Rotary Encoder for DRO? anybody done it.
Close
Login to Your Account
Page 1 of 2 12 LastLast
Results 1 to 20 of 22
  1. #1
    Join Date
    Sep 2018
    Country
    UNITED STATES
    State/Province
    Michigan
    Posts
    54
    Post Thanks / Like
    Likes (Given)
    12
    Likes (Received)
    29

    Default Rotary Encoder for DRO? anybody done it.

    Has Anybody used a rotary encoder for a DRO?

    Currently I have an old acurite DRO on my bridgeport, over the weekend the x axis scale bit the dust.

    Instead of buying a new DRO setup, I am tempted to buy a good ground ballscrew set that wouldn't cost much more than a dro, and add a rotory encoder to it to keep track of position. The added benefit of course would be that I can easily add a couple of servos for a cnc setup in the future.

    The kind of work I use the mill for would not be affected by the couple tenths of backlash in a ballscrew.

    Anybody done anything like this?

  2. #2
    Join Date
    Nov 2003
    Country
    UNITED STATES
    State/Province
    Texas
    Posts
    5,520
    Post Thanks / Like
    Likes (Given)
    189
    Likes (Received)
    1865

    Default

    There is no reason why that couldn't be made to work. The problem, as I see it, is that the non-functional scale and the rotary encoder are not likely to be talking the same language. So some kind of conversion board will be needed to convert the rotary code into whatever that scale output. One of those PIC style chips would be a good candidate: either a plain PIC device or one of the higher level ones like Arduino. Of course, it will require programming and to do that you will need detailed information about how your existing, non-functioning scale communicates.

  3. #3
    Join Date
    Dec 2013
    Location
    Fort Wayne, IN
    Posts
    771
    Post Thanks / Like
    Likes (Given)
    812
    Likes (Received)
    272

    Default

    As long as your Acu-Rite is a TTL signal box (anything by the MillMate) you can definitely add a rotary encoder to it. It might not be as simple as plug and play, but yet you can definitely get it hooked up.

    +1 to EPAIII

  4. #4
    Join Date
    Oct 2005
    Country
    UNITED STATES
    State/Province
    Pennsylvania
    Posts
    18,807
    Post Thanks / Like

    Default

    Quote Originally Posted by TheBigLebowski View Post
    The added benefit of course would be that I can easily add a couple of servos for a cnc setup in the future.

    The kind of work I use the mill for would not be affected by the couple tenths of backlash in a ballscrew.

    Anybody done anything like this?
    Yup, over at "The cnc zone dot com".....

  5. #5
    Join Date
    Feb 2009
    Location
    Colchester UK (where the lathes used to be made)
    Posts
    1,989
    Post Thanks / Like
    Likes (Given)
    1417
    Likes (Received)
    604

    Default

    That's pretty much how every cnc machine works (a few have additional scales) , The SWI prototrak system on my knee mill works exactly that way .

    One word of caution though: ball screws will back drive so you may have to lock the stationary axes while making cuts (the extra drag of motors and belts is usually enough on my machine).

    Bill

  6. Likes HWElecRepair, AlfaGTA liked this post
  7. #6
    Join Date
    Nov 2004
    Location
    Webster Groves, MO
    Posts
    7,690
    Post Thanks / Like
    Likes (Given)
    1974
    Likes (Received)
    3681

    Default

    A friend installed a ball screw kit on a BP and belted a rotary encoder to it. The readout works fine but the ball screw sucks. You can read the backlash on the original dials. It is definitely in the screw. He asked the manufacturer why it had .002" backlash and was told that was normal. A good ballscrew setup uses a second ballnut with a spring between them, removing almost all the backlash. My 1977 Boston Digital CNC mill has about .0004" backlash on the Y axis after 41 years of use. They did it right. It turns out that the BP replacement unit had only single ballnuts that have the balls selected to reduce backlash and naturally they will wear. If you can't get a proper ballscrew, don't bother.

    Bill

  8. Likes Newman109 liked this post
  9. #7
    Join Date
    Mar 2012
    Location
    Geilenkirchen, Germany
    Posts
    2,605
    Post Thanks / Like
    Likes (Given)
    1528
    Likes (Received)
    1323

    Default

    I'm not a CNC guy, but I wouldn't do it because it is poor engineering with what is available today. If the encoder is connected to the screw, the reading is indirect. There is simply too much between the encoder and the work that is invisible to the electronics, like backlash and temperature. I feel feedback must come from table movement only. Then you only have to deal with linearity between the scale and the table.

  10. #8
    Join Date
    Jun 2001
    Location
    St Louis
    Posts
    18,576
    Post Thanks / Like
    Likes (Given)
    2095
    Likes (Received)
    3279

    Default

    Quote Originally Posted by steve-l View Post
    I'm not a CNC guy, but I wouldn't do it because it is poor engineering with what is available today. If the encoder is connected to the screw, the reading is indirect. There is simply too much between the encoder and the work that is invisible to the electronics, like backlash and temperature. I feel feedback must come from table movement only. Then you only have to deal with linearity between the scale and the table.
    The right answer ^^^^^^^^^^^^^

  11. Likes HWElecRepair, ATA91710 liked this post
  12. #9
    Join Date
    Mar 2011
    Location
    Denver, CO USA
    Posts
    10,214
    Post Thanks / Like
    Likes (Given)
    58
    Likes (Received)
    5698

    Default

    Yes it will work.
    It may be a pain to communicate.
    It’s a suboptimal solution, but no worse than using the hand dials.

  13. Likes HWElecRepair liked this post
  14. #10
    Join Date
    Dec 2004
    Location
    uk
    Posts
    14,046
    Post Thanks / Like
    Likes (Given)
    4552
    Likes (Received)
    6716

    Default

    Check out the price of the chinese scales, its kinda not worth it and there going to be way more accurate than trying to translate screw angular position to actual movement, way better to just measure the movement directly :-)

  15. #11
    Join Date
    Jan 2007
    Location
    Flushing/Flint, Michigan
    Posts
    9,144
    Post Thanks / Like
    Likes (Given)
    501
    Likes (Received)
    7571

    Default

    Quote Originally Posted by steve-l View Post
    I'm not a CNC guy, but I wouldn't do it because it is poor engineering with what is available today. If the encoder is connected to the screw, the reading is indirect. ....
    I wonder how cnc machine tools without dual loop linear scales work?
    Seems there are lots of those in the field making parts.......
    Better maybe but once you mount the scale you sort of blindly trust you get caught up in Abbe errors and all that so its not a perfect world either.
    If you think mounting a scale to the side of a table will produce stuff to its resolution as a dead nuts number on the table you are not understanding how this works.
    Bob

  16. Likes HWElecRepair liked this post
  17. #12
    Join Date
    Mar 2011
    Location
    Denver, CO USA
    Posts
    10,214
    Post Thanks / Like
    Likes (Given)
    58
    Likes (Received)
    5698

    Default

    Quote Originally Posted by CarbideBob View Post
    I wonder how cnc machine tools without dual loop linear scales work?
    Seems there are lots of those in the field making parts.......
    Bob
    Really?
    I’m surprised to hear that.
    Who is selling machine tools without direct reading feedback?
    As I said, it’s no worse than using the dials on the crank, but why bother.

  18. #13
    Join Date
    Sep 2018
    Country
    UNITED STATES
    State/Province
    Michigan
    Posts
    54
    Post Thanks / Like
    Likes (Given)
    12
    Likes (Received)
    29

    Default

    Lots of the lower end VMCs come without glass scales standard. only one of the six fadals I used to work on had glass scales. Same story with haas. So long as the machine is kept tight (biggest issue on many fadals being thrust bearings) they are acceptably accurate for most parts. This is also how most of the later protrak type mills operate.

    That being said, anybody have recommendations for an ok cheapo dro.

  19. Likes Ed Call liked this post
  20. #14
    Join Date
    Jan 2007
    Location
    Flushing/Flint, Michigan
    Posts
    9,144
    Post Thanks / Like
    Likes (Given)
    501
    Likes (Received)
    7571

    Default

    Quote Originally Posted by Miguels244 View Post
    Really?
    I’m surprised to hear that.
    Who is selling machine tools without direct reading feedback?
    I can only do "duh" here.
    Do you even know how dual feedback loops work and the problems involved?
    In holding size part after part in many cases adding a linear scale can actually be worse than running a single motor encoder off the screw.
    Any servo system has direct feedback, scale and motor become two feedbacks which disagree and the working solution becomes much harder.
    Bob

  21. #15
    Join Date
    Mar 2012
    Location
    Geilenkirchen, Germany
    Posts
    2,605
    Post Thanks / Like
    Likes (Given)
    1528
    Likes (Received)
    1323

    Default

    Bob,
    You are not the only person on this forum that understands servo systems and feedback schemes. Yes, in years past rotary encoders were common, but not anymore and for very good reason. In those days thermal compensation sensors were also very common along with other band-aids all to compensate for those accuracy factors not visible to the electronics. Today everybody uses direct feedback simply because it is better and cheaper. Please note that I did not say optical scale feedback. Feedback can also be provided by magnetics as well at much greater cost.

  22. Likes HWElecRepair liked this post
  23. #16
    Join Date
    Dec 2004
    Location
    uk
    Posts
    14,046
    Post Thanks / Like
    Likes (Given)
    4552
    Likes (Received)
    6716

    Default

    Quote Originally Posted by TheBigLebowski View Post

    That being said, anybody have recommendations for an ok cheapo dro.
    SINO there cheap as chips, a std bridgport size setup is sub $400 its a no brainer, there also pretty dang reliable too!

  24. Likes steve-l liked this post
  25. #17
    Join Date
    Mar 2012
    Location
    Geilenkirchen, Germany
    Posts
    2,605
    Post Thanks / Like
    Likes (Given)
    1528
    Likes (Received)
    1323

    Default

    Adama is correct the SINO DROs are very good. Just as good as any I have used. I bought a BP clone in China 12 years ago or more and the SINO 3 axis DRO was factory installed. It has been flawless and accurate. You will also find that the SINO DROs are also sold under several other brand names.

  26. Likes HWElecRepair liked this post
  27. #18
    Join Date
    Mar 2011
    Location
    Denver, CO USA
    Posts
    10,214
    Post Thanks / Like
    Likes (Given)
    58
    Likes (Received)
    5698

    Default

    Quote Originally Posted by CarbideBob View Post
    I can only do "duh" here.
    Do you even know how dual feedback loops work and the problems involved?
    In holding size part after part in many cases adding a linear scale can actually be worse than running a single motor encoder off the screw.
    Any servo system has direct feedback, scale and motor become two feedbacks which disagree and the working solution becomes much harder.
    Bob
    I build automated machinery for a living.
    For anything except rough positioning we used direct read from the axis wer were controlling.
    Screw variance, vibration and loading made that generally necessary.
    Never built a machine tool though, I guess the loads and precision are lower than I would guess off the top of my head.

  28. #19
    Join Date
    Jan 2007
    Location
    Flushing/Flint, Michigan
    Posts
    9,144
    Post Thanks / Like
    Likes (Given)
    501
    Likes (Received)
    7571

    Default

    Quote Originally Posted by steve-l View Post
    Bob,
    You are not the only person on this forum that understands servo systems and feedback schemes. Yes, in years past rotary encoders were common, but not anymore and for very good reason. In those days thermal compensation sensors were also very common along with other band-aids all to compensate for those accuracy factors not visible to the electronics. Today everybody uses direct feedback simply because it is better and cheaper. Please note that I did not say optical scale feedback. Feedback can also be provided by magnetics as well at much greater cost.
    Ok, so I'm wrong.
    Guessing that since you know such you also know pitch,yaw,roll and abbe errors due to reader head placement.
    Can you provide me a link to magnetic capable at 1/10 micron? I'm sort of stuck with optical or laser scale to achieve this.
    Bob

  29. #20
    Join Date
    Mar 2011
    Location
    Denver, CO USA
    Posts
    10,214
    Post Thanks / Like
    Likes (Given)
    58
    Likes (Received)
    5698

    Default

    Quote Originally Posted by CarbideBob View Post
    Ok, so I'm wrong.
    Guessing that since you know such you also know pitch,yaw,roll and abbe errors due to reader head placement.
    Can you provide me a link to magnetic capable at 1/10 micron? I'm sort of stuck with optical or laser scale to achieve this.
    Bob
    Have you looked at the Sony scales?
    Micro Epsilon makes some interesting s nsors for narrow range high resolution work too.
    MTS magnetorestrictive sensors can give high resolution but you need to build the correction map.

  30. Likes HWElecRepair liked this post

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
  •