Servo tuning or mechanical? - Page 3
Close
Login to Your Account
Page 3 of 3 FirstFirst 123
Results 41 to 60 of 60
  1. #41
    Join Date
    Dec 2008
    Country
    UNITED STATES
    State/Province
    New York
    Posts
    8,998
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    2355

    Default

    had a big cnc mill the way covers were vibrating jumping quite a bit even coming unhooked occasionally.
    .
    finally they take way covers off to look at them. they had little roller ball bearings on the bottom that locked up or werent turning. amazing how much smoother a machine can get when all the bearings are lubricated and actually turn freely.
    .
    had another cnc mill the servos were vibrating oscillating bad. finally they replace ball screw and nut. that machine now moves very smooth with no oscillation. sure boss didnt want to buy a new ball screw but eventually had too in order to hold tolerances

  2. #42
    Join Date
    Jul 2006
    Location
    Hillsboro, New Hampshire
    Posts
    5,258
    Post Thanks / Like
    Likes (Given)
    1232
    Likes (Received)
    3678

    Default

    Sometimes industrial machinery can be affected by the incoming 3ph power not being "clean", with ripple, spikes, or other voltage anomalies influencing the machinery. 3ph line reactors are basically large "chokes" that clean up incoming power and remove or limit these deviations.

    Someone with more electrical experience can expand on that, but a very basic primer on them can be checked ot here: Line Reactor Manual | What Is A Line Reactor? Reactors suitable for industrial use can be found on ebay, running from under a hundred to ~$250 or so.

    https://www.ebay.com/sch/i.html?_fro...actor&_sacat=0

    Not saying this will solve everything (or anything), but it's worth looking into. Your service tech should be able to hook an oscilloscope and the appropriate probes to incoming power to see if there's questionable spikes on display.

  3. #43
    Join Date
    Feb 2011
    Location
    Boise ID
    Posts
    52
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    1

    Default

    Put the parts on the CMM, looks normalish. Maybe .001 out of normal but still in tolerances.
    yes even 45degs i can feel something.
    It still bothers me that its feels the worst when more than one axis is on the move.
    A guy just suggested to unhook the servo cupler and try it without the ways and ball screws interfering. Seams like a good option but i think the patience for that kind of stuff is already gone.
    If it where my personal machine i would. But i cant afford the other options, and labor is free. Well almost free just the wife yelling to get out of the garage.

  4. #44
    Join Date
    Feb 2011
    Location
    Boise ID
    Posts
    52
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    1

    Default

    I do have an oscilloscope that needs some exercise. But someone would have to walk me through barney style of what to hook to, look for, setting etc.
    I can tell you we brought the power line straight to the machine. It is on its own meter because there was not enough power for the rest of the shop otherwise. That kind of leads back to noise in the lines i suppose.
    I have today and tomorrow then i will be out of the shop for two weeks. don't know whats going to happen in that time.

  5. #45
    Join Date
    Jul 2006
    Location
    Hillsboro, New Hampshire
    Posts
    5,258
    Post Thanks / Like
    Likes (Given)
    1232
    Likes (Received)
    3678

    Default

    Quote Originally Posted by shanej45 View Post
    I do have an oscilloscope that needs some exercise. But someone would have to walk me through barney style of what to hook to, look for, setting etc.
    Won't be me - I don't want to get you killed!

    I can tell you we brought the power line straight to the machine. It is on its own meter because there was not enough power for the rest of the shop otherwise. That kind of leads back to noise in the lines i suppose.
    That's interesting! That's a new data point, so is worth exploring. But if the machine's gone by the time you get back it's moot.

  6. #46
    Join Date
    Feb 2011
    Location
    Boise ID
    Posts
    52
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    1

    Default

    I know enough not to go stabbing at things randomly. But not enough to keep me out of trouble, which leads me back to trying to get this machine running. (HA!)
    We most likely would move a different machine on to the same incoming line. It would be some sort of irony if it had the same problem. Although highly doubt it. Newer machine I imagine would have a lot more in the system to reduce electrical noises.

  7. #47
    Join Date
    Feb 2011
    Location
    Boise ID
    Posts
    52
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    1

    Default

    Seeing is this might be the last day i get to try or check anything. Just do me one favor, O people of practical machinist. If anybody ends up with this machine you HAVE to tell me what the fix was. Its killing me thinking is something like a power supply but no way to diagnose it.
    Until the check is signed i am still going to see if i can get this thing working. Unfortunately any funding to try anything will be gone.
    A few good things is i learned a lot about repairing machines, made a tech friend, and rather enjoy working on this stuff. Much more enjoyable when its actually fixed in the end. Maybe i should have a career shift, rather than program get into machine tool technician.

  8. Likes Ox liked this post
  9. #48
    Join Date
    Feb 2011
    Location
    Boise ID
    Posts
    52
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    1

    Default

    looks like the machine is still here. I can get back to poking and checking. Still a chance to save the machine.
    I have the feeling there is something that a Matsuura tech that has been around for a while would know what the problem is.

  10. #49
    Join Date
    Jan 2005
    Country
    CANADA
    State/Province
    Saskatchewan
    Posts
    9,773
    Post Thanks / Like
    Likes (Given)
    1220
    Likes (Received)
    3397

    Default

    Quote Originally Posted by shanej45 View Post
    looks like the machine is still here. I can get back to poking and checking. Still a chance to save the machine.
    I have the feeling there is something that a Matsuura tech that has been around for a while would know what the problem is.
    Can't sell it if you can't fix it? Generally doesn't work too well.

    If you can fix it, then may as well keep it

  11. #50
    Join Date
    Feb 2011
    Location
    Boise ID
    Posts
    52
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    1

    Default

    Here is a question I should ask. Does anyone know of a tech and or company that can fix this this kind of problem? Especially if they are in the Idaho area.

  12. #51
    Join Date
    Nov 2014
    Country
    UNITED STATES
    State/Province
    Colorado
    Posts
    1,666
    Post Thanks / Like
    Likes (Given)
    2310
    Likes (Received)
    472

    Default

    Quote Originally Posted by shanej45 View Post
    Maybe this will help with some ideas.


    Attachment 228631

    Ok so why does the test sheet say "Circularity" 0.000657"

    6 tenths ? (almost 7).

    I would have thought on that machine circularity of 5 to 8 micron would not be such a challenge. Given that everything else on the ball bar tests shows that the machine has the capability to be surprisingly accurate.

    1 micron ≈ 40 millionths.

    circularity should be of the order of a tenth or better ?


    circularity 'reading" is 6 (times x) less precise than it should be ? 6 times out of tolerance compared to what the rest of the test sheet / ball bar tests seems to indicate what the machine should be vaguely capable of ?


    What am I missing here ?

    _____________________________


    Most of that seems to come from scaling mis-match (between axes) … Also 6 tenths... Am I smoking crack on that ?

    Not mechanical.

    Servo tuning / calibration thing.

    Although rumbling sounds + smooth / uniform scaling error seems puzzling... Does that machine use ball screws ?

    __________________________________________________ ______________________________________________


    I'm wondering knackered / failing / corroded + foreign materials gunking up a ball nut. If the machine has scales then that might explain the rumbling noises and facetted appearance on the part + geometric scaling mis match between axes. I.e. Ball screw takes too much force to "turn" and the servos etc. have to overcompensate in an uneven way to try and join the "dots" in an attempt to "accomplish" each commanded position (but lags in time and maybe overshoots as well). Torque/friction for turning one of those ballscrews is uneven leaving a residual scaling error that's pretty big (by modern standards)... Circularity of 6.5 tenths + facets on parts yet the machine has no reversal spikes to speak of ? Rumble rumble.

    I think you can find the bad axis by cutting a feature with a big radii/ big arc (one arc principal axis of the feature going Up and down (Y axis), curve) and the other large radii feature long axis going side to side X axis (might be worth making separate cuts from low to high and then from high to low (Y axis). … One of those will have more faceted janky appearance than the other . Cuz at 45 degree you can't "Parse" which axis

    Also wondering if a ball screw/nut assembly was replaced more recently ?

    --> Each division on that chart/ plot is a thousandth of an inch not 5 micron
    !
    Last edited by cameraman; 07-26-2018 at 04:24 PM.

  13. #52
    Join Date
    Feb 2011
    Location
    Boise ID
    Posts
    52
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    1

    Default

    The tech did note the scaling miss match and circularity are of concern. More than likely they are related to the facets.
    cant get anything about the previous history Selway lost all there info and the previous owners will no longer help. I would like to see that ball bar test sheet with something finer than a .001 scale. I always wondered why it didn't really show much of these facets. Those facets could be pretty small, but enough to be unacceptable.
    There is a good machine in there somewhere. Takes someone that knows more than I do. Possibly a turnip.

  14. #53
    Join Date
    Jul 2009
    Location
    Peoria, IL
    Posts
    11,174
    Post Thanks / Like
    Likes (Given)
    41
    Likes (Received)
    8455

    Default

    OK. I tried to read through this. Seems like a puzzler.

    Right off the bat, I highly doubt this is an issue with incoming power. It seems like your machine uses the spindle drive to power everything, which is the most common system on newer machines. The 3 phase is rectified to DC power in the spindle drive and that power is used in the spindle drive and the axis drives. That power gets inverted in the drives to power the motors. The drive should be able to monitor the condition of the DC bus and will alarm if it is out of spec.

    My first inclination would be to get an amp clamp and check each phase of each motor while the machine is running a circle or some program with interpolation. See if the phases of each motor are balanced. You can probably check the total axis amps on the control, but it would be more useful to see each phase.

    What I'm driving at is possibly a motor winding issue.

    The drives and motors could all be different depending on the axis load. The Y is probably larger since the machine has no counterweight. But, for the purposes of testing, you should be able to swap them from around. You just don't try to run under a load like that. Most likely the commutation and everything is the same, even though the drives are different in size.


    Do you have a manual that shows how to check and adjust the gibs? That's a reputable brand, they should have a good procedure you can follow.


    It can be tough to troubleshoot servo issues with an oscilloscope, just because of the high voltages involved.

  15. #54
    Join Date
    Feb 2011
    Location
    Boise ID
    Posts
    52
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    1

    Default

    Motor winding could possibly explain some servo mismatch. The mismatch could make the machine struggle to maintain position while more than one servo is on the move. hmmmmmmm
    This rumble is not noticed if only one servo is going. Which makes me think the gibbs are not the source of the problem.
    One of our old Forman's said it is just a servo tuning problem and older machines you just turned a pot usually on the servo drive.
    I cant find the right way to go about that. There is one unidentified pot on the servo drives. It has a little red paint pen on a previous setting and two of the three no longer line up. I'm tempted just to turn it and find out, but know I need more info.

  16. #55
    Join Date
    Jul 2009
    Location
    Peoria, IL
    Posts
    11,174
    Post Thanks / Like
    Likes (Given)
    41
    Likes (Received)
    8455

    Default

    These are AC servos. It's not like the old days of DC brushed motors with gain and overshoot pots. It is possible to tune the drives, but it would have to be done with specialized equipment, and most likely by a Yaskawa factory rep.

    Unlike in the old analog DC systems where brushes would wear and things would drift, your system should not need to be tuned unless something has been damaged.

  17. Likes cameraman liked this post
  18. #56
    Join Date
    Nov 2014
    Country
    UNITED STATES
    State/Province
    Colorado
    Posts
    1,666
    Post Thanks / Like
    Likes (Given)
    2310
    Likes (Received)
    472

    Default

    Quote Originally Posted by shanej45 View Post
    The tech did note the scaling miss match and circularity are of concern. More than likely they are related to the facets.
    cant get anything about the previous history Selway lost all there info and the previous owners will no longer help. I would like to see that ball bar test sheet with something finer than a .001 scale. I always wondered why it didn't really show much of these facets. Those facets could be pretty small, but enough to be unacceptable.
    There is a good machine in there somewhere. Takes someone that knows more than I do. Possibly a turnip.
    I agree with what Elsley said regarding power and phasing.

    However almost an identical set of problems happened to Nashero's (old) mill.

    Facets on part.. Leading to replacement ball screw assembly. + scaling mis match.

    However one thing that can happen is that if a replacement ball screw is not mapped out properly and the old calibration file is used you can get a more chaotic result which would show up on a better ball bar plot. {Summation of waves) these ball screws are not geometrically perfect. But probably would not "rumble".


    Wrong screw to wrong calibration. That's an outlier though.


    I think maybe phasing issue on servos would be more subtle and smooth some how... Not sure you'd get the "rumbling thing " going on... But Ewlsey has 1000 X more experience than I on that.


    I'm wondering if servos have been "Sucked" into smoothing out/ compensating a mechanical issue like 'effed up ball nut ? scaling error 6 tenths/ artifact ?

    Basically a combination of two problems started by a bad ball nut. (Possibly). [Happens to horses a lot that one thing starts going wrong then another injury happens as result of a horse compensating for the first inherent problem. First problem is not obviously injurious but the injury resulting from the "compensation" IS. ].

  19. #57
    Join Date
    Feb 2011
    Location
    Boise ID
    Posts
    52
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    1

    Default

    I still have not located those pesky comp values. I will have to get back to that Monday if I can find the time in the shop.

  20. #58
    Join Date
    Sep 2014
    Country
    NORWAY
    Posts
    20
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    7

    Default

    Have you checked the manuals for that specific servo drive?, should be a mention of your little unknown switch there.
    If you have model nrs, serials etc for them the mfg, should be able to supply you the info also.

    Have you gotten anything checked out of the electrical?

    As ewlsey said checking the motors amp pull, while cutting air, interpolating, while you have your rumble going fx as it should all be equal then, on all the windings and checking the current pull is the same, and within spec of your motor, with a ampere clamp meter. will if nothing else remove a few possible error sources.
    I have no idea how your machine is built up electrically.

    A bad winding on a motor, a bearing or voltage issues can give you your rumble.
    But since it only happens when both are going at the same time, it is not the most likely culprit. as it should be constant on one of them if the windings etc are bad?.

    Your incoming power is likely not the problem also.
    And as mentioned the drives will most likely alarm out on dc bus issues. But checking incoming voltage to all your drives and outgoing voltages under load, would pretty much rule out the electrical side of it. Drives even vintage one are usually good at detecting issues with they're load side, not so much with incoming power issues, unless phase is missing.

    Also being a vintage machine, if you got any contact elements in there like contactors they also get old and worn, a corroded or lightly burnt phase on one will, before complete failure act just like a problematic motor winding in many cases.

  21. #59
    Join Date
    Feb 2011
    Location
    Boise ID
    Posts
    52
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    1

    Default

    I will have to wrangle up a ampere meter clamp. They are not willing to pay someone to come out at this point. Been meaning to get one anyway i suppose.
    If I remember right, all three servos are different. So i should probably get different readings regardless. That is going to make that process more difficult. I could at least see if they are within proper range (hopefully i can find those numbers).
    The servo manuals do not detail that one switch for unknown reasons. That manual does cover a lot of stuff that only the machine tool builder would care about. Well over my head
    Apparently the backlash and lead screw error parameters are not listed in the manuals that i have available. We did obtain the factory parameter settings. Most all values are from the factory, or they make sense as to why they are changed. There is a lot of parameters that are not in our factory parameter print offs, it is probably hiding in there. I wonder if Matsuura would be kind enough to send me one of those handy dandy how to prints. With regard to backlash, lead screw error, or just the parameters in general.

  22. #60
    Join Date
    Feb 2011
    Location
    Boise ID
    Posts
    52
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    1

    Default

    Its officially unplugged and nobody is interested so its going to the scrap yard. We couldn't even give it away (we tried donating it to a school).
    This all really surprises me how it all turned out in the end. We do have another Matsuura with I80 control so we will take out a lot of its electronics and doohickeys.
    I guess I don't have tons of experience with this kind of stuff. But was I wrong in thinking you don't scrap this kind of machine/problem that you get it fixed?


Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  
2