Announcement

Collapse
No announcement yet.

Need larger pulse range on HSO modules

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts


  • Need larger pulse range on HSO modules

    So I am playing around with some newer gen servo drives from Delta, and they have new 17-bit and 20-bit encoders on the drives. These things are extremely smooth running, love them. Only problem is that because the encoder feedback count is so huge, the actual pulses per revolution is 1,280,000. In Productivity hardware setup I cant actually input that number, see screenshot below:

    Click image for larger version  Name:	HSO.png Views:	1 Size:	53.9 KB ID:	119723



    I can get it to 999,999.0 as the highest, when I need it to go to 1,280,000.0. I can get around this by using 640,000.0 instead, and using a P1-44 numerator of 2 and a P1-45 denominator of 1. However it would be really nice if Productivity modules and software could accommodate pulse counts per revolution into the multiple millions. Hopefully this can be something like a simple software change to allow another digit or two in that box there. It would be nice to be able to set the revolution info properly and run 1:1 ratio in the servo drive itself to do some testing and for good low speed performance.


    Also, it would be great if you could come out with a newer HSO module for both the P3k and P2k that allowed up to something like a 25MHz output. The 1MHz we have now is kinda limiting... With this servo drive's PPR, and the speed of 4,500RPMs that I am going to be running it, the drive woiuld honestly need an input frequency of 96MHz. lol. I realize then you would run into cabling issues and things like that, and I doubt there are any servo's out there that even support 25MHz inputs right now (the Delta one I am using only support 4MHz input). But it would be nice to have the module support already there for future servo drives, and there is cabling that easily supports 25MHz on it. Cat6 ethernet cables comes to mind with its 550MHz ratings. I normally use ethernet cable for my encoder signals anyway, and it is working well for line driver HSO to servo signals in this application right now too.
    Last edited by MikeN; 12-19-2018, 09:58 AM.


  • #2
    Im no help, but there are actually encoders with 1.28 million pulses per rev, and applications that need that much resolution? Wow, that blows my mind!

    Comment



    • #3
      Running at 4500rpm and doing so with 1.28m ppr strikes me as being way out on the edge. I'm sure the need exists, but I would think it is well outside 3 sigma.

      Comment



      • #4
        When you pulse the drive to advance, you tell the drive how much you want it to advance, in encoder pulses. Wether the encoder can read one millionth of a full rotation does not mean you can actually increment your shaft position by 0.1 seconds of arc. That encoder accuracy supports the pid in controlling the speed at any time during the motion. The motion path is a sum of motion segments, each one with its own slip ( as in slide) error. A high encoder count helps better compensate those errors, such as the final stop results with the least possible error.

        Think of the motion of a robotic arm: it needs to start and stop at the exact location of pick and place. The errors along the path of motion can not add up. They need to cancel each other at the end of the path. The high encoder count supports this.

        When you drive a servo with pulse and direction, the servo controller will need to know:
        1. what is the frequency range of the pulses.
        2. what is the encoder count you want to see per pulse

        these both determine the target speed which the servo will autonomously try to ramp up to.
        if you tell the servo to move one encoder count per second, that could result in a speed ramp too flat to achieve. Will either fault or accumulate errror until is able to move. Undesirable situation.

        if you try to tell the servo to move a predetermined number of pulses- therefore encoder counts, then best is to do it in position mode, programmatically.

        The hso is used with stepper drives. Can be used with servo as well, but I donít see the point.

        Comment



        • #5
          Originally posted by g.mccormick View Post
          Im no help, but there are actually encoders with 1.28 million pulses per rev, and applications that need that much resolution? Wow, that blows my mind!
          And thats barely anything. lol. There are some, like Rockwell stuff for example, that go up to near 8.4 million encoder feedback pulses per revolution:
          Click image for larger version  Name:	KinetixVP.png Views:	1 Size:	101.1 KB ID:	119739








          Originally posted by Alexandru View Post
          The hso is used with stepper drives. Can be used with servo as well, but I donít see the point.
          This application I am doing right now with the Delta drive does not necessarily need to be on an HSO and use a PT input type, I am doing it this way initially as it is far easier to set up and test drive and motor stuff rather than learn the new multi-parameter 32-bit hexidecimal coding and bit configuration for programming the necessary moves in PR mode. I will be learning that new stuff down the line for this project, but not right now.

          I do normally need to use an HSO module on many of our machines because pre-set movements done in PR mode do not let me have continually variable velocity of the movement profile, which is a necessity for my applications. I need the torque and speed of a servo that steppers cant give, but I need the pulse train movement used with many steppers. I will have a few thousand different velocities all through a day, and it depends on what product is being run and what the batch of liquid is like for the product.

          I also have needed to use an HSO with the servo in PT mode with using the PLC and HSO in basically just velocity mode, with a velocity move going constantly but the speed being variable. I tried velocity mode in the servo first, but even with accel and decel set to 2ms and the drive tuned up the response was just too sluggish to use. I had to use PT mode with the velocity controlled in the PLC and it fixed the issue right up.
          Last edited by MikeN; 12-20-2018, 10:15 AM.

          Comment

          Working...
          X