PC Base DRO - Page 2

Page 2 of 2 FirstFirst 12
Results 13 to 23 of 23

Thread: PC Base DRO

  1. #13
    Registered aadityadengle's Avatar
    Join Date
    Jul 2011
    Location
    India
    Posts
    26
    Downloads
    0
    Uploads
    0

    Default

    @bobsch

    Only converting the signals from RS422 to RS232 or the serial is not the solution. Basically we have to decode the signals so that it can gives out the values and the direction(+/-). i agree that this can be done with software but that will create a timing lack and will result in to missing out some pulses. missing a pulse will cost me about 5 micron where as my scale will move with the speed 10mm (10K micron) / sec.
    if we consider a idle pc it will take almost 3 microseconds to calculate the direction and the values.
    with the help of internal card i am planning to give an packaged output to the pc which is already decoded and hence will take the lesser time to read. also the timing lack will be less as the computing is done at IC level with 33 or 66MHz clock (PCI Clock runs on the 33 or 66 MHz frequency) also by choosing the pci card we are avoiding to build the extra ref. clock and other.



  2. #14
    Registered
    Join Date
    Aug 2008
    Location
    UK
    Posts
    573
    Downloads
    0
    Uploads
    0

    Default

    Which scales do you have (or intend to buy)? - the document describes three types PPL TTL and HTL which have different power and output voltages.

    If the cable run is short <2m then you could probably use the 5v ones and only one leg of the TTL differential output (saving the need for a 422 receiver) . You could feed this straight into one of a DSPIC's quadrature inputs. (see: http://ww1.microchip.com/downloads/e...Doc/70063D.pdf) - The DSPIC 33 series have two encoders on-chip - The dsPIC33FJ32 is only about $3 in quantity (and Microchip will supply free samples!)

    Personally, I would avoid using PCI or any other PC card format, they are becoming obsolete rapidly. If you stick to USB (or serial via USB) it will give it some universal utility.

    The DSPIC's don't have USB hardware but you could simply use a serial - USB converter chip (or module, or adapter) .

    The PIC's firmware would be very simple - it just has to read the on-board quadrature counter registers and format the result out onto the serial port. The output to the PC should be in a software friendly form - e.g. if you intend to use a spreadsheet to display the data on the PC then choose a numeric string.

    Bill

    Bill


  3. #15
    Registered aadityadengle's Avatar
    Join Date
    Jul 2011
    Location
    India
    Posts
    26
    Downloads
    0
    Uploads
    0

    Default

    Quote Originally Posted by BillTodd View Post
    Which scales do you have (or intend to buy)? - the document describes three types PPL TTL and HTL which have different power and output voltages.

    If the cable run is short <2m then you could probably use the 5v ones and only one leg of the TTL differential output (saving the need for a 422 receiver) . You could feed this straight into one of a DSPIC's quadrature inputs. (see: http://ww1.microchip.com/downloads/e...Doc/70063D.pdf) - The DSPIC 33 series have two encoders on-chip - The dsPIC33FJ32 is only about $3 in quantity (and Microchip will supply free samples!)

    Personally, I would avoid using PCI or any other PC card format, they are becoming obsolete rapidly. If you stick to USB (or serial via USB) it will give it some universal utility.

    The DSPIC's don't have USB hardware but you could simply use a serial - USB converter chip (or module, or adapter) .

    The PIC's firmware would be very simple - it just has to read the on-board quadrature counter registers and format the result out onto the serial port. The output to the PC should be in a software friendly form - e.g. if you intend to use a spreadsheet to display the data on the PC then choose a numeric string.

    Bill
    @Bill,
    Thanks for reply. My scales gives 5V TTL output. and just to confirm and make my grounds clear which will take shorter time calculate the position one USB decoder or the PCI decoder. as i have stated in my last post i cant afford loosing too many pulses.



  4. #16
    Registered
    Join Date
    Aug 2008
    Location
    UK
    Posts
    573
    Downloads
    0
    Uploads
    0

    Default

    It is still not clear in my mind what you are trying to do; Are you trying to capture positional data on the move (for a servo system, for instance) or are you digitising (locating the position of some point, stopping, then reading the position into a PC)???


    which will take shorter time calculate the position one USB decoder or the PCI decoder. as i have stated in my last post i cant afford loosing too many pulses.
    The PIC's hardware will ensure that the position counter will not lose any pulses (unless the scale is moved far too fast) However, there will be a small delay between reading the position register and getting that position (number) into the PC. If the scale is moving when the position capture process is triggered then there will be a small position error (because the scale will have moved a little by the time the PC has the data).

    It is possible to minimise the errors by calculating the delay and allowing for the velocity of the scale (i.e. you add, or subtract, a correction to the position count).

    Bill


  5. #17
    Registered aadityadengle's Avatar
    Join Date
    Jul 2011
    Location
    India
    Posts
    26
    Downloads
    0
    Uploads
    0

    Default

    humm..... thanks bill, i have not considered your delay correction method.

    and yes what i want to do is read or capture the current position at the axis move and control the axis movement as per the read out. Say if i started the axis in forward from home and i have to stop (set) the same at 500 mm with slowing down the axis movement at 400 mm onwards.
    now in my software / program i will put the input for final position. after starting from the home scale should report its position (absolute / relative no issue) to the pc. so that pc can cross verify the scale's position with what i have set in my program. and triggers the motor accordingly. now here i can not use any kind of hardware limit switch to slow down or to end the movement as both of them are completely programmable between 0 to 900 mm.
    More even my Z axis is hydraulically driven and needs precise operation. as the movement will control the bend angle.

    i require the following logic.
    X Axis Run => Read position => check with input => if ok then stop + if not ok then again read but do not stop


    all at speed 10mm / sec.



  6. #18
    Registered
    Join Date
    Aug 2008
    Location
    UK
    Posts
    573
    Downloads
    0
    Uploads
    0

    Default

    Sounds like you could do what you want with a classic PID (Proportional , Integral, Derivative) controller, programmed into the PIC or PC.

    Your 10mm/s is only 10,000 pulses per second even at 1 micron resolution.

    Bill

    Bill


  7. #19
    Registered aadityadengle's Avatar
    Join Date
    Jul 2011
    Location
    India
    Posts
    26
    Downloads
    0
    Uploads
    0

    Default

    @bill
    what about DRO_Plus Project - by Tom Kulaga ?

    in addition to that we can use W83629D and W83628AG to convert PCI to ISA so that we can use the card with the now days pc

    Attached Thumbnails Attached Thumbnails PC Base DRO-w83628ag-pdf   PC Base DRO-w83629d-pdf  
    Last edited by aadityadengle; 10-06-2012 at 08:14 AM.


  8. #20
    Registered
    Join Date
    Aug 2008
    Location
    UK
    Posts
    573
    Downloads
    0
    Uploads
    0

    Default

    You seem to be confusing two different things:

    1 - displaying the position - for example a DRO
    Here the axis is moved (usually manually), the DRO counts the pulses from the scale and a display in a suitable unit (mm or inch) is displayed to the user.


    2 - controlling the position - using a Servo for instance
    Here an arbitrary number (machine steps) is sent to a servo controller (usually converted from Inches or millimetres). The servo controller drives a motor back and forth until the position of the axis (counted from the scale)matches the desired number.

    It does not matter to the servo system if it counts the pulses from a rotary encoder on the motor or from a linear scale. However, it is important for the stability of the system, that the coupling between the drive (motor) and the feedback (e.g. the counter) is as 'stiff' and free from backlash (or time delays) as possible. The longer it takes for the positional feedback to arrive at the controller, harder it is to maintain stability.



    ATM I am still guessing that you are trying to use a linear scale as feedback for a positioning system

    Bill

    Bill


  9. #21
    Registered aadityadengle's Avatar
    Join Date
    Jul 2011
    Location
    India
    Posts
    26
    Downloads
    0
    Uploads
    0

    Default

    yes bill i am using the scales just for feedback and ensure the axis position. the scales will provide actual travel or position to my software and on that basis the software logic will control the stepper motor.
    I have done with the stepper controlling part in my project and now working on the feedback reading. the idea of using linear scales for the feedback is just exactly as you said. i want to minimize the backlashes and do my cnc a bit to the professional level.



  10. #22
    Registered
    Join Date
    Aug 2008
    Location
    UK
    Posts
    573
    Downloads
    0
    Uploads
    0

    Default

    My thoughts only but ...

    Steppers either have enough power to get the job done, or they don't. If they are losing steps then adding extra ones seems a bit of a bodge (I suppose if speed is not important then slowing to ensure the extra steps actually do something might be OK - but then why not just go slow enough not to lose steps in the first place???)

    On the other hand, if there is an inherent non-linearity in the stepper-motor drive mechanism then correcting by linear feedback would make some sense (although a similar effect could be achieved with a correction table)

    In short steppers are good at low speed only their torque drops of at speed and there is no way to add extra 'oomph' when the going get tough.

    So...

    If you are going to the trouble of adding linear feedback , why not use a servo motor instead of a stepper?

    Bill

    Bill


  11. #23
    Registered aadityadengle's Avatar
    Join Date
    Jul 2011
    Location
    India
    Posts
    26
    Downloads
    0
    Uploads
    0

    Default

    finally used US Digital QSB..... it rocks



Page 2 of 2 FirstFirst 12

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

PC Base DRO

PC Base DRO

PC Base DRO