Closed Loop CNC - Page 5


Page 5 of 5 FirstFirst ... 2345
Results 81 to 99 of 99

Thread: Closed Loop CNC

  1. #81
    Member samco's Avatar
    Join Date
    Jul 2003
    Posts
    1753
    Downloads
    2
    Uploads
    0

    Default Re: Closed Loop CNC

    I shouldn't say same.. Similar magic..



  2. #82
    Member mactec54's Avatar
    Join Date
    Jan 2005
    Location
    USA
    Posts
    15362
    Downloads
    0
    Uploads
    0

    Default Re: Closed Loop CNC

    Quote Originally Posted by samco View Post
    The thing is - the magic in the closed loop drives - is the same magic in linuxcnc. It can control

    step/dir (variants like - up/down, quadrature)
    voltage
    current
    torque
    velocity
    smart serial
    Ethercat
    It can directly control 3 phase servos...
    Others that I can't think of atm...

    sam

    I think most people know this, the problem is the pulse rate is to low in LinuxCNC by it's self to be of any use for a direct drive for high speed servo

    Mactec54


  3. #83
    Member samco's Avatar
    Join Date
    Jul 2003
    Posts
    1753
    Downloads
    2
    Uploads
    0

    Default Re: Closed Loop CNC

    Hmm.. the thing is - when linuxcnc is closing the servo loop - there isn't a pulse rate. That is a step/dir thing.

    I have tested linuxcnc with pretty high acceleration and velocities - so if you can get the servos/drives big enough to handle it - linuxcnc will do it.

    example.. 400in/sec^2 and 500ipm (and that is not the limit..)





    Quote Originally Posted by mactec54 View Post
    I think most people know this, the problem is the pulse rate is to low in LinuxCNC by it's self to be of any use for a direct drive for high speed servo




  4. #84
    Member samco's Avatar
    Join Date
    Jul 2003
    Posts
    1753
    Downloads
    2
    Uploads
    0

    Default Re: Closed Loop CNC

    BTW - with linuxcnc you can easily get software step rates into the 40khz. External interface hardware step rates are in the mhz...



  5. #85
    Member
    Join Date
    Jun 2010
    Location
    Australia
    Posts
    4252
    Downloads
    4
    Uploads
    0

    Default Re: Closed Loop CNC

    I am sure the Dyna AC servo drive is well made and powerful, but it uses serial comms in a custom format instead of step/dir, and there are no modern NIST-compatible CNC systems which use that format. That means you are locked into ONE of the older generation proprietary systems with little chance of an upgrade to a different brand. It often also means that adding a feature can be difficult or very expensive.

    I remember the complaints of one business owner who wanted to add some feature (I think it was a 4th axis) to his rather expensive CNC. The upgrade was going to cost him several thousand dollars. Sigh, OK he said. The tech came out to do the upgrade. Did he install several bits of hardware, some new PCBs, or what? No, he just inserted a USB key, supplied a password, and basically unlocked the existing extra feature. Proprietary inaccessible systems. I understand the business model, I just don't want to be the victim of it.

    Being locked into an older generation proprietary system is not the way of the future.

    Cheers
    Roger



  6. #86
    Member ger21's Avatar
    Join Date
    Mar 2003
    Location
    Shelby Township
    Posts
    35538
    Downloads
    1
    Uploads
    0

    Default Re: Closed Loop CNC

    DMM DYN2 drives are step/dir.

    Gerry

    UCCNC 2017 Screenset
    [URL]http://www.thecncwoodworker.com/2017.html[/URL]

    Mach3 2010 Screenset
    [URL]http://www.thecncwoodworker.com/2010.html[/URL]

    JointCAM - CNC Dovetails & Box Joints
    [URL]http://www.g-forcecnc.com/jointcam.html[/URL]

    (Note: The opinions expressed in this post are my own and are not necessarily those of CNCzone and its management)


  7. #87
    Member mactec54's Avatar
    Join Date
    Jan 2005
    Location
    USA
    Posts
    15362
    Downloads
    0
    Uploads
    0

    Default Re: Closed Loop CNC

    Quote Originally Posted by RCaffin View Post
    I am sure the Dyna AC servo drive is well made and powerful, but it uses serial comms in a custom format instead of step/dir, and there are no modern NIST-compatible CNC systems which use that format. That means you are locked into ONE of the older generation proprietary systems with little chance of an upgrade to a different brand. It often also means that adding a feature can be difficult or very expensive.

    I remember the complaints of one business owner who wanted to add some feature (I think it was a 4th axis) to his rather expensive CNC. The upgrade was going to cost him several thousand dollars. Sigh, OK he said. The tech came out to do the upgrade. Did he install several bits of hardware, some new PCBs, or what? No, he just inserted a USB key, supplied a password, and basically unlocked the existing extra feature. Proprietary inaccessible systems. I understand the business model, I just don't want to be the victim of it.

    Being locked into an older generation proprietary system is not the way of the future.

    Cheers
    Roger
    Right on the first page of there AC Servo Drive Specification PDF it tells you what and how it can be controlled, they can be controlled, by step/dir Analog Can Modbus and more, they are way ahead of the game, find another system that can use just 4 wires for Encoder commutations, and 2 of the wires are for power and ground, plus use a 8-bit security, for the sent / received encoder information, this was a first in the world to do this

    It is common practice to use a USB to upgrade or open other features in a control, with a Haas machine you only have to enter a code number which they send you for any extras they have to offer

    Many company can connect direct to the servo drives over the internet and see what is wrong if there is a problem, they can correct it or replace / upgrade the software

    Attached Thumbnails Attached Thumbnails Closed Loop CNC-control-method-png  
    Mactec54


  8. #88
    Community Moderator Al_The_Man's Avatar
    Join Date
    Dec 2003
    Location
    Canada
    Posts
    24216
    Downloads
    0
    Uploads
    0

    Default Re: Closed Loop CNC

    Quote Originally Posted by Jim Dawson View Post
    It seems to me that using torque control would be most appropriate for applications requiring maintaining a constant tension on a web material or something like that. The velocity control would seem the be most appropriate for positioning control, letting the drive apply the torque required to maintain minimum following error, like in a CNC machine application.

    Is my thinking incorrect here?
    The velocity mode is outdated now, this is the reason you don't see tach's on DC motors anymore that was required when used with velocity drives and the inner (velocity loop) had to be tuned first and then the outer Controller PID loop.
    Galil recommend torque mode control and this is the way I have always used them, in fact they will drive the most basic version of transconductance (torque mode) amplifier drives using 2 axis sine wave input signals where the third is extrapolated from the other two.
    The down side is they take up two command outputs from the Card instead of the 'regular' single per axis.
    Also A-M-C recommend the torque mode for their drives using CNC/PID closed loop.
    When Using a flexible drive such as A-M-C ±10vdc analogue input in torque mode, the amplifier cannot be used in open loop mode like you can the velocity mode, where general purpose open loop motor drive applications are used.
    The beauty of these drives is they require minimum tuning and do not contain any 'Intelligent' electronics or require encoder input.
    Al..
    .

    CNC, Mechatronics Integration and Custom Machine Design

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


  9. #89
    Member
    Join Date
    Jun 2010
    Location
    Australia
    Posts
    4252
    Downloads
    4
    Uploads
    0

    Default Re: Closed Loop CNC

    Yes, I can see that the DYN2 drives have had Step/Dir added onto their interface. Technically, this probably means they have bypassed much of the older control electronics. This is not being 'way ahead of the game', this is scrabbling to catch up with modern trends while trying to preserve their older market for replacement parts. Seems reasonable to me.

    A really major FAULT of the older techniques is that none of them were suited to 4, 5 or 6 axis synchronous control at high speed. Each axis has a different acceleration parameter, and trying to figure out what serial commands to send to each axis to allow for the differing accelerations verged on the impossible. Synchronising them all was equally difficult. Obsolete.

    Proof of this? The extreme cost of such multi-axis machines from the older vendors, PLUS the existence of Coordinate Measuring Machines being used to check the shape of parts which had been made on the older CNCs. You have to check up on those parts when multi-axis operations had been performed because you could not be sure of the result. Contrast this with the product of a modern CNC using centrally-controlled multi-axis Step/Dir: if the program runs you KNOW the part will be as the program directed. You do not need a CMM with a modern CNC. (You may need to check your program - but that has always been with us.)

    find another system that can use just 4 wires for Encoder commutations,
    Are you aware that almost any hobby CNC using Mach3 or the other modern CNC programs uses encoders with a 4-wire communication link? +5 V, 0 V, A and B. This has been stock standard on modern CNCs for decades. You can go more sophisticated if you want, with A&A', B&B', and even I&I' for the index pulse, but we do not need the extra. FOUR wire is the modern standard.

    use a 8-bit security, for the sent / received encoder information
    That is not 'security' in the modern sense, it is just a data checksum to handle noise on the serial port. With proper system design (wiring, shielding and grounding), we simply do not need that - and we don't need serial comms either. Obsolete. Ah yes: modern motor drivers do check the encoder signals for validity as well. A failed link will halt the system.

    It is common practice to use a USB to upgrade or open other features in a control,
    Oh, I can believe that. I just don't like the business model. If I buy Mach3/4, LinuxCNC, UCCNC or other modern control SW, I get ALL the features right from the start at no extra cost. This included fully synchronised 6-axis operation (at least with Mach3) with full synchronous control over acceleration/deceleration as well. It's all just SW.

    Many company can connect direct to the servo drives over the internet
    Which is a really bad idea as it means the internet can interrupt the CNC controller for an unknown period of time, crashing everything. By and large, we run our modern CNCs with NO internet connection at all. I am aware that some do leave their machines connected: this is possible if you know exactly what you are doing, but it is risky. An 8 GByte USB key makes a very good file transfer mechanism.

    It will take another decade or two for most the obsolete CNC machines to be replaced with modern designs. I know that. Owners will want to maintain their technically obsolete machines until they are mechanically worn out, and maybe they will just resist making the change for some time too. I can understand the commercial problems. I do note that quite a few business owners today are however slowly stripping out the older controllers and drives and replacing them with modern versions. Clearly, they must think that the longer-term benefit is worth the short-term cost.

    Cheers
    Roger



  10. #90
    Member mactec54's Avatar
    Join Date
    Jan 2005
    Location
    USA
    Posts
    15362
    Downloads
    0
    Uploads
    0

    Default Re: Closed Loop CNC

    Quote Originally Posted by RCaffin View Post
    Yes, I can see that the DYN2 drives have had Step/Dir added onto their interface. Technically, this probably means they have bypassed much of the older control electronics. This is not being 'way ahead of the game', this is scrabbling to catch up with modern trends while trying to preserve their older market for replacement parts. Seems reasonable to me.
    Roger
    From the very first AC Servo Drive DMM have ever made, they have always had Step / Dir, it's not something that has just been add, how do you come up with this rubbish


    Quote Originally Posted by RCaffin View Post
    Are you aware that almost any hobby CNC using Mach3 or the other modern CNC programs uses encoders with a 4-wire communication link? +5 V, 0 V, A and B. This has been stock standard on modern CNCs for decades. You can go more sophisticated if you want, with A&A', B&B', and even I&I' for the index pulse, but we do not need the extra. FOUR wire is the modern standard. Roger
    Not sure what you are thinking about, but they where granted a Patent for this, and the first to file for this technology, all I can think of is that you have no understanding of what we are even talking about

    I not going to waste my time with the rest of your post, as it is just more of your uneducated ramblings

    Mactec54


  11. #91
    Member
    Join Date
    Jun 2010
    Location
    Australia
    Posts
    4252
    Downloads
    4
    Uploads
    0

    Default Re: Closed Loop CNC

    A Patent ... well, I am afraid that does not mean very much these days. If you put enough waffle in the application, avoid any mention of competing technologies, and hassle the poor examiner, you get away with it. Having inspected their web site, it seems that the Patent you may be referring to is for a magnetic encoder with a 4-wire serial comms link (probably RS232C). They do mention 'interpolation' in their advertising - as I explained earlier. Just digitisation of the initial analog sine waves. Ho hum.

    The control algorithm uses only 3 parameters to adjust gain and internally optimizes position accuracy and torque ripple during real time operation. The only tuning method of this kind in the industry.
    The diagram shows a standard s-plane feedback diagram. The rest of the world calls this PID tuning - with 3 parameters.

    Regardless, most of us currently use 4-wire encoder connections with PID tuning.

    Have you looked at the patent war between Sony and Apple? Round corners on the case no less. Conflicting patents were granted! Amazing stuff.

    uneducated ramblings - I have been working in this general area since the mid-80s, successfully developing robotic systems. Ah well.

    Cheers
    Roger (PhD)



  12. #92
    Community Moderator Al_The_Man's Avatar
    Join Date
    Dec 2003
    Location
    Canada
    Posts
    24216
    Downloads
    0
    Uploads
    0

    Default Re: Closed Loop CNC

    Quote Originally Posted by RCaffin View Post

    find another system that can use just 4 wires for Encoder commutations,
    Are you aware that almost any hobby CNC using Mach3 or the other modern CNC programs uses encoders with a 4-wire communication link? +5 V, 0 V, A and B. This has been stock standard on modern CNCs for decades. You can go more sophisticated if you want, with A&A', B&B', and even I&I' for the index pulse, but we do not need the extra. FOUR wire is the modern standard.
    Don't really agree with that!
    The differential feature is is used with quadrature encoders for RS485 transmission in order for to prevent EMI or other interference affecting the signal. It is converted at both ends so the differential signals only really exist for transmission purposes.
    If I have no choice to use a single ended type where it is already supplied, I just add a RS485 RX/TX IC each end.
    Al.

    Last edited by Al_The_Man; 07-07-2018 at 02:53 PM.
    CNC, Mechatronics Integration and Custom Machine Design

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


  13. #93
    Member mactec54's Avatar
    Join Date
    Jan 2005
    Location
    USA
    Posts
    15362
    Downloads
    0
    Uploads
    0

    Default Re: Closed Loop CNC

    Quote Originally Posted by RCaffin View Post
    A Patent ... well, I am afraid that does not mean very much these days. If you put enough waffle in the application, avoid any mention of competing technologies, and hassle the poor examiner, you get away with it. Having inspected their web site, it seems that the Patent you may be referring to is for a magnetic encoder with a 4-wire serial comms link (probably RS232C). They do mention 'interpolation' in their advertising - as I explained earlier. Just digitisation of the initial analog sine waves. Ho hum.

    The control algorithm uses only 3 parameters to adjust gain and internally optimizes position accuracy and torque ripple during real time operation. The only tuning method of this kind in the industry.
    The diagram shows a standard s-plane feedback diagram. The rest of the world calls this PID tuning - with 3 parameters.

    Regardless, most of us currently use 4-wire encoder connections with PID tuning.

    Have you looked at the patent war between Sony and Apple? Round corners on the case no less. Conflicting patents were granted! Amazing stuff.

    uneducated ramblings - I have been working in this general area since the mid-80s, successfully developing robotic systems. Ah well.

    Cheers
    Roger (PhD)
    I wondered when you would quote your education again, those that pull the education card are a lost cause, you were a researcher, nothing more, you can pretend all you like, you have absolutely no substance in your posts but negative comments which tells us who you really are

    They don't advertise the many Patents they have, they just make a lot of money selling there technology to other Servo Drive manufactures, so you have no clue how and what they are doing, if you think they have on there website information on there servo drives workings you are a very mistaken

    Another feature they have is Adaptive Tuning on the fly, another first, the PID tuning is not required with there Adaptive Tuning,

    It's Amazing how you guys dwell on your past, and have no knowledge of what is happening around you, lack of exposure=lack of knowledge, which you are showing us in your posts

    Mactec54


  14. #94
    Member mactec54's Avatar
    Join Date
    Jan 2005
    Location
    USA
    Posts
    15362
    Downloads
    0
    Uploads
    0

    Default Re: Closed Loop CNC

    Quote Originally Posted by samco View Post
    BTW - with linuxcnc you can easily get software step rates into the 40khz. External interface hardware step rates are in the mhz...
    You need around 200Khz to run a modern servo at max speed the higher the rpm of the servo motor the higher the KHz you need to run them, so 40Khz is not going to do a very good job even with using electronic gearing, I though LinuxCNC had a 50Khz output

    Mactec54


  15. #95
    Member samco's Avatar
    Join Date
    Jul 2003
    Posts
    1753
    Downloads
    2
    Uploads
    0

    Default Re: Closed Loop CNC

    Again - with linuxcnc you don't have to use step/dir to run servos... So there isn't a problem of pulse rate.... You push dmm - I assume you have skin in the game. Linuxcnc is not limited to step/dir solutions.

    Software step generation speed is totally dependent on the computer. I touted 40khz as a average. Some can do faster - some slower. It is something you test. External interface hardware (mesa, pico - whatever) can do mhz step rates.

    Quote Originally Posted by mactec54 View Post
    You need around 200Khz to run a modern servo at max speed the higher the rpm of the servo motor the higher the KHz you need to run them, so 40Khz is not going to do a very good job even with using electronic gearing, I though LinuxCNC had a 50Khz output




  16. #96
    Member mactec54's Avatar
    Join Date
    Jan 2005
    Location
    USA
    Posts
    15362
    Downloads
    0
    Uploads
    0

    Default Re: Closed Loop CNC

    Quote Originally Posted by samco View Post
    Again - with linuxcnc you don't have to use step/dir to run servos... So there isn't a problem of pulse rate.... Linuxcnc is not limited to step/dir solutions. .
    There is a problem if a Hobby user wants to run servo's direct, like you posted using step/dir, 40Khz is ok for steppers as most Hobby users are using, but for some reason they don't jump through hoops to use LinuxCNC, even though it has got quite user friendly

    Push Dmm no, I have been installing them for 10 years side by side with Yaskawa, so know them very well, and have watched them grow into a great company

    Mactec54


  17. #97
    Member samco's Avatar
    Join Date
    Jul 2003
    Posts
    1753
    Downloads
    2
    Uploads
    0

    Default Re: Closed Loop CNC

    Yes - for software step generation. for $89 you can get a mesa external interface card and use your existing printer port bobs (emulates 2 printer ports). But you get mhz step rates (and expandable). well within hobby price range.

    Quote Originally Posted by mactec54 View Post
    There is a problem if a Hobby user wants to run servo's direct, like you posted using step/dir, 40Khz is ok for steppers as most Hobby users are using, but for some reason they don't jump through hoops to use LinuxCNC, even though it has got quite user friendly

    I have not used any servo drives that take step/dir. I am sure they work just great. (I just like more control )

    Quote Originally Posted by mactec54 View Post
    Push Dmm no, I have been installing them for 10 years side by side with Yaskawa, so know them very well, and have watched them grow into a great company




  18. #98
    Member mactec54's Avatar
    Join Date
    Jan 2005
    Location
    USA
    Posts
    15362
    Downloads
    0
    Uploads
    0

    Default Re: Closed Loop CNC

    Quote Originally Posted by samco View Post
    Yes - for software step generation. for $89 you can get a mesa external interface card and use your existing printer port bobs (emulates 2 printer ports). But you get mhz step rates (and expandable). well within hobby price range.




    I have not used any servo drives that take step/dir. I am sure they work just great. (I just like more control )
    That sounds like a it would be a good board to have for anyone that wants to use LinuxCNC with step / dir and needs a higher step rate

    You can have the same control, no matter what you use, you are using Analog because that is what your system drives required, any quality Servo Drives use all the methods for controlling them, the best is Fiber Optic's, but not many servo drives support it's use, Modbus, I'm starting to use, Can similar to Modbus, then comes Analog and Step /Dir Etc

    Mactec54


  19. #99
    Member rvectors's Avatar
    Join Date
    May 2019
    Posts
    16
    Downloads
    0
    Uploads
    0

    Default Re: Closed Loop CNC

    I was all set on which motors I'd use (57N-DHT/640-DST), then after reading this thread, I'm back to being undecided.

    (brief background. I have absolutely no prior CNC experience, and any knowledge is that of busily researching this past week. So total greenhorn, gotta start somewhere)


    I had decided on DMM servos based on the build I was after, and its future uses. For accuracy, speed, noise, power, smoothness and being 1/2-1/3rd cheaper than the next (servo) marque brands up.


    Now, after doing more research (including thoughts triggered by this thread), with a good driver (eg. Gecko GR214V) with micro-stepping, resonance compensation etc, steppers can get comparable accuracy to the DMM, smoothness and noise level? The key servo advantage remaining on my list, is then speed, and maybe peak torque.
    Am I wrong? Considering the revelation that max resolution/accuracy is effectively 4000 (guessing physical limitation of motor design), despite the encoder resolution.

    Having no CNC experience I cannot gauge how much the speed equation is worth, considering steppers setup cost £300-500 vs DMM approx £1500 (imported)... and I'd likely have to add some constant spring/counter balance arrangement, on Z without a brake (another $100)



    Doing some calculations, based on the CNC mechanics I'm looking at.
    Gantry 60-75kg
    Y 2x 16x10mm BS 1330mm
    X 16x10mm BS 1040mm
    one motor each, both driven nut via belt 1.25x, fixed ends
    Z 16x5mm 1:1 ratio 190mm


    The Y inertia ratio will be between 2-4x. The servos should provide great linear acceleration, and permissible 'potential' rapids of 1181IPM. The machine will be in a small room 14ftx10ft, and monitored by me, hence my interest in noise level and speed factor too. My use case will be wood, non ferrous metals, maybe a laser upgrade in the future. I understand that the stepper's torque curve drops off precipitously (useful speeds <750-900RPM), but can the dmm really make a difference in job speed. Is anyone able to quantity this when they've switched to servos from steppers? Are there any other advantages? I believe the servos would be better if I were to fit a laser at some point.

    I saw the video posted by mactec54, showing load inertia/response time but I didn't understand them running a test of 1hz-20hz range. If I'm not mistaken, to run the servos near their rated speed, with my intended machine, the step frequency would be anything up to 200khz. I'm assuming these two are different operations.



    I hope my question fits in with the thread. The discussion is relevant to my new indecision.
    Thanks in advance.



Page 5 of 5 FirstFirst ... 2345

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

Closed Loop CNC

Closed Loop CNC