadscrew.
Last thing I wanted to hear, but its very likely you are correct. Fact that even under load it would move predictably would strongly suggest that its lead-screw. However, I did observe that I could do a very heavy cut and get lines to almost disappear... We will see what Tormach says. What would cause Y-axis lead screw to bend though?
Last edited by coffeetek; 12-21-2017 at 07:38 PM.
Dennis
It is either a bent lead screw or the ball nut may be misaligned. Could be loose too, but then I would expect to see slop in the form of backlash from the nut rocking.
Do you know anyone that has one of those little inspection cameras?
I bought one a while back from Harbor freight and it is very cool.
Perfect tool for inspecting things under tables and other tight places. I have used it quite a bit more than I initially thought I would.
https://www.harborfreight.com/digita...era-62359.html
Lee