504,015 active members
4,352 visitors online
Register for free
Login
Page 2 of 3 123
Results 13 to 24 of 25
  1. #13
    Registered
    Join Date
    Jan 2016
    Posts
    66

    Re: Closing the loop and controllers

    Thanks for the additional advice and replies. There's a lot to think about. It seems that maybe some of the advantages in choosing between closing the loop in the controller versus the driver may depend on hardware specs of the controller - control refresh rates, D/A resolution and update frequency, and whatnot.

    I really appreciate your help as I learn more about this!

    Micah

  2. #14

    Join Date
    Jan 2005
    Posts
    10124

    Re: Closing the loop and controllers

    Quote Originally Posted by burbingus View Post
    Hey Mac, why wouldn't there be a noticeable difference in closing in the driver versus the controller?
    You tell me, modern Servo drives can do a better job, both ways are doing the same thing, one less complicated than the other, I have done them both ways, and I prefer to close the loop in the servo drive, you can do it any way you like, but I would be cutting parts while you will be trying to get your system tuned, and if there are complication's with your system you may never get it perfectly tuned
    Mactec54

  3. #15
    Registered
    Join Date
    Nov 2013
    Posts
    291

    Re: Closing the loop and controllers

    Hi,
    I agree with Mactec54. Modern AC servo drives have really flexible control options.

    For instance my Allen Bradley servo drive, and it is by no means the latest model, has a thermal model of my servo built in but also has 10 point magnetic
    model to simulate magnetic saturation. These sorts of details are not available for most servos or most controllers, thus the closed loop servo drive is
    materially better at predicting thermal and magnetic performance.

    The only time I would regard a loop closed by the controller to be a distinct advantage is if you were using a linear scale or
    feedback device OTHER THAN an encoder fitted onto the servo.

    Craig

  4. #16
    Registered
    Join Date
    Feb 2008
    Posts
    580

    Re: Closing the loop and controllers

    Quote Originally Posted by joeavaerage View Post
    Hi,
    I agree with Mactec54. Modern AC servo drives have really flexible control options.

    For instance my Allen Bradley servo drive, and it is by no means the latest model, has a thermal model of my servo built in but also has 10 point magnetic
    model to simulate magnetic saturation. These sorts of details are not available for most servos or most controllers, thus the closed loop servo drive is
    materially better at predicting thermal and magnetic performance.

    The only time I would regard a loop closed by the controller to be a distinct advantage is if you were using a linear scale or
    feedback device OTHER THAN an encoder fitted onto the servo.

    Craig
    Actually there are many advantages of having the controller close the loop or at least have bi-directional communication with the drives:

    1. Absolute encoder support (very awkward with step/dir interfaces)
    2. Real time load, and smart following error monitoring
    3. Complex Feed forward, the controller is aware of whats coming next the drive is not ( feedback is great but its always too late! )

    Most higher end CNC systems have position feedback to the controller

  5. #17

    Join Date
    Jan 2005
    Posts
    10124

    Re: Closing the loop and controllers

    Quote Originally Posted by PCW_MESA View Post
    Actually there are many advantages of having the controller close the loop or at least have bi-directional communication with the drives:

    1. Absolute encoder support (very awkward with step/dir interfaces)
    2. Real time load, and smart following error monitoring
    3. Complex Feed forward, the controller is aware of whats coming next the drive is not ( feedback is great but its always too late! )

    Most higher end CNC systems have position feedback to the controller
    ( 1 ) is incorrect I do it all the time, and with many different high end servo drives Absolute Encoder I have never had a problem with

    ( 2 )Following error is also covered when the servo drive has an encoder feedback to the control for monitoring any error which most quality servo drives have

    ( 3 ) There is not difference if the control is looking at the encoder which most good closed loop servo drives can send the encoder position to the control, if it is to late as you say you would have a fault which does not happen, Do you mean Look-ahead or something else,

    The drive does not know what's coming next even if the loop is closed in the control both ways the drive does not do anything until it gets a signal to do something from the control
    Mactec54

  6. #18
    Registered
    Join Date
    Feb 2008
    Posts
    580

    Re: Closing the loop and controllers

    ( 1 ) is incorrect I do it all the time, and with many different high end servo drives Absolute Encoder I have never had a problem with

    How does the controller know the initial position at startup to avoid homing? This is not possible with just step/dir

    ( 2 )Following error is also covered when the servo drive has an encoder feedback to the control for monitoring any error which most quality servo drives have

    Right, this requires bidirectional communication

    ( 3 ) There is not difference if the control is looking at the encoder which most good closed loop servo drives can send the encoder position to the control,
    if it is to late as you say you would have a fault which does not happen, Do you mean Look-ahead or something else,

    The drive does not know what's coming next even if the loop is closed in the control both ways the drive does not do anything until it gets a signal to do something from the control[/QUOTE]

    Yes, but the controller can send a smarter signal because it does effectively have lookahead

    This is why most high end CNC and robotic systems have position (and maybe velocity and torque) feedback to the controller

  7. #19

    Join Date
    Jan 2005
    Posts
    10124

    Re: Closing the loop and controllers

    Quote Originally Posted by PCW_MESA View Post
    ( 1 ) is incorrect I do it all the time, and with many different high end servo drives Absolute Encoder I have never had a problem with

    How does the controller know the initial position at startup to avoid homing? This is not possible with just step/dir

    ( 2 )Following error is also covered when the servo drive has an encoder feedback to the control for monitoring any error which most quality servo drives have

    Right, this requires bidirectional communication

    ( 3 ) There is not difference if the control is looking at the encoder which most good closed loop servo drives can send the encoder position to the control,
    if it is to late as you say you would have a fault which does not happen, Do you mean Look-ahead or something else,

    The drive does not know what's coming next even if the loop is closed in the control both ways the drive does not do anything until it gets a signal to do something from the control
    Yes, but the controller can send a smarter signal because it does effectively have lookahead

    This is why most high end CNC and robotic systems have position (and maybe velocity and torque) feedback to the controller[/QUOTE]

    You still have the same feed back to the control either way you want to do it, the control is only looking at the Encoder in both cases, whether you close the loop in the Drive or the Control
    Mactec54

  8. #20
    Super Moderator Al_The_Man's Avatar
    Join Date
    Dec 2003
    Posts
    23156

    Re: Closing the loop and controllers

    Quote Originally Posted by burbingus View Post
    Thanks for the additional advice and replies. There's a lot to think about. It seems that maybe some of the advantages in choosing between closing the loop in the controller versus the driver may depend on hardware specs of the controller - control refresh rates, D/A resolution and update frequency, and whatnot.
    Micah
    I guess I was spoiled as I came into the CNC field from the 'Top' end installing and commissioning Fanuc and Mitsubishi, etc, For my self build units I used Galil Motion, which to me was the closest thing to the 'big boys'.
    It is quite enlightening to watch the Galil educational video's by Dr Jacob Tall.
    There is one on the usage of a dual feedback system that is required when you use a Scale and an encoder on the motor to eliminate problems that occur if servo backlash exists (oscillation etc), splitting the PID loop between the two and how to tune it....
    The Galil is also capable of look-ahead motion. The 32 bit microprocessor, has feed-forward and notch filter, encoder feedback up to 12Mhz!
    I found the gearing ability very handy, for example using two servo motors for the Table X axis to synchronize them
    Incidentally the very simple analogue drives that Galil sell are relabeled Advanced Motion Controls, simple transconductance amplifiers.
    Just my 2¢
    Al.
    CNC, Mechatronics Integration and Custom Machine Design

    “Logic will get you from A to B. Imagination will take you everywhere.”
    Albert E.

  9. #21
    Registered
    Join Date
    Jul 2003
    Posts
    1670

    Re: Closing the loop and controllers

    Quote Originally Posted by Al_The_Man View Post
    I guess I was spoiled as I came into the CNC field from the 'Top' end installing and commissioning Fanuc and Mitsubishi, etc, For my self build units I used Galil Motion, which to me was the closest thing to the 'big boys'.
    It is quite enlightening to watch the Galil educational video's by Dr Jacob Tall.
    There is one on the usage of a dual feedback system that is required when you use a Scale and an encoder on the motor to eliminate problems that occur if servo backlash exists (oscillation etc), splitting the PID loop between the two and how to tune it....
    The Galil is also capable of look-ahead motion. The 32 bit microprocessor, has feed-forward and notch filter, encoder feedback up to 12Mhz!
    I found the gearing ability very handy, for example using two servo motors for the Table X axis to synchronize them
    Incidentally the very simple analogue drives that Galil sell are relabeled Advanced Motion Controls, simple transconductance amplifiers.
    Just my 2¢
    Al.
    Linuxcnc can do all of that also..

  10. #22

    Join Date
    Jan 2005
    Posts
    10124

    Re: Closing the loop and controllers

    Quote Originally Posted by Al_The_Man View Post
    I guess I was spoiled as I came into the CNC field from the 'Top' end installing and commissioning Fanuc and Mitsubishi, etc, For my self build units I used Galil Motion, which to me was the closest thing to the 'big boys'.
    It is quite enlightening to watch the Galil educational video's by Dr Jacob Tall.
    There is one on the usage of a dual feedback system that is required when you use a Scale and an encoder on the motor to eliminate problems that occur if servo backlash exists (oscillation etc), splitting the PID loop between the two and how to tune it....
    The Galil is also capable of look-ahead motion. The 32 bit microprocessor, has feed-forward and notch filter, encoder feedback up to 12Mhz!
    I found the gearing ability very handy, for example using two servo motors for the Table X axis to synchronize them
    Incidentally the very simple analogue drives that Galil sell are relabeled Advanced Motion Controls, simple transconductance amplifiers.
    Just my 2¢
    Al.
    I did retro fits also and used / installed Galil as well, I went through the Galil training classes they have, that is one of the reasons why I don't recommend Galil for Hobby users, Galil is not for everyone especially someone at the Hobby level

    One thing they do have is lots of information about there products
    Mactec54

  11. #23

    Re: Closing the loop and controllers

    Quote Originally Posted by Al_The_Man View Post
    ..............encoder feedback up to 12Mhz! ...............

    Al.
    The latest generation Accelera controllers (18x6 and 40x0) are 22MHz now, with loop update times as low as 62?s. Much smoother operation.
    Jim Dawson
    Sandy, Oregon, USA

  12. #24
    Registered
    Join Date
    Jul 2017
    Posts
    12

    Re: Closing the loop and controllers

    such analysis is provided by the CAM that allows the analysis of the program controlling the processing of any other https://machine.report/

Page 2 of 3 123

Similar Threads

  1. Closing the loop with Path Pilot. Possibly dual (global) closed loop?
    By SwampDonkey in forum Tormach Personal CNC Mill
    Replies: 29
    Last Post: 05-04-2016, 09:39 PM
  2. Replies: 0
    Last Post: 11-14-2015, 06:52 AM
  3. Replies: 6
    Last Post: 06-06-2015, 10:30 PM
  4. Thoughts on "Closing the Loop"
    By Fish4Fun in forum Dynomotion/Kflop/Kanalog
    Replies: 13
    Last Post: 01-15-2014, 06:43 PM
  5. Replies: 1
    Last Post: 01-04-2013, 04:45 AM

Tags for this Thread

Posting Permissions

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