Kflop i/o confusion


Results 1 to 10 of 10

Thread: Kflop i/o confusion

  1. #1
    Registered
    Join Date
    Dec 2016
    Location
    Israel
    Posts
    10
    Downloads
    0
    Uploads
    0

    Default Kflop i/o confusion

    Hello,

    i'm thinking of purchasing a Kflop motion controller for my DIY cnc mill
    but i'm having issues understanding from the manuals how the IO works.

    i need to control
    - 4 movement servos and one spindle servo (all with encoders feedback).
    - some relays and solenoids

    Q
    - is there any simple breakout board that provides simple screw terminals to all IO? maybe some relays? kanalog seems to add i/o's which i don't need (and costs as much as the kflop itself)
    - i don't understand the difference between general purpose IO pins and ones designated to a specific task. like on JP7 - you can either have a step/dir axis or home switches? can't you have both?
    - regarding the PC itself, i'm thinking of coupling Kflop with Lattepanda's single board windows 10 PC, it's a quad core Intel atom with 4 gigs of ram, any reason why it shouldn't work?

    thanks,
    Gilad

    Similar Threads:


  2. #2
    Member TomKerekes's Avatar
    Join Date
    May 2006
    Location
    USA
    Posts
    4043
    Downloads
    0
    Uploads
    0

    Default Re: Kflop i/o confusion

    Hi Gilad,

    i need to control
    - 4 movement servos and one spindle servo (all with encoders feedback).
    - some relays and solenoids
    What kind of Servos do you have? Do you have Amplifiers/Drives? What kind of encoders do you have? Differential or single ended?

    is there any simple breakout board that provides simple screw terminals to all IO?
    Winford.com sells simple breakout boards for cables such as:
    2x13 0.1" Header (26 position IDC ribbon connector) Breakout Board with Screw Terminals - Winford Engineering

    maybe some relays?
    They also sell simple Relay boards that can be driven with LVTTL signals:
    Relay Board: TTL Logic Level Inputs, 2 SPDT 15A Relays - Winford Engineering

    i don't understand the difference between general purpose IO pins and ones designated to a specific task. like on JP7 - you can either have a step/dir axis or home switches? can't you have both?
    Yes you can. Those are only suggested Inputs for Limits if they are available. Home/Limits are handled in software and can be configured to any Input.

    regarding the PC itself, i'm thinking of coupling Kflop with Lattepanda's single board windows 10 PC, it's a quad core Intel atom with 4 gigs of ram, any reason why it shouldn't work?
    I don't see any reason it shouldn't work.

    Regards

    Regards
    TK http://dynomotion.com


  3. #3
    Registered
    Join Date
    Dec 2016
    Location
    Israel
    Posts
    10
    Downloads
    0
    Uploads
    0

    Default Re: Kflop i/o confusion

    hello,

    thanks for the reply Tom.
    i will use 4 clearpath servos (digital, integrated amp) for movement and a common chinese AC servo for the spindle (also, the amp is digital think).

    so if i understand correctly:
    only step/dir outputs are fixed(hard wired), so i will have to connect the 4 movement servos to JP7 and the spindle to JP5?

    thanks!



  4. #4
    Member TomKerekes's Avatar
    Join Date
    May 2006
    Location
    USA
    Posts
    4043
    Downloads
    0
    Uploads
    0

    Default Re: Kflop i/o confusion

    Hi Gilad,

    Yes for simple Step/Dir Drives with no encoder feedback to the Controller that should work. For more info see:
    Step and Direction Setup

    Regards

    Regards
    TK http://dynomotion.com


  5. #5
    Registered
    Join Date
    Dec 2016
    Location
    Israel
    Posts
    10
    Downloads
    0
    Uploads
    0

    Default Re: Kflop i/o confusion

    well, i learned too late clearpath servos do not offer encoder feedback .
    which means my system will not be truly closed loop. kflop will not be able to coordinate my axes based on feedback that way (only on wishful thinking...)
    so my only solution as i see it now, is to integrate linear scales on my axes, this should be possible with kflop, right?

    thanks!



  6. #6
    Member
    Join Date
    Jun 2010
    Location
    Australia
    Posts
    4252
    Downloads
    4
    Uploads
    0

    Default Re: Kflop i/o confusion

    I think you misunderstand.
    If it is a SERVO, then it will use encoder feedback. But the feedback will go to the driver, not to the controller.
    The controller will then treat the driver&motor as a stepper, with Step and Dir signals.

    If you want to use linear scales, that gets a bit more complex. There can be a serious stability problem which needs a dual-loop control. Galil offer this, but I am not sure whether anyone else does at this stage.

    Cheers
    Roger



  7. #7
    Registered
    Join Date
    Dec 2016
    Location
    Israel
    Posts
    10
    Downloads
    0
    Uploads
    0

    Default Re: Kflop i/o confusion

    no, i might have just explained myself badly, i'm fully aware that any servo will have a loop with its driver.
    i'm also aware that i don't have to connect feedback to kflop and it will still work but 'blind'.

    the thing is, i do want a full closed loop system. assuming the linear scales uses an appropriate signal,
    it should be possible to use their signal theoretically... but lets see what Tom has to say

    thanks,
    Gilad



  8. #8
    Registered
    Join Date
    Apr 2017
    Location
    United States
    Posts
    22
    Downloads
    0
    Uploads
    0

    Default Re: Kflop i/o confusion

    Hi Gilead,

    This is Tom from Teknic. I believe I understand your requirements and concerns and Roger has shed some good insight. Let's start at the beginning.

    There are many thousands of high production, mid production, and DIY CNC machines in use that use closed loop servo systems that do not feedback the encoder signal to the controller so the design is field tested and proven successful by many different machine integrators. So how is this possible?

    Old school designs from many years ago used the controller to close the position servo loop. The servo drive would close the velocity and torque servo loops. This worked ok but had some deficiencies by design. One of the main drawbacks was there there were two cycle clocks being used by two different devices. The controller used its internal cycle clock while the driver used its internal cycle clock. This meant that the device sending/updating the control signals and controlling the position loop was on one clock cycle and the drive controlling the velocity and torque loops was on a different clock cycle. What is a better way?

    It is true that the controller sends out the signal to the servo drives like a stepper, but this is where the similarity ends. The controller is following a cut pattern to send out the step and direction signals. The servo drives follow the command by controling all three servo loops (position, velocity, torque) using one cycle clock. This is the tightest architecture with the least amount of synchronization and latency issues. Since the servo drives are closed loop, they know what the command is and they know where they are at any point in time. They compare the commanded position with the actual position and will increase and decrease current/torque to achieve the least amount of tracking error (the difference between the command position and actual position). If the tracking error exceeds a specified value (due to machine misalignment, an obstruction, etc.), the servo drive will shutdown and let the controller know that there is a machine problem. So the axes are fully closed loop and are not "running blind" like a stepper.

    As Roger mentioned, adding external linear scales is much more complicated and is typically outside the scope of a DIY project. (and virtually never needed for a CNC type machine)

    That being said, if you needed to trigger a device (like a solenoid or laser) based on a specific encoder position, then there is a different ClearPath model that can do that. But for your typical CNC type of machines (laser, router, plasma, waterjet), the ClearPath SD family works very well.



  9. #9
    Member TomKerekes's Avatar
    Join Date
    May 2006
    Location
    USA
    Posts
    4043
    Downloads
    0
    Uploads
    0

    Default Re: Kflop i/o confusion

    Hi Gilad,

    Assuming the Clear Path Servos can be tuned to close the loop with the same performance as KFLOP closing the loop, then the overall system precision/performance should be the same in both cases. If there is excessive following error due to a motor stall, excessive friction, hitting an obstacle, etc. the system will still fault and stop (assuming the ClearPath faults are interfaced back to KFLOP). The advantage of having the position feedback connected back to the controller is that after a fault or Drive disable/manual motion the controller can still track the position so the system can resume without loss of position that would require re-homing. With only Step/Dir going to the Drive once there is a fault or disable and the Axis moves there is no way to know where it is. The feedback to KFLOP also permits monitoring and plotting performance.

    Adding Linear Scale feedback to KFLOP would allow real-time corrections for mechanical errors such as backlash, compliance, and lead screw errors. Many of our Users do this with great success. Note there are limits based on the system dynamics on how quickly errors can be corrected. For example backlash makes it impossible for the motor to hold any position perfectly when external forces are applied to the Axis. I mention this because some Users have the misconception that by adding Linear Scales any amount of backlash will be entirely removed.

    As Roger describes when adding Linear Scales a dual loop/dual feedback approach is usually required. That is because with only Linear Scale feedback it is difficult to control motor motion well. For example the motor might be accelerating rapidly through the backlash or it might be stationary and the Linear Scale Position would show no motion in both cases. Extra information from the motor shaft from a tachometer or encoder allows the motor speed to be known and controlled independently from the Linear Scales. Such a dual loop configuration is exactly what you would have if you added Linear Scales.

    HTH
    Regards

    Regards
    TK http://dynomotion.com


  10. #10
    Registered
    Join Date
    Dec 2016
    Location
    Israel
    Posts
    10
    Downloads
    0
    Uploads
    0

    Default Re: Kflop i/o confusion

    Tom/Tom/Roger,

    thanks for the explanations, things are clearer for me now.

    TK - my main concern was not in case of a catastrophic failure like a crash.
    i'm more concerned with cases in which one axis have a much bigger resistance than another axis (lets say while they're trying to create circular motion)
    even if the driver compensates quickly, it will still take some time to react and reach the desired speed again.

    having said that, after your explanations i do feel more confident with my setup

    cheers



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
  •  


About CNCzone.com

    We are the largest and most active discussion forum for manufacturing industry. The site is 100% free to join and use, so join today!

Follow us on


Our Brands

Kflop i/o confusion

Kflop i/o confusion