Newbie From NEMA 27 to NEMA 34 and voltage type selection


Page 1 of 2 12 LastLast
Results 1 to 20 of 32

Thread: From NEMA 27 to NEMA 34 and voltage type selection

  1. #1
    Member sn4k3's Avatar
    Join Date
    Mar 2019
    Location
    Alvito
    Posts
    51
    Downloads
    0
    Uploads
    0

    Question From NEMA 27 to NEMA 34 and voltage type selection

    Hello,

    I have an all steel 6090 CNC router with linear guides.

    From NEMA 27 to NEMA 34 and voltage type selection-img_20190808_165340-jpg

    Currently i have this longs 23HS9442B in every axis (XYZ): https://ae01.alicdn.com/kf/HTB1pm9yd...716&width=1000
    Active Current: 4.2A
    Microstepping: 1/16

    And i plan to upgrade with NEMA 34: https://www.aliexpress.com/item/32836484729.html

    Why?
    1) Because i can only hit 2000 mm/m velocity, if i go for 2500 on Y or 3000 on YX stepper will stall and my machine can benefit from more speed, i would like to have 4000 to 5000.
    2) The high temperature, i had to put a 360º heatsink cover on steppers, without them my finger burn in a second can't handle more than 2 seconds, with the heatsinks is much better but still hot.
    3) Closed loop seens better option to have

    My questions are:

    1) Driver support AC and DC, i read somewhere that not regulated AC is better for the driver, i understand that when using DC it will be rectified and filtered when not needed to and also drop a bit of voltage. Now with unregulated AC the voltage can fluctuate with with load, my question is if i use AC how good is the driver with the regulation, mantaining the same voltage over the load and if i should use it pior DC. Put a good capacitor at output is a good pratice?

    2) On ac route i'm considering this to feed the 3 drivers: https://www.aliexpress.com/item/32821396714.html 1000W 220AC to 60AC. 60AC because 70AC seens a bit near the limit, and we know things can go wrong. Seller also recommend 60AC and says 200W is ok per set.
    Is this Toroidal a good option? Also have search arround mouser but only find Toroidal with dual outputs which i don't need.

    3) Motor will try to keep the steps and in a lose condition the driver will compensate, what happen if i send a move command that crash the axis to the limit of the granity and then stop the command? Motor will try to compensate and keep crashing forever or theres a limit?

    4) My current motors in a crash condition will stall and have some kind of brake, with this much powerfull motors can it damage something with it stronger muscle or will behave the same?

    Similar Threads:


  2. #2
    Member
    Join Date
    Nov 2013
    Posts
    4280
    Downloads
    0
    Uploads
    0

    Default Re: From NEMA 27 to NEMA 34 and voltage type selection

    Hi,
    I suspect that the real reason that your steppers can't go as fast as you want is because they have too high inductance (3.8mH).

    The lower the inductance the faster the stepper can go before it loses torque and stalls. Every stepper loses torque the faster it goes.
    A 23 size stepper with 1mH inductance will retain about 40% of its torque at 1000 rpm whereas a 3.8mH stepper of the same size might have only
    10% at 1000rpm.

    To be honest I doubt the 34 size steppers will be any faster than what you have. Sure they have a lot more torque at slow speeds but what about at 1000 rpm?

    Manufacturers of steppers know that first time buyers don't know about inductance so they make cheap high torque steppers but with such a high
    inductance that they can ONLY be used at low speeds. Don't be duped out of your money.....you need LOW LOW LOW inductance.
    For 23/24 size steppers get 1-2mH with 1mH preferred and reject anything over 2mH.
    For 34 size steppers get 2-4mH with 2mH preferred and reject anything over 4mH.

    The classic way to overcome inductance is to use the highest possible voltage driver and power supply you can get. At the current time Gecko's (the
    gold standard for performance and reliability) are rated to 80V. Leadshine AM882's are also rated to 80V and quite a bit cheaper than Geckos.

    To me it matters not whether you have a driver that accepts AC (it rectifies and smooths that to produce DC) or one that accepts DC only, the only difference
    is the voltage available to drive the stepper, the higher the better. My Vexta 5-phase stepper drivers accept 230VAC but rectify and regulate to 150VDC to drive
    the steppers. With 150VDC up their chuff they spin at 2400rpm no trouble!

    There are some advantages of closed loop steppers but not enough to justify the expense. The manufactures claim that they go faster and have more power.
    That's ABSOLUTE RUBBISH. They are when all said and done ARE STILL steppers, they lose torque as they go faster. A closed loop drive DOES NOT CHANGE
    the physics of stepper motors. The manufacturers are using a bulls****t line to get your money.

    If you really want closed loop performance then get AC servos, don't waste your money on closed loop steppers. Delta (Taiwanese manufactured in China) or DMM
    (Canadian manufactured in China) are two very good brands that won't break the bank. There are even cheaper Chinese AC servos much less well known and of less
    certain quality and support but at the price they beat any stepper ever made.

    Craig



  3. #3
    Member
    Join Date
    Nov 2013
    Posts
    4280
    Downloads
    0
    Uploads
    0

    Default Re: From NEMA 27 to NEMA 34 and voltage type selection

    Hi,

    Is this Toroidal a good option?
    Toroidal power supplies are very reliable power sources, way more reliable and forgiving and vastly better overload tolerance than any switchmode supply.

    3) Motor will try to keep the steps and in a lose condition the driver will compensate,
    The only reason a closed loop driver will insert extra pulses to allow the stepper to catch up is when the stepper is at or near overload condition. Any
    extra pulse inserted by the drive will suffer the same fate as the ordinary pulses.....they too get lost because the stepper is overloaded and CANNOT catch up.
    If the stepper lags too far behind the commanded position it will fault 'following error' and signal the controller of the fault and stop. An ordinary stepper
    will just stall, it has no way to signal a following error fault.

    what happen if i send a move command that crash the axis to the limit of the granity and then stop the command? Motor will try to compensate and keep crashing forever or theres a limit?
    That's the purpose of limit switches...if you crash into the gantry the switch will signal the controller to Emergency Stop, called EStop. The drive might
    try initially to compensate but will soon fault following error or be signalled from the controller to EStop.

    Craig



  4. #4
    Member sn4k3's Avatar
    Join Date
    Mar 2019
    Location
    Alvito
    Posts
    51
    Downloads
    0
    Uploads
    0

    Default Re: From NEMA 27 to NEMA 34 and voltage type selection

    Thank you for such information.

    To complete my topic:

    My spindle is only 0.8Nm max torque

    NEMA 23 23HS9442B torque curve: https://i.ebayimg.com/00/s/NDY0WDkzM..._id=8800005007

    NEMA34 86HB250-156B Motor specs: https://ae01.alicdn.com/kf/H3af294cc...283efa7e9q.jpg
    Similar motor torque curve: https://ae01.alicdn.com/kf/HTB1wG5vOpXXXXaaXFXXq6xXFXXXM/118948726/HTB1wG5vOpXXXXaaXFXXq6xXFXXXM.jpg?size=596514&heig ht=334&width=1000

    As i can see the 86HB250-156B at 1000rpms wins 23HS9442B, but not a by a far margin, i see that would not be a real improvement over what i already have...
    What steppers or servos do you recommend for a router?
    Also i see ac servos are much more power hungry, they will have dynamic current based on load or always at constant current?

    Regards



  5. #5
    Member
    Join Date
    Nov 2013
    Posts
    4280
    Downloads
    0
    Uploads
    0

    Default Re: From NEMA 27 to NEMA 34 and voltage type selection

    Hi,
    yes the 34 size motor has more torque at 1000 rpm, but not by that much. Given that it is SO MUCH bigger and yet has such a small advantage at 1000 rpm
    might come as a surprise.

    My recommendation is the lowest inductance 23/24size stepper you can find matched with the highest voltage driver you can get. I would not consider
    closed loop....if I really want closed loop I would go for servos.

    This is the Leadshine D57CM31, a 23 size stepper but with REALLY low inductance (1.18mH):
    Object reference not set to an instance of an object.
    I would match it with this driver:
    Object reference not set to an instance of an object.
    I would power it with the biggest, toughest 80VDC toroidal transformer power supply I can get.

    If you want servos I would recommend Delta B2 series(160,000 count/rev encoder). They accept 230VAC input, so no power supply required.
    https://www.ebay.com/itm/CNC-Delta-4...item1c8ecce597
    I have bought one 400W Delta servo and three Delta 750W servos from this supplier, they are great. The same supplier has a cheaper brand, it is not as well known but the price
    is very compelling:
    https://www.ebay.com/itm/400W-1-26Nm...item1c7c51f79a

    Also i see ac servos are much more power hungry, they will have dynamic current based on load or always at constant current?
    AC servos, in fact all servos, only use current when they are running, unlike steppers that require continuous current. Servos have a rated current but will have a temporary overload rating
    of three to four times normal current. A stepper has no overload, once it reaches its maximum torque it stalls...no ifs or buts....it stalls. A servo on the other hand will suck more current
    and deliver the extra torque until it overheats, so servos REALLY punch above their weight. AC servos have come down in price but they are still more expensive than open loop steppers....but
    then they are that much better, it will depend on your budget.

    Craig



  6. #6
    Member sn4k3's Avatar
    Join Date
    Mar 2019
    Location
    Alvito
    Posts
    51
    Downloads
    0
    Uploads
    0

    Default Re: From NEMA 27 to NEMA 34 and voltage type selection

    This is the Leadshine D57CM31, a 23 size stepper but with REALLY low inductance (1.18mH):
    Object reference not set to an instance of an object.
    I would match it with this driver:
    Object reference not set to an instance of an object.
    As EM882S alternative can i use: DM2280A 110-230VAC input stepper drive - CNCdrive - webshop so i can power it with 220VAC directly instead of buy Power supply?

    The stepper i can find it cheap (https://www.aliexpress.com/item/4000192126645.html) and are perfect drop in replacements, but with driver (https://kocomotion-shop.de/EM882S-V30) kills the price compared with 400W servos, almost same price.
    I found these ac servos: https://www.aliexpress.com/item/33015560148.html which look the same from ebay, but 750W cheaper than 400W!?
    EDIT: Also a cheap delta: https://www.aliexpress.com/item/32918315088.html

    I would power it with the biggest, toughest 80VDC toroidal transformer power supply I can get.
    How? toroidal transformer convert ac to ac and driver only accept DC, or you mean make own DC circuit to convert the ac to dc using the toroidal?
    Also the max input voltage is 80VDC, is not that dangerous cause of spikes? they recommend from 24 to 68 VDC

    Last edited by sn4k3; 02-22-2020 at 04:54 PM.


  7. #7
    Member
    Join Date
    Nov 2013
    Posts
    4280
    Downloads
    0
    Uploads
    0

    Default Re: From NEMA 27 to NEMA 34 and voltage type selection

    Hi,
    yes those steppers are good value and with a good high voltage driver will go much faster than what you have now. While it will go faster, I'm guessing
    about twice as fast.....'is that enough to justify the money?'

    The cheap AC servo (Yun Duan) you linked to are a perfect example of the cheap Chinese made servos. The problem is that the quality and backup is unknown, even if the price is good.
    Yes, it is entirely possible to find 750W servos at a better price than 400W....it all depends on quality and backup.

    The price for the B2 series Delta 400W is very sharp indeed. I favor Delta because while they are still cheap, at least by comparison to Yaskawa, Siemens and Fanuc, they are very
    good quality and excellent manufacturer support. I bought my 400W Delta about six months ago and now receive emails from the Ebay supplier. Because of the volume of trade
    they do with Delta they have been granted an even better discount and the supplier is glad to pass some of those savings on to me. When I bought it cost $460USD which included
    three day DHL shipping to New Zealand. It thought that price was pretty sharp. Now I could get another one for about $350USD.

    The real point is here that the price differential between good steppers and stepper drives by comparison to genuine AC servos and drives has come down dramatically.
    I personally won't be buying any more steppers, not that I haven't had excellent service from my 5-phase Vexta's but for just a little more I can get Delta servos which are in
    another performance class altogether.

    How? toroidal transformer convert ac to ac and driver only accept DC, or you mean make own DC circuit to convert the ac to dc using the toroidal?
    Also the max input voltage is 80VDC, is not that dangerous cause of spikes? they recommend from 24 to 68 VDC
    There are a number of companies that make DC supplies with a transformer, conventional or toroidal, with a rectifier and filter capacitors. I will try to find a link to them.
    Otherwise you could bet a transformer with a 50-55V AC output, add a bridge rectifier and some big filter capacitors to end up with a 72-80V DC supply.
    The filter capacitors serve not only to smooth the rectifier ripple but severely dampen any voltage spikes from the drives. If I had 80V drives I would drive them at
    80V.....I sure as hell didn't pay all the extra to get the highest voltage drives I can get only to then not utilise that capacity. If you buy Geckos, they'll handle 80V or
    Gecko will replace them. Gecko's customer support is legendary, and that's what you pay the extra for.

    Craig



  8. #8
    Member sn4k3's Avatar
    Join Date
    Mar 2019
    Location
    Alvito
    Posts
    51
    Downloads
    0
    Uploads
    0

    Default Re: From NEMA 27 to NEMA 34 and voltage type selection

    After doing some math, stepper + driver would be arround 180€ plus any tax if it stop at customs. And not including the power supply,thats maybe 100€ to 150€. It will end at same cost of a china servo... And a little more a delta. think not worth when servos are better.
    I also found this cheaper delta: https://www.aliexpress.com/item/4000218057693.html it's cheaper with more torque but it's 6.4mH but not much information on article sale, while https://www.aliexpress.com/item/32918315088.html i can't find the inductance. Who's the best?



  9. #9
    Member
    Join Date
    Nov 2013
    Posts
    4280
    Downloads
    0
    Uploads
    0

    Default Re: From NEMA 27 to NEMA 34 and voltage type selection

    Hi,

    It will end at same cost of a china servo... And a little more a delta. think not worth when servos are better.
    That is EXACTLY the same conclusion I came to....so for my new mill build I bought three B2 series 750W servos and drives, one is braked
    for the Z axis.

    Both of those links are very competitive prices, and both appear to be genuine B2 series servos and drives. Just beware that there are earlier models on sale,
    the A, B and AB series, none of which are bad but not up to the B2 series. The A2 and A3 series are later model and more expensive again.

    Inductance is a very important parameter in stepper motors but not so in AC servos. A B2 series Delta servo will blow your socks off!

    One point to consider is that each drive will need to be programmed and there are many hundreds of parameters. You can program by pushing buttons on the
    front of the drive but it is slow and confusing. Much better to use the programming and tuning software provided by Delta and use a comm cable to program
    the drive. I bought this one, being a genuine Delta accessory, but have seen them cheaper since:

    https://www.ebay.com/itm/NEW-ASDA-B2...72.m2749.l2649

    You'll need only one cable, it will program as many drives as you have got.

    Craig



  10. #10
    Member sn4k3's Avatar
    Join Date
    Mar 2019
    Location
    Alvito
    Posts
    51
    Downloads
    0
    Uploads
    0

    Default Re: From NEMA 27 to NEMA 34 and voltage type selection

    Both of those links are very competitive prices, and both appear to be genuine B2 series servos and drives.
    I think i found the difference at least on driver... The Delta logo is missing, ground screws are not green, ltp port not covered. It seens a clone of delta and not the original? I'm split, both have 5 star feedback and cheaper look better but it's a bit suspicious...

    One point to consider is that each drive will need to be programmed and there are many hundreds of parameters. You can program by pushing buttons on the
    front of the drive but it is slow and confusing. Much better to use the programming and tuning software provided by Delta and use a comm cable to program
    the drive. I bought this one, being a genuine Delta accessory, but have seen them cheaper since:

    https://www.ebay.com/itm/NEW-ASDA-B2...72.m2749.l2649

    You'll need only one cable, it will program as many drives as you have got.
    Thank you, all your information is a life saver.

    I'm studing the manual and i'm a bit confused with this diagram: https://ae01.alicdn.com/kf/Hfec7ee72...82c1ca0d63.jpg
    Do i really need to connect 220VAC to R and S? Or just L and N for single phase and leave R S open?

    Other thing i'm unsure is the best way to connect them with AXBB-E, on manual they have 4 examples, two using 24v logic and other 5v. I know 24V require isolated output and 5V non-isolated output. Which should i use for the best and safe?
    Manual: http://www.deltaww.com/filecenter/Pr...N_20141217.pdf
    Page: 9 (Wiring)



  11. #11
    Member
    Join Date
    Nov 2013
    Posts
    4280
    Downloads
    0
    Uploads
    0

    Default Re: From NEMA 27 to NEMA 34 and voltage type selection

    Hi,

    Do i really need to connect 220VAC to R and S?
    Yes, connect the line and neutral to R and S, polarity is unimportant. You need to connect line and neutral to L1c and L2c as well.
    If you had three phase power you would connect one phase to each R,S and T, being the main AC input terminals. You would still need to connect L1c
    and L2c, being the AC input for the control electronics.

    The simplest means of signaling the servo drive is with 'single ended 24V'. That will allow you to signal up to 200kHz. If you wish to signal faster then you
    will require 'differential signalling' which uses 5V differential. I doubt the ABBX-E has line driver (differential) outputs so I would go with single ended, sometimes
    called open collector, signalling.

    When you get the drive ther will be a sticker on it showing the correct way to wire an input and an incorrect way. If you do it incorrectly you will blow the input
    photodiode....so don't so it! Study and UNDRESTAND the sticker because if you blow the photodiode you wont get a warranty claim for it. BE WARNED!!!


    The difference is that you have access to the photodiode through low (50-75 Ohm) resistors, suitable for 5V max differential signalling OR by using different
    terminals through 2kOhm current limit resistors suitable for 24V signalling. If you attempt to apply 24V to the photodiode but ONLY through the low
    (50-75Ohm) resistors that's when you will damage the photodiode.

    In any event the internal electronics of the drive are optoisolated from your controller, so using isolated outputs on your controller is superfluous. The drive provides
    a low current (500mA max) isolated 24V supply be used for signalling. Just be sure you understand the signalling arrangement BEFORE you hook it up and switch
    it on.

    My own BoB's are 5V TTL output so I use a transistor between the output of the BoB and the drive.

    Craig



  12. #12
    Member
    Join Date
    Nov 2013
    Posts
    4280
    Downloads
    0
    Uploads
    0

    Default Re: From NEMA 27 to NEMA 34 and voltage type selection

    Hi,

    I think i found the difference at least on driver... The Delta logo is missing, ground screws are not green, ltp port not covered. It seens a clone of delta and not the original?
    Yes, on closer inspection it is a clone....the Chinese are famous, or should that be infamous, for passing off items made to look genuine but are not.
    Quite frankly I wouldn't consider it, I want genuine and that's what I got. I will long have forgotten the price when genuine drives are still working.

    Craig



  13. #13
    Member sn4k3's Avatar
    Join Date
    Mar 2019
    Location
    Alvito
    Posts
    51
    Downloads
    0
    Uploads
    0

    Default Re: From NEMA 27 to NEMA 34 and voltage type selection

    The simplest means of signaling the servo drive is with 'single ended 24V'. That will allow you to signal up to 200kHz.
    That way, there are two possible options, using external 24V psu (the same from AXBB-E) like show on sticker, or using delta power supply internal 24v and merge every GND. Which is the safest? As GND can leak i guess option 1 better?

    Also do i need or better have an external braker resistor? Or internal will do just fine?

    About shaft, keyway shafts are new to me, i was searching for couplings but they are hard to find on china, currently i have a Jaw type and i can find them as keyway (https://www.aliexpress.com/item/32876128828.html) but i found that disc type: https://www.aliexpress.com/item/4000199514575.html and https://www.aliexpress.com/item/32876191524.html

    Which one should i go? I guess single disc only allow angle disalignment and dual disc add parallel disalignment?



  14. #14
    Member
    Join Date
    Nov 2013
    Posts
    4280
    Downloads
    0
    Uploads
    0

    Default Re: From NEMA 27 to NEMA 34 and voltage type selection

    Hi,

    I think i found the difference at least on driver... The Delta logo is missing, ground screws are not green, ltp port not covered. It seens a clone of delta and not the original?
    Either is fine, I used the internal supply.

    If you are going from the low level acceleration of a stepper to the acceleration of a servo, that's going to be a big step up. I doubt you'll require an external braking resistor,
    UNLESS you demand the absolute fastest deceleration, and that is getting away from a hobby machine.

    About shaft, keyway shafts are new to me,
    Keyways are the way industrial machines are put together, have an engineer cut keyways into your couplers.

    Disc couplers are probably best....but it doesn't matter hugely. May I suggest get cheap ones to start with....if they are adequate, all well and good. If they are not
    then consider disc or bellows types.

    Craig



  15. #15
    Member sn4k3's Avatar
    Join Date
    Mar 2019
    Location
    Alvito
    Posts
    51
    Downloads
    0
    Uploads
    0

    Default Re: From NEMA 27 to NEMA 34 and voltage type selection

    After order some parts i just notice that these Delta servos use NEMA24 frame holes, it dismatch my nema23 holes by just 2 to 3mm. That mean i can't directly mount. Also i can't use an adapter because i cant place screws on the other side of the plate, the screw head will not fit into space.
    I have attach the axis mounts.
    What would be the best solution to overcome this and install the servos? Do a NEMA24 solid block with larger holes diameter to allow screw insert in a angle and match the NEMA27 holes? Or that would be a bad idea?

    Attached Thumbnails Attached Thumbnails From NEMA 27 to NEMA 34 and voltage type selection-img_20200307_173518-jpg   From NEMA 27 to NEMA 34 and voltage type selection-img_20200307_173555_1-jpg   From NEMA 27 to NEMA 34 and voltage type selection-img_20200307_173605_1-jpg  


  16. #16
    Member
    Join Date
    Nov 2013
    Posts
    4280
    Downloads
    0
    Uploads
    0

    Default Re: From NEMA 27 to NEMA 34 and voltage type selection

    Hi,
    I'm not sure where your 'NEMA 27' is coming from....to my knowledge there is no 'NEMA 27'. I assumed it was a typo on your part.

    NEMA is, or was, an American standard and was therefore in inches. A NEMA 23 is 2.3 inches (58mm) across the flats of the square mounting flange
    and a NEMA 34 is 3.4 inches (86mm) across the flats.

    Your new servos are made in China and metric is the norm there. It has become almost standard in metric oriented countries to make servos at
    60mm and 90mm across the flats. To make matters more confusing some of those manufacturers call their 60mm motor NEMA 24, when to my
    knowledge there is no such designation officially.

    The difference between NEMA 23 and 60mm is only about 2mm. If you can fit a nut to the coupler side (your third pic) then you should be able to elongate
    the holes in the mount with a small round file.

    The mount in the first pic is more difficult. It that situation you have to fit the servo and rotate it a few degrees so you can drill and tap fresh holes at the slightly
    wider spacing.

    Craig

    Attached Thumbnails Attached Thumbnails From NEMA 27 to NEMA 34 and voltage type selection-couplersidenut-jpg  


  17. #17
    Member sn4k3's Avatar
    Join Date
    Mar 2019
    Location
    Alvito
    Posts
    51
    Downloads
    0
    Uploads
    0

    Default Re: From NEMA 27 to NEMA 34 and voltage type selection

    All my steppers and mounts are NEMA23, sorry for confusion, my bad, sometimes i type 27 instead of 23 don't know why... So where i said 27 should be 23.

    The difference between NEMA 23 and 60mm is only about 2mm. If you can fit a nut to the coupler side (your third pic) then you should be able to elongate
    the holes in the mount with a small round file.
    Yes that the easiest axis (Y) to modify, still i was about to replace that thin steel bridge with an strong adapter that goes to the thick steel of structure, will see.


    The mount in the first pic is more difficult. It that situation you have to fit the servo and rotate it a few degrees so you can drill and tap fresh holes at the slightly
    wider spacing.
    Dam, sometimes simpler solutions are harder to think of, thank you for enlighten.
    I already have an 3d printer model of servo, so i can start testing.

    I also start studing the parameters, it seens we can choose the amount of torque? Some installations have a pot to regulate the torque on fly, if i understand correctly it will keep that constant torque all the time or will be dynamic up that value giving the motor load?

    About steps/rev, are choosen with P1-44 and P1-45 Electronic Gear Ratio (1st Numerator) (N1)
    If i set P1-44 1600 it will match stepper drivers 1600 steps/rev? With 320 steps per unit on software? It seens the default are 16, which is the best value to use with this servos?

    I found a quick setup guide here: https://www.damencnc.com/userdata/fi...tart_Guide.pdf



  18. #18
    Member
    Join Date
    Nov 2013
    Posts
    4280
    Downloads
    0
    Uploads
    0

    Default Re: From NEMA 27 to NEMA 34 and voltage type selection

    Hi,

    I also start studing the parameters, it seens we can choose the amount of torque?
    Yes, you can program a maximum torque, and it is appropriate in certain circumstances but probably not for your machine.
    What you are talking about here is servo motor tuning and setup. These servos are very flexible and have been designed so they
    can be used in a huge and bewildering array of situations which maximizes the market for them. Many of the operating modes
    are of no interest to you.

    For example there is a 'constant torque' mode. A motor in this mode rotates at whatever speed it needs to produce its commanded torque against
    a load. A printing press might be an example. The roller to which this motor is attached generates a constant torque and therefore holds the paper under constant tension.
    If the torque varies then so would the tension in the paper, including the situation that the paper droops between rollers or worse if the tension increases too much the paper tears.

    As I say, this is very clever, and no doubt would be exploited if you are designing and building a printing machine or maybe conveyor systems but you want
    position control for a CNC axis.Your CNC motion controller will command a position by issuing step and direction pulses and the servo and drive is
    expected to execute that command. This is called 'tool path following'. When following a tool path the higher the acceleration the servo can attain the better it
    can follow sudden changes in direction and speed that it may be commanded to execute. Maximum acceleration means maximum torque.
    Torque is directly proportional to current, and you will often see a basic specification of a servo 'so many Nm's per Amp', its a basic
    figure of merit of a servo. Essentially the more torque it can produce at the lowest possible current the better. For best tool path following you want
    maximum acceleration, therefore maximum torque, therefore maximum current WITHOUT the servo overheating.

    With Delta servos the drives are supplied by default to produce rated current, that is the maximum continuous current the servo can accept without overheating
    but also a 10 second overload of about 300% of rated current. This is fine for your application. You can of course change it but you shouldn't have to, you
    might find it advantageous to do so later on but not initially.

    Usually Delta servo drives are by default supplied in position mode. All you really have to do is set the 'electronic gearing' and maybe the pulse mode
    (step/dir OR CCW/CW for example) and you should be good to go.

    Electronic gearing gives people a lot of headaches until they sort it out in their own mind.

    The first thing to realise and fix firmly in your brain is that the NATIVE encoder in the servo has 40,000 lines (ppr) which with quadrature
    results in a 160,000 count (cpr) encoder. You could theoretically cause the servo to step 1/160,000 th of a revolution or
    0.00225 degrees or 8.1 arc seconds clockwise by issuing one step pulse to the drive. You'd have to agree that is exceptionally fine resolution,
    way WAY WAY more than we would ever need....but its there.

    The second thing to understand is that the servo drive has three 'encoder outputs, A, B, I'. These encoder outputs are NOT the
    native encoder but are DERIVED from it. For instance you may have a computer monitoring the servo, say with our printing press example, to keep
    track of the number of turns and therefore the paper used but also the speed which might be used to determine how much ink should be pumped
    into the print head say.

    This auxillary encoder can be programmed (P!-46) to have anywhere from 4 lines (16 cpr) to 40,000 lines (160,000 cpr) with a default of 2500 lines (10,000 cpr).
    The auxillary encoder does not affect the commanded step rate (or electronic gearing) IT IS SOLEY for you, the machine builder, to monitor the position of the servo.

    Lets imagine you had to pulse the servodrive 160,000 times to get the servo to turn one revolution and then you want to spin at 50 revolutions per second (3000 rpm)
    the pulse rate is:
    PulseRate=50 x 160,000
    =8,000,000 pulses per second or 8Mhz!! That's HF radio frequency!!

    Using the NATIVE encoder count of 160,000 cpr results in a required signalling rate from your motion controller to the servo drive which is impractically
    high. To overcome this shortcoming you use electronic gearing. This involves setting two numbers, a numerator and denominator.
    For example lets say you controller is issuing pulses at a rate of 12,500 pluses per second. If you set the numerator at 21534 and the denominator at 113:
    10000 x 21534/113 = 1,905,663 pulses per second or 11.91 revolutions per second with a NATIVE encoder count of 160,000 (1905663/160000=11.91).

    Whats happened here is that the 'electronic gearing' has stepped up the fairly slow pulse rate from the controller to a much higher rate to spin the servo faster
    than you could otherwise. Note I have chose these numbers pretty much at random...they have no significance otherwise.

    Lets say you need to replace a servo or stepper motor in a machine. You could bolt in a Delta servo and use the electronic gearing to make it mimic the motor its
    replacing WITHOUT having to alter the controller AT ALL.

    Another example of the use of electronic gearing is if you were using Mach's parallel port in default configuration of 25kHz max pulse output rate. If you want
    to be able to spin the servo to its rated maximum of 3000rpm then you need a step rate at NATIVE encoder resolution of 8Mhz
    25Khz= N/M X8000 kHz

    Solving for M and N:
    N=320 M=1
    Thus if you programmed the drive with N=320 (P1-44) and M=1 (P1-45) and applied Mach's max frequency to the drive the servo would spin at its full 3000rpm.
    Note also that there is more than one combination of N and M that would work. For instance if N=640 and M=2 or N=5440 and M=17 both work to.

    As you can see electronic gearing adds a lot of flexibility in application. Note that the numerator and denominator are programmed into the drive and are
    (normally) left that way. You will notice that Delta have given you the opportunity to program four numerators N1, the primary one is P1-44 but three others,
    N2 though N4 at P2-60 through P2-62. You can select one of these numerators to be active by using two digital inputs (GNUM0 and GNUM1). Thus you could
    be cruising along at 20000 pulse per rev but then you decide you need to change what you are doing so can select another 'gear' and go to 5000 pulse pre rev.
    This can be changed on the fly.Clever or what???

    You can spend many hours programming different parameters, especially if you use Delta's tuning and setup software, to make the servo do as you wish,its
    all rather fascinating. A little time doing so will make you appreciate just how far FAR FAR advanced modern AC servos are by comparison to brushed DC servos
    and steppers.

    Craig



  19. #19
    Member sn4k3's Avatar
    Join Date
    Mar 2019
    Location
    Alvito
    Posts
    51
    Downloads
    0
    Uploads
    0

    Default Re: From NEMA 27 to NEMA 34 and voltage type selection

    Only a man who loves this can write such explanation

    There are a few things i can't sort out and get confused, such as:

    For example lets say you controller is issuing pulses at a rate of 12,500 pluses per second. If you set the numerator at 21534 and the denominator at 113:
    10000 x 21534/113 = 1,905,663 pulses per second or 11.91 revolutions per second with a NATIVE encoder count of 160,000 (1905663/160000=11.91).
    Where's the 10000 value comes from? Shouldn't be 12500?


    25Khz= N/M X8000 kHz

    Solving for M and N:
    N=320 M=1
    320*8000 = 2560000 but != 25Khz ?
    Software wise, what value would go into motor steps per unit? 320?

    Under UCCNC i can choose kernel frequency of 25Khz, 50Khz, 100Khz, 200Khz and 400Khz.
    Giving your example, and if i'm on 100khz then to match at full 3000rpm would be 320/4 = 80 for N at M=1?
    Will i have any benefits running higher kernel frequency when the electronic gear can solve the problem at any selected frequency?
    The numerator and denominator can be used "any" value as long they match the desired result, set the denominator other than 1 would have any benefit?



  20. #20
    Member
    Join Date
    Nov 2013
    Posts
    4280
    Downloads
    0
    Uploads
    0

    Default Re: From NEMA 27 to NEMA 34 and voltage type selection

    Hi,
    Where's the 10000 value comes from? Shouldn't be 12500
    Yes you are correct, I wanted to select numbers that were not a 'slam dunk', and then managed to confuse myself.

    The corrected calculation is:
    12500 x 21534/113=2,382,079 or 14.88 revs/sec or 893.3rpm

    320*8000 = 2560000 but != 25Khz
    Again you are correct. Writing the equation correctly per page 8-70 of the manual:
    8000kHz= 25kHz X 320/1

    Software wise, what value would go into motor steps per unit? 320?
    That depends on your machine. Lets say you have a servo direct coupled to a 5mm pitch ballscrew. One revolution of the servo results
    in the axis advancing 5mm. If we retain N=320 and M=1 per my example:

    160000/320=500, ie 500 pulses at the step input of the drive will result in the servo rotating 160,000 of its native encoder, ie one revolution.
    Note how this works, if you issue one pulse to the input of the drive the electronic gearing means that it issues 320 steps to the native
    servo drive.

    The steps per unit vale is:
    StepsPerUnit=500 (steps/rev) / 5 (mm per rev)
    =100

    Under UCCNC i can choose kernel frequency of 25Khz, 50Khz, 100Khz, 200Khz and 400Khz
    In my example I used Mach3's default kernel speed of 25kHz. If you want to be able to drive your servo to full speed with such a low
    input frequency you have to use a 'high (electronic) gear ratio' with a loss of resolution. Lets do the same calculation but using the highest
    kernel frequency you can in UCCNC, 400kHz.

    8000kHz=N/M x 400Khz
    therefore N=20 and M=1

    Lets now compare the resolution of the two setups, assuming the same direct coupled 5mm pitch ballscrew:
    With 25kHz kernel (N=320,M=1) one step is 5 x 320/160000 =0.01mm=10um

    With 400kHz kernel (N=20 M=1) one step is 5 x 20/160000=0.000625mm=0.625um

    So you see that the higher the kernel speed and therefore the max pulse frequency allows greater resolution AND STILL be able to achieve full speed of the servo.
    Thus higher kernel speeds are favoured. However it may be that the PC and motion controller benefit from going a little slower

    I use an Ethernet SmoothStepper and it is capable of 4Mhz. The max theoretical resolution while still being able to get 3000 rpm is:
    8000kHz=N/M x 4000kHz, ie N=2, M=1
    so with 4MHz kernel one step is 5 x 2/160000=0.0000625mm=0.0625um.
    I say this is theoretical....its way more than I need or is practical. The maximum practical signaling rate to the servo is 500kHz using differential signaling
    or 200kHz using single ended 24V signaling. It doesn't really matter, at least as far as the servo goes, that the ESS is way faster than that. The kernel speed
    is fixed at 4MHz, there is no adjustment offered or required.

    The numerator and denominator can be used "any" value as long they match the desired result, set the denominator other than 1 would have any benefit?
    Not to my knowledge. Having said that I'm getting to the limit of what I understand, there may be advantage in selecting one of the possible combinations
    of N and M but unknown to me.

    Craig



Page 1 of 2 12 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

From NEMA 27 to NEMA 34 and voltage type selection

From NEMA 27 to NEMA 34 and voltage type selection