Delta servo drives and servos. - Page 2


Page 2 of 5 FirstFirst 12345 LastLast
Results 21 to 40 of 81

Thread: Delta servo drives and servos.

  1. #21
    Member
    Join Date
    Feb 2017
    Location
    Estonia
    Posts
    13
    Downloads
    0
    Uploads
    0

    Default Re: Delta servo drives and servos.

    yes, I think exactly the same. my coal is to find OPTIMAL product. OPTIMAL=best performance and $$$ relation. yaskawa maybe have sligth advantage on support or I dont know where, but if it costs 3 times more. then it is far far from optimal from my solution.



  2. #22
    Member
    Join Date
    Jan 2005
    Location
    USA
    Posts
    11543
    Downloads
    0
    Uploads
    0

    Default Re: Delta servo drives and servos.

    Quote Originally Posted by joeavaerage View Post
    Hi,



    That is not correct in the case of the Delta A2-L series servos, they have two encoder inputs on the drive, one the regular encoder mounted on the servo
    and another mounted on the load. The attached pic is of the A2 user manual and describes the connection CN5 being for a linear encoder.
    The regular encoder is connected to CN2 per the attached.

    Thus BOTH encoders are connected to the drive....not the motion controller.

    Craig
    That's a nice feature they have, it is quite normal to do it the way I said with, hardware and the control, the control would still have to be part of the closed loop or it would not work, what is the linear encoder max resolution this may be where the limitation is by closing the duel loop in the drive

    Mactec54


  3. #23
    Member
    Join Date
    Jan 2005
    Location
    USA
    Posts
    11543
    Downloads
    0
    Uploads
    0

    Default Re: Delta servo drives and servos.

    Quote Originally Posted by joeavaerage View Post
    Hi,
    and the control parameters to set it all up are in the pics attached.

    Craig
    It's just the same as any modern top quality servo drive can do the duel encoder feed back but it is not widely used because of the added cost

    Have you used or installed this to know how well it works or even seen a machine where it is being used, I know of a few machines that use duel feed back but they cost between 250K to 500K

    I don't think to many hobby guys will be using a duel closed loop anytime soon a lot can dream

    Mactec54


  4. #24
    Member
    Join Date
    Jan 2005
    Location
    USA
    Posts
    11543
    Downloads
    0
    Uploads
    0

    Default Re: Delta servo drives and servos.

    Quote Originally Posted by joeavaerage View Post
    fasttobuy2012 | eBay Stores

    I found this Ebay store to be if not the lowest price it was close to. They are responsive and can source direct from Delta.
    If you buy Delta servos you want to use the Delta tuning software in which case you need the programming cable. I bought
    a genuine Delta accessory cable, you'll want one. Craig
    They are manufactured in China as for a lot of these servo drives and motors, that is why you were able to buy them cheap, but if they work as per the spec's then you have a good deal

    Mactec54


  5. #25
    Member
    Join Date
    Feb 2017
    Location
    Estonia
    Posts
    13
    Downloads
    0
    Uploads
    0

    Default Re: Delta servo drives and servos.

    Quote Originally Posted by mactec54 View Post
    It's just the same as any modern top quality servo drive can do the duel encoder feed back but it is not widely used because of the added cost

    Have you used or installed this to know how well it works or even seen a machine where it is being used, I know of a few machines that use duel feed back but they cost between 250K to 500K

    I don't think to many hobby guys will be using a duel closed loop anytime soon a lot can dream

    No I am not tested it jet. but I already ordered encoder and probably next month I order servo.
    and my plan:
    1. to make some test platform and test how good it works and what results I get. if all works well and results are good.
    2. mount first motor to my x axis and test it on real action, if it is all good, then i order some more motors and install them on y axis aswell.

    I know this is not maybe not the most practical thing in the world, but I realy like to tinkering, and my cnc makes some money, so on some time it pays back.



  6. #26

    Default Re: Delta servo drives and servos.

    Hi,

    It's just the same as any modern top quality servo drive can do the duel encoder feed back but it is not widely used because of the added cost
    I paid $450USD (excluding shipping) for 750W B2 series serovs, drives and cables, that is with a 160,000 count per rev encoder.
    I did not know at that time about the dual encoder reading feature of the A2 series (1,280,000 count per rev encoder), and had I done so I would
    have bought them instead as they are only $500USD (excluding shipping) for 750W servo, drive and cables.

    So the dual reading feature can be had for only really a small amount extra.

    I use an Ethernet SmoothStepper which is a non-feedback step/direction controller and I could not use the traditional scheme that you outline,
    namely the linear scale loop is closed by the controller not the drive.

    This is an example of my obersvation that 'servos are becoming ever more sophisticated such that CNCmotion controllers are shedding functions
    which are now performed by the servo drive'. The ultimate is distributed motion control like EtherCAT, each servo drive becomes its own motion controller
    so the PC is the trajectory planner only, it has no need of a motion control.

    Craig



  7. #27
    Member
    Join Date
    Jan 2005
    Location
    USA
    Posts
    11543
    Downloads
    0
    Uploads
    0

    Default Re: Delta servo drives and servos.

    Quote Originally Posted by joeavaerage View Post
    Hi,



    I paid $450USD (excluding shipping) for 750W B2 series serovs, drives and cables, that is with a 160,000 count per rev encoder.
    I did not know at that time about the dual encoder reading feature of the A2 series (1,280,000 count per rev encoder), and had I done so I would
    have bought them instead as they are only $500USD (excluding shipping) for 750W servo, drive and cables.

    So the dual reading feature can be had for only really a small amount extra.

    I use an Ethernet SmoothStepper which is a non-feedback step/direction controller and I could not use the traditional scheme that you outline,
    namely the linear scale loop is closed by the controller not the drive.

    This is an example of my obersvation that 'servos are becoming ever more sophisticated such that CNCmotion controllers are shedding functions
    which are now performed by the servo drive'. The ultimate is distributed motion control like EtherCAT, each servo drive becomes its own motion controller
    so the PC is the trajectory planner only, it has no need of a motion control.

    Craig
    These features can't be working very well as there are no CNC Manufacturers using it in the drive only, they have duel encoder feed back but are all closing the loop in the control

    Yes I have been using EtherCat and fiber optics since 1999 with my CNC Controls

    When you close the loop in the control Both Encoders are controlled not just the linear scale

    You will find if you buy from delta direct, drive and motor you won't have much change out of around $1200 the Chinese made are cheap but not all are created equal

    Mactec54


  8. #28

    Default Re: Delta servo drives and servos.

    Hi,
    when I bought the B2 series servos they did not have a braked one and yet I wanted one. The Ebay company (fast-to-buy) I bought from
    got one in for me, they said from Delta, maybe a warehouse, and within18 hours. I presume therefore that they deal fairly closely with Delta.
    While Delta is a Taiwanese company most manufacturing is done in China

    In subsequent emails with them, they now always try to tell me about their latest special....., they claim that on the basis of Delta sale volumes in the
    previous year that they had been offered a better discount and they want to pass (some) of that on to me, presumably to induce me to buy more of them.

    For example I paid $438USD (excluding shipping) for a 400W B2 series servo/drive and cables in August last year. They offered me the exact same
    model servo/drive and cables for $387USD just before the COVID thing blew up.

    I cannot, to my knowledge, buy direct from Delta, but I suspect the company I deal with is reasonably close to the source and all factory brand new
    with Delta warranty cards etc.

    I have done a little research and found that Schnieder Electrical offer a plugin module that allows their LUMIEX32 servo drive to become dual
    sensing. Allen Bradley KINETIX 5700 drives also offer dual sensing.

    I rather suspect that dual sensing and that loop being closed at the drive is more widespread than you imagine.

    Craig

    Attached Thumbnails Attached Thumbnails Delta servo drives and servos.-lexium32servodrive-png  


  9. #29

    Default Re: Delta servo drives and servos.

    Quote Originally Posted by mactec54 View Post
    I doubt that you would ever have a need for a braking ( regenerative ) resistor unless you have a mass that you are trying to stop in a hurry which could be the case when using high pitch Ballscrews

    With quality servo drives the loop is closed in the servo drive, and can be corrected by the control if out of position, but not normally not needed to be controlled by the control for what you are building

    Even with 3:1 the 400w servos will fail to perform with a 2020 Ballscrew even 750w would be at there max to run a 2020 Ballscrew so if you are looking for performance nothing less than 750w is going to do it for you

    Most servo drives can be setup ( Programed ) by using a USB cable suited for the Servo Drive

    How good are the Delta Servo system they are in the mid range in the market place, if you want a comparison Yaskawa is the number ( 1 )
    Hi.

    About the servo performance, I am building a machine with a Z-axis weight of 30kg, an X-axis weight of 60kg and gantry weight of approximately 100kg (including X and Z-axis).

    I will be running 1605 ballscrew on the Z-axis, 2:1 with a braked servo. The X-axis will be run with 2020 and 2:1. The Y-axis will be dual motor and dual 2020 screws with 2:1 ratio.

    But as you mentioned here that even 3:1 ratio with a 750W servo would be the bare minimum, I am having second thoughts about my desing. Originally I was planning 2010 screws on X and Y, with 400w servos and 2:1, but now I want more speed.

    Don't you think that this would work well:

    - 750W (maybe even 400W?) with 2:1 with 1605 for 30kg Z-axis
    - 750W with 2:1 for 60kg with 2020
    - 2x750W with 2:1 with 2x2020

    Here is my build log if you are interested: https://r.tapatalk.com/shareLink/top...ink_source=app

    Skickat från min SM-A530F via Tapatalk



  10. #30

    Default Re: Delta servo drives and servos.

    Hi,
    my new build mill has axis beds of 75kg. Thus I would expect the Y axis partiularly to be required to accelerate the X axis bed, linear rails,
    ballscrew and servo, vice and workpiece AND control cutting forces. I anticipate the mass alone will be about 110kg. I am planning
    direct coupling the servo to the 32mm diameter 5mm pitch screw.

    I am of the opinion, and have calculations to support it, that it will work. The ratio of the moments of inertia between the servo armature and the ballscrew
    and attached beds/mass suggest that I may have some fun trying to tune the combination. A medium inertia servo would be better in this regard.

    Follow this calculation:

    750W servo, rated continuous torque =2.4Nm
    Effective radius of a 32mm ballscrew= 16mm =0.016m
    Force@0.016m=2.4/0.016
    =150N

    Mechanical advantage of a 32mm diameter screw or 5mm pitch is:
    mechanical advantage=32 x pi / 5
    =20

    Thrust ( at rated torque)= 20 x 150
    =3kN

    Lets say that half of that thrust is available to accelerate the axis and the remainder overcomes friction and controls cutting forces:
    Accelerating thrust= 3kN x 0.5
    =1.5kN
    F=M x A or A=F /M
    acceleration =1500/110
    =13.6 m/s2

    The acceleration of my 110kg axis could be as high as 1.3g.....and thats only rated torque, if I allow axis tuning to use the the overlaod torque
    I could accelerate at close to 4g!!!, and that still assumes ONLY ONE HALF of the torque is available for axis movement. More likely
    3/4 to 7/8 in actual practice.

    In short a 750W direct coupled servo is way more power than I can really use, 400W would still be more than adequate, but 750W servos
    were only another $30 each so I got them.

    I can have G1's of 15m/min (3000 rpm) and G0's of 25m/min (5000rpm).

    Yes you can get your 3000rpm Delta servo to do 5000 rpm, using a feature called feild weakening. It loses some of it top end
    torque, but hey I've got swags to spare!

    I think you'll have more power and speed that you can possibly believe with the servos you have listed.

    Craig

    Last edited by joeavaerage; 04-19-2020 at 05:03 PM.


  11. #31

    Default Re: Delta servo drives and servos.

    Quote Originally Posted by joeavaerage View Post
    Hi,
    my new build mill has axis beds of 75kg. Thus I would expect the Y axis partiularly to be required to accelerate the X axis bed, linear rails,
    ballscrew and servo, vice and workpiece AND control cutting forces. I anticipate the mass alone will be about 110kg. I am planning
    direct coupling the servo to the 32mm diameter 5mm pitch screw.

    I am of the opinion, and have calculations to support it, that it will work. The ratio of the moments of inertia between the servo armature and the ballscrew
    and attached beds/mass suggest that I may have some fun trying to tune the combination. A medium inertia servo would be better in this regard.

    Follow this calculation:

    750W servo, rated continuous torque =2.4Nm
    Effective radius of a 32mm ballscrew= 16mm =0.016m
    Force@0.016m=2.4/0.016
    =150N

    Mechanical advantage of a 32mm diameter screw or 5mm pitch is:
    mechanical advantage=32 x pi / 5
    =20

    Thrust ( at rated torque)= 20 x 150
    =3kN

    Lets say that half of that thrust is available to accelerate the axis and the remainder overcomes friction and controls cutting forces:
    Accelerating thrust= 3kN x 0.5
    =1.5kN
    F=M x A or A=F /M
    acceleration =1500/110
    =13.6 m/s2

    The acceleration of my 110kg axis could be as high as 1.3g.....and thats only rated torque, if I allow axis tuning to use the the overlaod torque
    I could accelerate at close to 4g!!!, and that still assumes ONLY ONE HALF of the torque is available for axis movement. More likely
    3/4 to 7/8 in actual practice.

    In short a 750W direct coupled servo is way more power than I can really use, 400W would still be more than adequate, but 750W servos
    were only another $30 each so I got them.

    I can have G1's of 15m/min (3000 rpm) and G0's of 25m/min (5000rpm).

    Yes you can get your 3000rpm Delta servo to do 5000 rpm, using a feature called feild weakening. It loses some of it top end
    torque, but hey I've got swags to spare!

    I think you'll have more power and speed that you can possibly believe with the servos you have listed.

    Craig
    Thanks for the thorough explanation. I was asking because mactec54 claimed the power being way too low. I am not saying that he is wrong but according to the calculations, even 400w would be sufficient!

    Skickat från min SM-A530F via Tapatalk



  12. #32

    Default Re: Delta servo drives and servos.

    Here are the calculations done with Sigma Select (Sigma software for calculating Yaskawa servo sizes).

    According to the reports when using 400W servos (see attachments), peak toque would be at the rated torque on the X-axis (0,1s acceleration, 30m/s, 2010 screw, 2:1 ratio) and a little bit over on the Z-axis (0,1s acceleration, 7,5m/s, 1605 screw, 2:1 ratio). The Y-axis I am pretty confident that will be fine with 2x 400W motors.

    Attached Files Attached Files


  13. #33

    Default Re: Delta servo drives and servos.

    Hi,
    another way of looking at it is to use 750W, being approx double the torque of 400W units and drop the 2:1 belt reduction.

    Cant say about Yaskawa but 750W Deltas are only about $30 more than 400W ones. I bet the belt and pulleys cost more than $30.

    Craig



  14. #34

    Default Re: Delta servo drives and servos.

    Quote Originally Posted by joeavaerage View Post
    Hi,
    another way of looking at it is to use 750W, being approx double the torque of 400W units and drop the 2:1 belt reduction.

    Cant say about Yaskawa but 750W Deltas are only about $30 more than 400W ones. I bet the belt and pulleys cost more than $30.

    Craig
    That is a good point. I made the calculations for direct drive (coupling) and to my surprise the peak torque is actually lower (see attachment)! Seems like the diameter of the pulleys increases the inertia a lot! With 1:1 coupling ratio, I will achieve same speed with 50% lower rpm (only 1500rpm). Then there is room for ever faster speeds, given that the ballscrews wont whip!

    According to the reports, 400W will have more than enough torque. What do you think? Can these calculations be trusted at all?

    Attached Files Attached Files
    Last edited by NordicCnc; 04-20-2020 at 08:42 AM.


  15. #35

    Default Re: Delta servo drives and servos.

    Hi,

    What do you think? Can these calculations be trusted at all?
    I'd love to see the formulas that are used then I would know definitively. Having said that if we can't trust Yaskawa to do these simple calculations
    then who can you trust? So yes I trust them.

    Before I committed to buy three (one braked) 750W servos for my new build I bought a 400W B2 series to expeiment with. It is a 23 size which is
    a bolt in replacement for the stepper/gearbox combo in my existing mini-mill. To say I was impressed is an understatement. The rated torque of
    a 3000rpm 400W servo is 1.27Nm, pretty modest. Yet actually in service it appears to have considerably greater capacity than the spec suggests.

    I came to the conclusion that a servos prodigious overload torque lends a servo as 'a giant killer'. I am not suggesting that you should design to
    overload but if you design to rated torque the servo will perform effortlessy in practice.

    You have calculations that show that 400W is ample and my experience informs me a 400W servo will outperform your expectation.

    As I posted earlier I'm using 32mm BNFN double nut C5 ballscrews and they have a stack height of 95mm. I needed, at least the X axis servo
    to occupy the intersitial space, ie I could go to 86mm (34 size) servos, which is the readily available low inertia 750W model.
    There is a medium inertia model, less readiliy available of 110mm, however I decided that the complication of trying to provide the extra
    room for it counted against it.

    As I posted earlier also that a 750W B2 series Delta servo cost only $30USD more than a 400W servo....so I got the 750's. They are truly
    awesome.

    I will not know how well or otherwise they tune up until the machine is (near) complete, I have used low inertia servos where medium inertia
    would be better. It will be a least a couple of months before I'm ina position to asses how significant it is.

    Craig



  16. #36
    Member
    Join Date
    Jan 2005
    Location
    USA
    Posts
    11543
    Downloads
    0
    Uploads
    0

    Default Re: Delta servo drives and servos.

    Quote Originally Posted by NordicCnc View Post
    That is a good point. I made the calculations for direct drive (coupling) and to my surprise the peak torque is actually lower (see attachment)! Seems like the diameter of the pulleys increases the inertia a lot! With 1:1 coupling ratio, I will achieve same speed with 50% lower rpm (only 1500rpm). Then there is room for ever faster speeds, given that the ballscrews wont whip!

    According to the reports, 400W will have more than enough torque. What do you think? Can these calculations be trusted at all?
    You have plugged in 2:1 ratio you said you where wanting direct drive which is not going to happen with a 20mm pitch ballscrew and a 400w motor you need a 750W motor and even that will be maxed out with a 20mm pitch Ballscrew direct drive

    Mactec54


  17. #37

    Default Re: Delta servo drives and servos.

    Hi,
    my calculation was done with a 32mm ballscrew with 5mm pitch. the mechanical advantage is (within a very close approximation):

    mech avd. = circumference / pitch
    = 32 x pi / 5
    =20.1

    Most mechanical enginners will recognise that I have made an approximation here. The precise formula is:
    mech adv. = 1 / sin(tan-1 pitch / circumference)
    =1 / sin( tan-1 (5 / 32 x pi)
    =1 / sin ( tan-1 (0.04973))
    =1/sin (2.847 degrees)
    =1/ 0.0496
    =20.13

    So the difference between the approximation and the precisce formula are very very close indeed. As the pitch to diameter ratio decreases however
    the approximation gets worse. Lets takes NordicCnc's example of 20mm diameter and 20mm pitch.

    The approximation is:
    mech adv.= circumference/ pitch
    =20 x pi /20
    = 3.141
    The precise formula is:
    mech adv. =1/sin (tan-1(pitch /circuference)
    = 1/sin (17.65 degrees)
    =3.296

    So the variation between the approximation and the precise forumla is 5%

    Let say we have a servo (or other motor) that applies a 1Nm torque to NordicCnc's 20mm diameter 20mm pitch screw:
    The force at 10mm radius:
    Force @0.01m=1/0.01
    =100N
    With a mechanical advantage ( previously calculated) of 3.296 means a thrust of:
    thrust= 100 x 3.296
    =329.6N

    Note also the the Yaskawa calculation applies a 'ballscrew efficiency' of 0.9, so in interests of comparison lets do the same:
    thrust =329.6 x 0.9
    =296.6

    If NordicCnc's gantry weighs 60 kg then the acceleration is:

    accel= force/ mass
    =296.6/ 60
    =4.944m/s2
    or 1/2g, not shabby for a hobby machine but not great either, and this is for a 1Nm input.

    A 400W servo has 1.27Nm, so an accel of 6.2m/s2 results.

    A 750W servo has 2.4Nm so an accel of 11.9m/s2 results.

    Clearly with such a coarse pitched screw then a 750W servo is indicated.

    The problem here is that the pitch of the screw is high therefore the mechaincal advantage is low. To counteract the low mechaincal advantage
    a belt reduction is in order. The other alternative is to use a lower pitched screw direct coupled. The same calculation but with 5mm pitch and the
    same 20mm diameter:
    mech adv.=12.5
    thrust (per N applied torque)=1130.4N (allowing ballscrew efficieny of 0.9)
    accel (60kg axis)=18.8m/s2 per N applied torque
    accel (400W servo)=23.9m/s2
    accel (750W servo)=45.2m/s2

    Clearly reducing the pitch of the ballscrew has markedly increased the achievable acceleration.

    This would be my recomendation to NordicCnc, reduce the pitch of your screws to 5mm and then direct couple the servos. As you have established
    400W is adequate, even more than adequate, then 750W is doubly 'more than adequate' and costs as little as $30 extra each.

    Craig

    Last edited by joeavaerage; 04-20-2020 at 10:11 PM.


  18. #38

    Default Re: Delta servo drives and servos.

    Quote Originally Posted by mactec54 View Post
    You have plugged in 2:1 ratio you said you where wanting direct drive which is not going to happen with a 20mm pitch ballscrew and a 400w motor you need a 750W motor and even that will be maxed out with a 20mm pitch Ballscrew direct drive
    Actually no. If you check page 2 on respective report you will find that the reports named Z- and X-axisReport have 2 in gear ratio. These reports are for 2:1 pulley transmission. If you check the other 2 report named Z- and X-axis direct driveReport, you will find that the gear ratio is one. These 2 report are from the direct drive with coupling calculations.



  19. #39

    Post Re: Delta servo drives and servos.

    Quote Originally Posted by joeavaerage View Post
    Hi,
    my calculation was done with a 32mm ballscrew with 5mm pitch. the mechanical advantage is (within a very close approximation):

    mech avd. = circumference / pitch
    = 32 x pi / 5
    =20.1

    Most mechanical enginners will recognise that I have made an approximation here. The precise formula is:
    mech adv. = 1 / sin(tan-1 pitch / circumference)
    =1 / sin( tan-1 (5 / 32 x pi)
    =1 / sin ( tan-1 (0.04973))
    =1/sin (2.847 degrees)
    =1/ 0.0496
    =20.13

    So the difference between the approximation and the precisce formula are very very close indeed. As the pitch to diameter ratio decreases however
    the approximation gets worse. Lets takes NordicCnc's example of 20mm diameter and 20mm pitch.

    The approximation is:
    mech adv.= circumference/ pitch
    =20 x pi /20
    = 3.141
    The precise formula is:
    mech adv. =1/sin (tan-1(pitch /circuference)
    = 1/sin (17.65 degrees)
    =3.296

    So the variation between the approximation and the precise forumla is 5%

    Let say we have a servo (or other motor) that applies a 1Nm torque to NordicCnc's 20mm diameter 20mm pitch screw:
    The force at 10mm radius:
    Force @0.01m=1/0.01
    =100N
    With a mechanical advantage ( previously calculated) of 3.296 means a thrust of:
    thrust= 100 x 3.296
    =329.6N

    Note also the the Yaskawa calculation applies a 'ballscrew efficiency' of 0.9, so in interests of comparison lets do the same:
    thrust =329.6 x 0.9
    =296.6

    If NordicCnc's gantry weighs 60 kg then the acceleration is:

    accel= force/ mass
    =296.6/ 60
    =4.944m/s2
    or 1/2g, not shabby for a hobby machine but not great either, and this is for a 1Nm input.

    A 400W servo has 1.27Nm, so an accel of 6.2m/s2 results.

    A 750W servo has 2.4Nm so an accel of 11.9m/s2 results.

    Clearly with such a coarse pitched screw then a 750W servo is indicated.

    The problem here is that the pitch of the screw is high therefore the mechaincal advantage is low. To counteract the low mechaincal advantage
    a belt reduction is in order. The other alternative is to use a lower pitched screw direct coupled. The same calculation but with 5mm pitch and the
    same 20mm diameter:
    mech adv.=12.5
    thrust (per N applied torque)=1130.4N (allowing ballscrew efficieny of 0.9)
    accel (60kg axis)=18.8m/s2 per N applied torque
    accel (400W servo)=23.9m/s2
    accel (750W servo)=45.2m/s2

    Clearly reducing the pitch of the ballscrew has markedly increased the achievable acceleration.

    This would be my recommendation to NordicCnc, reduce the pitch of your screws to 5mm and then direct couple the servos. As you have established
    400W is adequate, even more than adequate, then 750W is doubly 'more than adequate' and costs as little as $30 extra each.

    Craig
    Thanks for the formulas and examples, this helps a lot to understand the requirements. I have a few questions though:

    1. I think that these formulas are not accounting for inertia of the pulley, ballscrew or coupling at 3000rpm (I might be wrong?)
    2. I think 1/2g is more than enough for a hobby machine (attached excel calculation or picture to visualize this). I would reach full speed in 0,16s in worst case scenario. This is for the Y-axis where I will anyway be using two motors and ballscrews, so theoretically the weight can be calculated as 60kg instead of the used 120kg. That would result in 0,08s.
    3. You are right that the high pitch is a problem. Theoretically using belt reduction or changing from 20mm to 5mm pitch would give way more torque. Belt reduction would be favorable here. The problem with decreasing the pitch is that if you want to achieve the same speed, you'd have to rotate the ballscrew 3000rpm. This is way over a 20mm ballscrews critical speed, at say a length of 1-1,5m

    So the original reason for wanting to use a 750W servo with 1:1 on the 2020 ballscrew is that the Yaskawa calculations (which accounts for inertia) has shown that the inertia from the pulleys in a 2:1 reduction with either 400W or 750W servo, will have a higher peak torque compared to a 1:1 ratio with a coupling.

    - In the 2:1 ratio, the small pulley Ø31mm (20 teeth) on the motor (weighs 0,05kg) spins with 3000rpm.
    - The larger pulley Ø62mm (40 teeth) on the ballscrew (weighs 0,25kg) will have to spin 1500rpm.
    - The coupling Ø45mm it would only have to spin 1500rpm to achieve the same speed given that the 2020 ballscrew is still used. This coupling weighs about 0,27kg.

    My conclusion is that as long as you are going to run the motor at the same RPM, belt reduction is good. In this case where the 750W (increased torque by 100% compared to the 400W servo, which means belt reduction is not needed) servo is only 30€ more than the 400W servo, then the 1:1 drive seems to be favorable.

    Correct me if I am wrong!!

    Attached Thumbnails Attached Thumbnails Delta servo drives and servos.-servo-motor-calculations-jpg  
    Attached Files Attached Files


  20. #40

    Default Re: Delta servo drives and servos.

    Hi,

    1. I think that these formulas are not accounting for inertia of the pulley, ballscrew or coupling at 3000rpm (I might be wrong?)
    You are correct, my calculations are correct for static thrust, but the accelaration calculation is an approximation and in effect ignores rotational inertia.
    With a very heavy axis the inertia is dominated by the linear acceleration of that mass, with a light axis then the rotational inertia of both
    the armature and ballscrew and whatever gearing is employed become significant.

    My understanding is that its common in engineering circles to 'translate' linear momentum to effective rotational inertia and that gives a more accurate
    picture. My own training is in electrical and electronic engineering and so my mechanical engineering calculations tend to be introductory or first
    approach affairs.

    This may be an appropriate time for me to extend my knowledge in this particular field.

    Such an approach would not be acceptable in a professional environment it is (ususally) sufficient to draw conclusions about a system, and that is
    good enough for me.

    2. I think 1/2g is more than enough for a hobby machine (attached excel calculation or picture to visualize this).
    But that makes no allowance for rotational inertia per the above nor does it allow for cutting forces. Note that when was calculating for the axes in
    my machine I allowed half the thrust be 'consumed' by cutting forces, whereas the 1/2g calculation did not. Note also that I said its not too shabby....
    but also implies thats its not that great either, in fact I would call it borderline. If it turns out your (or my) estimate of those countervailing forces
    are underestimated then once you've spent your money only to find that it performs sub-par.

    I paid nearly $3000NZD for my three servos.......a big, I mean very big chunk of my budget. I absolutely cannot afford to make a mistake here.
    Neither do I believe I have...but time will tell. What I've found over several years I've been fiddling with CNC is that if you get it right and chose
    the right piece of equipment even if its expensive, because its right youll use it and use it and use it again over many years. I've also had
    experience where I'll get something which is underspecified or poor quality because I was trying to save money, only to find that it in no way
    meets my needs and is in effect discarded. Despite being cheap to buy it is in fact the most expensive on the basis that I get no use of it.

    You have the time and a forum in which to conduct your research such that you'll have a very fair idea, even if not 'iron clad' idea about
    how your machine will perform BEFORE you spend your money. Use it well.

    Craig



Page 2 of 5 FirstFirst 12345 LastLast

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

Delta servo drives and servos.

Delta servo drives and servos.

Delta servo drives and servos.