What's new
What's new

Servo tuning or mechanical?

shanej45

Aluminum
Joined
Feb 10, 2011
Location
Boise ID
My shop has a new to us Matsuura MAM-600hf-pc2 1997 that has had loads of issue. We finally have it up and running but the finish is unacceptable. One of the issues was we had the spindle rebuilt. Another was the 4th axis indexer servo drive was toast, and the 4 axis was out of alignment. That all appears good now.
We ran a part/program that is well know to us to see if it was acceptable, and it is not. There appears to be a faceted appearance to the part.
especially on parts that are interpolated or X and Y movement like a 45.
Changed feed and speeds to see if harmonics was an issue and saw no real difference.
I am attaching a picture that really shows the 45 being odd. One part is from our normal machine the other off of the "new" machine. Hard to see in the photo but there are a lot of facets around the pivot hole area leading to the 45.
wonder if there is a servo tuning issue going on or if there is something else going on.


thumbnail.jpg
facets.jpg
Thank you
 
My shop has a new to us Matsuura MAM-600hf-pc2 1997 that has had loads of issue. We finally have it up and running but the finish is unacceptable. One of the issues was we had the spindle rebuilt. Another was the 4th axis indexer servo drive was toast, and the 4 axis was out of alignment. That all appears good now.
We ran a part/program that is well know to us to see if it was acceptable, and it is not. There appears to be a faceted appearance to the part.
especially on parts that are interpolated or X and Y movement like a 45.
Changed feed and speeds to see if harmonics was an issue and saw no real difference.
I am attaching a picture that really shows the 45 being odd. One part is from our normal machine the other off of the "new" machine. Hard to see in the photo but there are a lot of facets around the pivot hole area leading to the 45.
wonder if there is a servo tuning issue going on or if there is something else going on.


View attachment 228625
View attachment 228626
Thank you
.
many machines have a servo diagnostic screen it will show servo moving. if machine at a stop and you see it go +10 to -10 to +10 its moving. often if indicator on a part in handle mode you move .001 and indicator will show it still going back and forth often for many seconds
 
I have seen that screen on a lot of other machine but cannot locate it on my Yasnac I80. It is a 1995, earlier i said 97.
 
Before I rebuilt my Haas spindle, the surface finish it produced was similar to what you're getting. The bearing preload was lost on my spindle and a new set of properly adjusted bearings improved it immensely.
 
Yes, check the spindle, your ballbar chart looks pretty good (IMO) when considering the age of the machine.

Maybe try some additonal test cuts with a few different sizes of cutter, and within cutter size testing try varying feeds and speeds to see if you get to a "sweet spot". Just round test paths should be enough to figure out what's going on.

You can test the spindle with indicators and prying, as was discussed in this thread (and others): http://www.practicalmachinist.com/vb/haas/haas-vf2-surface-finishing-issue-350313/
 
Before we ever made a cut we had the spindle rebuilt. It had almost no draw bar tension when it arrived.
With that being said. After install it still sounded way too noisy for a Matsuura. I have the suspicion something is still going on. Could the spindle motor be toast? bad rebuild? Are there simple tests for this stuff?
we did put an indicator from the spindle onto the tombstone and shove the spindle up and down (it is a horizontal, so y axis).
watched the needle jump and stay .0005 could push it back as well. Put the indicators on different parts of the y axis ball screw shoved the spindle the same with no real movement. Could it be a Gibbs adjustment/cleaning?
We are trying to avoid trial and error. Replacing one thing after another until its found (too expensive).
We are about at the point that getting rid of the machine is up on the table. (It has been about 9 months of this now)
It was a really nice machine. Hate to get ride of it for something simple.
 
Applying test pressures to the spindle should demonstrate perfect elastic rebound. Make sure that your indicator base is attached to the spindle housing so that you are not bringing the backlash of other components into the same test.

I'm not a spindle repair specialist, even though I have replaced bearings on several of my own spindles with 100% satisfactory result. It is possible that something could go wrong, so just saying it's rebuilt does not remove all doubt that it is good. Can the rebuilder perform his own tests to prove the quality of the work? Was the taper reground in place? I have my doubts that a regrind would turn out well if there was something wrong with the bearing preload.
 
Minor point - Z axis is always on axis with the spindle, so by definition you're dealing with Z. Setting up multiple indicators off surfaces that won't (or shouldn't) move and then prying on bits of the spindle and housing will hopefully show if it's a gibs issue, or bad spindle rebuild, or whatever.

If you've got a quality pin gage, like a 3/4" Deltronic, you can place that in the spindle and do low speed spin tests, the only deviation you should get is one slow cycle of error due to eccentricity (Deltronic pins being round and cylindrical far beyond a typical spindle capacity), if you see excess movement or more than one complete deviation, or if the deviation high point moves with each rotation, then you've got a good case that the spindle was rebuilt incorrectly.
 
Placed an indicator from pallet base to spindle taper run out was less than .0002 Lifted up and down on the spindle (listed as the y axis) less than .0001. moved the indicator base to the spindle housing with the arm extended 12ish inches to the tombstone. lifted the the spindle again got about .002 and it would bounce back to about .0015 it stayed away from original start about .0005+.
The spindle does vibrate far more than i think it should. I am still leaning towards bad bearing in the motor. Possibly a through the tool coolant coupler
on the back that has not been in operation for years . They said it leaked, and they didn't ever use it anyway.

y axis.jpg
 
Your first pic looks like tool/toolholder issue.

But the second pic looks like a jumpy axis.


Is this a box way machine?



---------------------------

Think Snow Eh!
Ox
 
I have seen similar faceting caused by the code itself. I'm sure you've already checked this, but make sure both machines have similar movements - no pauses involved and the actual feedrate looks similar. You may try looking into the machines corner rounding settings and making sure they are the same.
 
yes, box way machine. Was in storage for a few years we left it in a warm up program for a few hours, did the test twice, on separate days also. way lube all looks good.
looks like all tools have similar problems. That job has about 15 tools. (some are drills so count them out.) from .25- .5, choked up, feed rates up/down etc.
yes the exact same master cam program the difference being post processor. But the code itself is near identical nothing that would cause what i am seeing.
 
Is the spindle belt driven, possibly using a multi-vee belt?
Our cnc machine with very low hours suddenly sounded like a bad bearing in the motor. Opened the cover and saw a multi-vee belt and my mind immediately jumped to a service call I did 30 years ago on a tire tread grinder with a multi-vee belt. Bearings had just been replaced and sounded awful. Problem was parallel misalignment of the sheaves. Found the exact same problem on our mill!
 
no belts. You can basically walk into the back side where the motor and through the tool coupler are located. We would have removed the coupler along time ago, but we just don't know very much about it and opted not to touch it.
 
The setting I'm talking about is machine side. On a HAAS, its set by settings 85 and 191 or using G187. It controls how close to the point the machine has to get before trying to move to the next point. If this setting is too small, you get jerky motion resulting in facets. Too big and you can have accuracy issues. It can be diagnosed by seeing if your facets directly line up with your code.

This is admittedly a long shot, but is something that could cause your facets. Also something "free" to check on and fix. Running the same code, your working machine could be set looser, hiding the facets, where the new machine is set tight, showing them.
 
In a lot of areas on the part it is a single line of code. But i see lots of facets. That is why I am thinking machine related problems.
I Tried looking for something in the parameters compared them with what came from the factory. Nothing i could find. They tend to keep some parameters to themselves. The ones i have access to, look good.
 
If its faceting on straight lines, its definitely not a smoothing issue. Hard to tell if those surfaces in the pictures were straight or gently curved though, and gently curved would be a prime candidate for bad smoothing to facet.

Still, didn't take long to rule it out. Guess you've got a more expensive problem on your hands.
 
If it is a one axis line it looks ok. If it uses 2 axis straight or interpolated it looks faceted. kind of why in the beginning was thinking it might be a servo tuning thing. servo mismatch was .05ms Not sure if that is good or bad.
 
If it is a one axis line it looks ok. If it uses 2 axis straight or interpolated it looks faceted.

"Either" one axis?

I believe that you mentioned a one axis X move.
Doo you have any one axis Y moves?


-----------------

Think Snow Eh!
Ox
 








 
Back
Top