Speed Control For SX3 ? - Page 5


Page 5 of 7 FirstFirst ... 234567 LastLast
Results 81 to 100 of 136

Thread: Speed Control For SX3 ?

  1. #81
    Community Moderator Al_The_Man's Avatar
    Join Date
    Dec 2003
    Location
    Canada
    Posts
    24216
    Downloads
    0
    Uploads
    0

    Default

    Quote Originally Posted by devers6 View Post
    I looked at the full AMC product line and saw the B30A40 was the next most appropriate drive.
    The thing with the B30A40 is that with a separate tailored supply, you can run the motor at max voltage/rpm, the AC version will give you the same result as the B20A25AC
    I will hook up the factory drive and read the Hall effect to see if I get the same rpm with the 'scope.
    The one mechanical cycle of the motor should display 4 electrical cycles for any hall output, so any reading of a Hall will be divided by 4 for RPM calcs.
    BTW, I took the end cover off to see what method of commutation is used, because modern BLDC use commutation off the encoder disk, not actual hall effect devices anymore.
    But unfortunately it appears that actual hall effect is used and so no unused encoder!!.
    On the original set up I am assuming they are reading one of the hall effect devices for display purposes, if so the rpm display should be spot on.
    Incidentally there was a B20A25AC listed last week for $45.00, it did not go!
    Also, one of the deciding factor for top rpm is the DC supply, so two controllers using the same DC bus voltage, should produce the same RPM.
    Al.

    CNC, Mechatronics Integration and Custom Machine Design

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


  2. #82
    Registered
    Join Date
    May 2007
    Location
    US
    Posts
    715
    Downloads
    0
    Uploads
    0

    Default

    Holding my breath,

    Good work on this guys. But the off thing is that I emailed AMC before about this with the specs of the motor as I had read them off of the motor and they said there products wouldn't work. I sure hope they were wrong. I will try to find that email.



  3. #83
    Community Moderator Al_The_Man's Avatar
    Join Date
    Dec 2003
    Location
    Canada
    Posts
    24216
    Downloads
    0
    Uploads
    0

    Default

    Quote Originally Posted by cjdavis618 View Post
    I emailed AMC before about this with the specs of the motor as I had read them off of the motor and they said there products wouldn't work. I sure hope they were wrong. I will try to find that email.
    That's strange, it is a standard BLDC motor?
    I would be interested to hear their reasoning also.
    Al.

    CNC, Mechatronics Integration and Custom Machine Design

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


  4. #84
    Registered
    Join Date
    May 2007
    Location
    US
    Posts
    715
    Downloads
    0
    Uploads
    0

    Default

    I have searched but I don't see it, come to think of it, it may have been a voicemail.

    If I remember correctly, it was due to the voltage of the motor and they didn't have one that matched it enough to get the full speed. For some reason 300V ac came to mind regarding the motor, but I can't exactly remember what that related to in thier answer.

    I didn't get a response from them for nearly 2 weeks though. I do remember that, and that is one reason I went and got the set from Keling.



  5. #85
    Community Moderator Al_The_Man's Avatar
    Join Date
    Dec 2003
    Location
    Canada
    Posts
    24216
    Downloads
    0
    Uploads
    0

    Default

    Quote Originally Posted by cjdavis618 View Post
    If I remember correctly, it was due to the voltage of the motor and they didn't have one that matched it enough to get the full speed. For some reason 300V ac came to mind regarding the motor, but I can't exactly remember what that related to in thier answer.
    Mmm.. well if using one of the non AC suffix types previously mentioned the power supply can be tailored to suit, as long as the drive falls somewhere in the range?
    They make drives that can cater for 24vdc to 400vdc supply?
    Al.

    CNC, Mechatronics Integration and Custom Machine Design

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


  6. #86
    Registered
    Join Date
    Mar 2004
    Location
    USA
    Posts
    92
    Downloads
    0
    Uploads
    0

    Default

    I apologize for not updating over the last couple of days - another home project took priority, and I was not able to get back to the shop to work on the mill. I admit I hadn't thought about the ramifications of leaving a post saying "I'm about to go out and stick my hands in a high-power AC circuit", and then not make any subsequent reply. So I thought I would let you know that I at least wasn't still standing rigidly at the mill, vibrating and smoking out the eyeballs.

    I know a lot of you are interested in this thread, and I should have some time tonight to do a bit of comparison in RPM between the old and new drivers. I think I know how it is likely to go - the Sieg/Grizzly driver will probably show about double the voltage at full speed as the AMC; and that will probably be the best we can expect from this drive model. I see a B30A40 drive is listed on ebay, but it is over $300 which is above my threshold to just buy and try out. If I can find one for a reasonable price, I might likely buy one and supply it with simple rectified/filtered 220V, and it should be roughly equivalent to the existing drive.

    Other possibilities are to go to 1:1 pulleys on the spindle (for 1/2 the torque), live with half the RPM, use a different motor, etc. All relatively non-optimal solutions for something that should have been included in the spindle drive in the first place (0-10v control). I would have rather had a pot on the front panel than the up/down keys anyway. It really surprises me that this has gone for so long without an adequate solution. It looks like there are now at least a few more options available than we had a couple of weeks ago, but it is a shame it has taken this much effort to re-engineer the design of a spindle motor on an otherwise excellent platform.



  7. #87
    Registered
    Join Date
    Mar 2004
    Location
    USA
    Posts
    92
    Downloads
    0
    Uploads
    0

    Default Yo! Light Bulb moment!

    So I have learned much over the past few days about BLDC motor control, what is available in existing hardware, and some of the concepts that are common across all the available controllers. With that new knowledge, I went back to some of the earliest research I did on this subject; and I now have another potential solution.

    Early on, I made contact with a user here named "oldraven", who had fixed a dead SX3 controller board; and in the process made schematics of the power board and the controller board. He was kind enough to share them with me, but I had little knowledge of the subject at the time and they meant nothing to me then. However, this week I dug them back out, and something jumped out at me - the SX3 controller uses a standard MC33035 BLDC motor controller chip. This chip is likely used in many of the off-the-shelf servo and BLDC controllers because it handles commutation and other motor control issues; but what struck me was that it also includes speed control by DC voltage.

    Pin #11 of the MC33035 is typically attached to the wiper of a pot, which when varied between 0 volts and the reference voltage supplied on pin #8, controls the speed of the motor. Since the SX3 includes this chip, voltage control should be just a simple matter of injecting the control voltage into pin #11 of the chip.

    I still know that I don't know much about this subject, so before I go tempting fate by hacking into this PC board, I thought I would run this by any of you motor control experts out there to see if this seems to be a reasonable idea.

    I have attached a jpg of the relevant portion of the schematic that oldraven had produced. It shows pin 11 basically going to 6V. It would seem to me that just disconnecting this trace, and instead feeding a 0-6 volt signal either through a pot or an external source should provide speed control. It should be relatively easy to test this, but before I tried it I thought I would see if anybody has any concerns or thoughts about this approach.

    I have attached a couple of PDF's of the schematics, and the datasheet for the MC33035 IC is here:

    http://www.onsemi.com/pub_link/Collateral/MC33035-D.PDF

    Any advice as to how an external signal should best be brought in to the circuit board? Any buffering required? Scaling to 10V and buffering could probably easily be done using a simple op-amp if necessary.

    Dave

    Attached Thumbnails Attached Thumbnails Speed Control For SX3 ?-sx3-controlboard-pdf   Speed Control For SX3 ?-sx3-powerboard-sheet2-a2-pdf  


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

    Default

    I will dig up my old notes this week end on old experiments with the MC33035 and see what they turn up.
    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
    Oct 2008
    Location
    United States
    Posts
    1632
    Downloads
    0
    Uploads
    0

    Default

    The Original SX3 board used a motor control chip. All the SX3's in the last couple of years have controlled the BLDC motor directly from an Atmel MCU with the appropriate drivers. Please check your board closely to see if the schematic and board matches.

    I have also spent a lot of time reverse engineering this board. I even planned on writing new firmware for an Atmel to drop in place, however time had been very limited and I just decided to add a motor and VFD. I may get back to it but for now there is lots of stuff to do outside on other things.

    Richard

    Quote Originally Posted by devers6 View Post
    So I have learned much over the past few days about BLDC motor control, what is available in existing hardware, and some of the concepts that are common across all the available controllers. With that new knowledge, I went back to some of the earliest research I did on this subject; and I now have another potential solution.

    Early on, I made contact with a user here named "oldraven", who had fixed a dead SX3 controller board; and in the process made schematics of the power board and the controller board. He was kind enough to share them with me, but I had little knowledge of the subject at the time and they meant nothing to me then. However, this week I dug them back out, and something jumped out at me - the SX3 controller uses a standard MC33035 BLDC motor controller chip. This chip is likely used in many of the off-the-shelf servo and BLDC controllers because it handles commutation and other motor control issues; but what struck me was that it also includes speed control by DC voltage.

    Pin #11 of the MC33035 is typically attached to the wiper of a pot, which when varied between 0 volts and the reference voltage supplied on pin #8, controls the speed of the motor. Since the SX3 includes this chip, voltage control should be just a simple matter of injecting the control voltage into pin #11 of the chip.

    I still know that I don't know much about this subject, so before I go tempting fate by hacking into this PC board, I thought I would run this by any of you motor control experts out there to see if this seems to be a reasonable idea.

    I have attached a jpg of the relevant portion of the schematic that oldraven had produced. It shows pin 11 basically going to 6V. It would seem to me that just disconnecting this trace, and instead feeding a 0-6 volt signal either through a pot or an external source should provide speed control. It should be relatively easy to test this, but before I tried it I thought I would see if anybody has any concerns or thoughts about this approach.

    I have attached a couple of PDF's of the schematics, and the datasheet for the MC33035 IC is here:

    http://www.onsemi.com/pub_link/Collateral/MC33035-D.PDF

    Any advice as to how an external signal should best be brought in to the circuit board? Any buffering required? Scaling to 10V and buffering could probably easily be done using a simple op-amp if necessary.

    Dave




  10. #90
    Registered
    Join Date
    Sep 2006
    Location
    USA
    Posts
    58
    Downloads
    0
    Uploads
    0

    Default

    devers6,

    I am not a motor control expert, however my expertise is in electronic circuit design. My interest is also the SX3 since I own a CNC one.

    Attached is a file on how I would hook the pot to control the speed. If you can afford de-soldering the two circled components, then you can add the 10K pot (this pot value is suggested on the ON spec sheet) as per drawing.

    You can use one side of the diode (cathode) to connect the top of the pot , the other side of the diode (anode) to the wiper (center) of the pot and of course the 3rd pot connection to GND. Make sure to de-solder or disable the PNP transistor.

    It should work there isn't anything magic on the circuit, no buffering, etc. should be required, and the rest to circuit blocks have not been compromised, only the control to pin 11 has been added. It is using the 6V regulated voltage that is provided.

    Hope it works for you. Let us know how it comes out.

    groov

    PS.Included is a zip file which you will have to download in order to see the posted picture, otherwise the resolution would not have been great.

    Attached Files Attached Files


  11. #91
    Community Moderator Al_The_Man's Avatar
    Join Date
    Dec 2003
    Location
    Canada
    Posts
    24216
    Downloads
    0
    Uploads
    0

    Default

    Myself, I would disconnect the pin 11 and feed this from the pot slider, the outer ends of the pot would go to common and pin 8.
    If using Mach the analogue would go between common and pin 11.
    Al.

    CNC, Mechatronics Integration and Custom Machine Design

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


  12. #92
    Registered
    Join Date
    Mar 2004
    Location
    USA
    Posts
    92
    Downloads
    0
    Uploads
    0

    Default Not so fast

    Well, all of this is moot, because as rwskinner reminded me, the current version of the motor control board did away with the MC33035 and replaced it with an Atmel ATMEGA8-16 mounted on a daughterboard with a bunch of optocouplers. Apparently the schematic oldraven sent was of the older variety, but I double checked my board and it indeed is the newer version.

    But all is not necessarily lost. Because the daughterboard is a plug-in, it is possible to make another that would substitute for it. I have begun ringing out the connections to the board (there is a 10 pin header on one side, and a 20 pin on the other, both single inline), and I have already verified that the hall effect sensors go directly to pins in these headers. So it looks like all the circuitry for commutation happens on the daughterboard.

    I can see one of two approaches:

    1) Replace the ATMEGA8-16 with another MCU, this one using an analog input to allow for speed control from a 0-10v source. The output would go to the main board where the power circuitry resides (a triple H-bridge using power mosfets or igbts). I found a very suitable application note here:

    http://www.atmel.com/dyn/resources/p...ts/doc4987.pdf

    It describes an Atmel chip in a BLDC control solution with external speed control. It is a little overkill for this application because it includes LIN communication for automotive functions, as well as a power stage (ATA6832) which would directly drive a small motor; but perhaps this could handle the isolation/output into the high-power section on the main PCB.

    2) Build a similar daughterboard using the MC33035 to feed the main board power stage. Obviously this is a step backward as far as technology goes (Sieg must have had a reason to move to the ATMEGA - might have been cost, part availability, problems, or the desire to have an updatable control stage); but it would seem that the MC33035 could handle the requirements of a variable-speed spindle motor with external speed control. I myself would rather have a pot on the front panel than the up and down speed buttons (or any of the membrane switches for that matter); so as long as the circuit could handle the job I don't think it going backwards is a bad thing. I think this is my preferred approach, because it does not have the added complication of firmware. I have a chip coming, and I will throw together a circuit that matches as much as possible the original SX3 driver with the addition of external control.

    I am going to spend some time this weekend trying to reverse engineer the daughterboard and the pathway to the power section. It seems possible to create a substitute daughterboard to handle commutation logic and motor phase signal generation, which would then drive the existing power stage. I could use a little help from you electronics types out there who wouldn't mind looking at your system, and help trace out the function of the interface pins on the daughterboard. I don't think we have to document the entire board (I think the majority of the rest of the circuitry on the main board is voltage-doubling and rectifying (the board takes a 110VAC input and ups it to about 300VDC for the motor); we just need to see where the 3 phases go when they leave the daughterboard on their way to the power devices.

    I really hope this pans out. There have been way too many false starts on this project; and I'm determined to see it through one way or another. I plan on prototyping a schematic, trying it out, then perhaps enlisting some help in laying out a PC board that could be made with the many PCB houses that are begging for business these days. If everything turns out OK, I would buy a number of boards for others who might want to build them (I would try to keep it through-hole and low parts count for the soldering-challenged).



  13. #93
    Registered
    Join Date
    Mar 2004
    Location
    USA
    Posts
    92
    Downloads
    0
    Uploads
    0

    Default A bit more information

    The controller daughterboard includes (1) ATMEGA8-16 24 pin DIP, (1) 74LS09 14 pin DIP, (3) EL817 4 pin DIP (opto), (6) 6N137 8 pin Dip. I believe the EL817 isolate the hall effect sensor inputs, and the 6N137 isolate the high and low drive to the IGBT's on the power board.

    The power board has (6) G60N100 60A 1000V IGBT's in a triple H-Bridge for the 3 motor phases. It also shows (3) IR2103 Half Bridge Drivers (http://www.irf.com/product-info/data...ata/ir2103.pdf) which look to take the top drive from the controller. I will verify the circuit path around these tomorrow.

    Attached Thumbnails Attached Thumbnails Speed Control For SX3 ?-controller-board-circuit-side-jpg   Speed Control For SX3 ?-controller-board-solder-side-jpg   Speed Control For SX3 ?-g60n100-igbt-jpg   Speed Control For SX3 ?-power-board-circuit-side-jpg  

    Speed Control For SX3 ?-power-board-solder-side-2-jpg  


  14. #94
    Community Moderator Al_The_Man's Avatar
    Join Date
    Dec 2003
    Location
    Canada
    Posts
    24216
    Downloads
    0
    Uploads
    0

    Default

    Quote Originally Posted by devers6 View Post


    It describes an Atmel chip in a BLDC control solution with external speed control. It is a little overkill for this application because it includes LIN communication for automotive functions, as well as a power stage (ATA6832) which would directly drive a small motor; but perhaps this could handle the isolation/output into the high-power section on the main PCB.
    I think the controller I got hold of is the same model, I haven't looked closely, but it could be the display is also being done by the Atmel chip if it has this feature, I use PicMicro so I am not that familiar with Atmel.
    It would make sense for all the control and display to be done by an embedded processor?
    If I get enough time, I may dig into it a bit deeper this week end.
    Good luck.
    Al.

    CNC, Mechatronics Integration and Custom Machine Design

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


  15. #95
    Registered
    Join Date
    Mar 2004
    Location
    USA
    Posts
    92
    Downloads
    0
    Uploads
    0

    Default Some excellent help

    I spent some time re-reading this thread:

    http://www.cnczone.com/forums/bencht...rol_sx3-3.html

    I had high hopes for this approach when it was ongoing, but it seems to have died on the vine about a year ago. I am also not so sure I like the idea of achieving speed control by injecting up/down button presses into the communication stream between the front panel and the controller board - seems like the long way around to get the desired action, and has got to be slow.

    Buried in this thread (post #30) was an offer of assistance by a group from Belgium who had done some investigation and reverse engineering of the SX3. They said they had attached a document with the results of their schematic tracing, although the post didn't include any attachments. So I reached out to them via their website, and one of them (Patrick Bayard) kindly responded by email with the missing attachments.

    These are excellent resources, so I thought I would share them here. It confirms the limited tracing I did, so I have high confidence in the accuracy of this work. I want to thank Patrick and the others at A2P for making this information available.

    Attached Thumbnails Attached Thumbnails Speed Control For SX3 ?-att00024-pdf   Speed Control For SX3 ?-spindel-sx3-1-pdf  
    Attached Files Attached Files


  16. #96
    Registered
    Join Date
    Mar 2004
    Location
    USA
    Posts
    92
    Downloads
    0
    Uploads
    0

    Default

    So after perusing the A2P schematics, it appears that the 20 pin header contains all the communication with the main board and the hall effect sensors. The 10 pin header looks to communicate with the front panel. I've never been much of a fan of the membrane switches on the front panel, so in my replacement I think I will just provide inputs for a maintained switch (push on/push off) for On/Off, a toggle switch for direction, and a 0-10v speed voltage along with another switch to select between manual and CNC operation. I will implement this on a separate connector, using the 10 pin header just for physical mounting. Theoretically, this would allow for the old daughterboard to be unplugged, the new daughterboard installed in its place, and the new manual and 0-10v controls connected on the additional connector.

    As I mentioned before, I am leaning toward using the MC33035 on the daughterboard, interfaced using the same optoisolators as the existing circuit (EL817 on the halls, 6N137 on the outputs). Although this is my preferred approach, it doesn't rule out using an MCU on a different daughterboard should the MC33035 not work out. I will try to post mechanical and electrical details as I figure them out, in case someone else wants to pursue the ATMEGA approach (probably could use any suitable MCU like PIC, etc). There are lots of BLDC application notes available, and they all seem very similar and easily adaptable to the power section on the main board.

    I haven't done much design work in a while, so I would appreciate any feedback on the design as it evolves. At this point, I will probably generate the schematic using PCB123 (free download, available here: PCB123 - Free PCB Design Software | Sunstone.com). The benefit is you can do the schematic, route the board, and then send it off to the PCB house for prototypes all in one program. There are others, and many of you probably have your favorites. I'm open to using other tools and vendors if you have someone you particularly like and that has advantages such as lower cost, faster turnaround, etc.

    As a starting point, here is the pinout of the 20 pin header (CN302):

    1 Hall Sensor A
    2 Hall Sensor B
    3 Hall Sensor C
    4 +5V
    5 GND
    6 GNDF
    7 U303 Hin (W Phase)
    8 U303 /Lin (W Phase)
    9 U302 Hin (V Phase)
    10 U302 /Lin (V Phase)
    11 U301 Hin (U Phase)
    12 U301 /Lin (U Phase)
    13 +5VF
    14 +5VA Power Good?
    15 GNDA
    16 GNDA
    17 CTRL1
    18 CTRL2
    19 +5VA
    20 +5VA

    The CTRL1 and CTRL2 seem to be signals derived from pots on the board for adjustments (current sense?). Could someone knowledgeable look at the circuitry around these and see what these look to be doing, and how the circuit should accomodate them? They are documented on Sheet 3 of the A2P schematic. I am going to ignore them for now, with the option to implement them later if necessary.



  17. #97
    Registered
    Join Date
    Mar 2004
    Location
    USA
    Posts
    92
    Downloads
    0
    Uploads
    0

    Default Mechanical for SX3 Daughterboard

    Here are the results of some quick measurements for the SX3 daughterboard. I would appreciate it if someone can verify these measurements on their machine.

    Attached Thumbnails Attached Thumbnails Speed Control For SX3 ?-sx3-daughterboard-jpg  
    Attached Files Attached Files


  18. #98
    Community Moderator Al_The_Man's Avatar
    Join Date
    Dec 2003
    Location
    Canada
    Posts
    24216
    Downloads
    0
    Uploads
    0

    Default

    Pretty close, I made it .497 and .1335 for the header position between the two.
    With the board removed.
    Al.

    CNC, Mechatronics Integration and Custom Machine Design

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


  19. #99
    Registered
    Join Date
    Mar 2004
    Location
    USA
    Posts
    92
    Downloads
    0
    Uploads
    0

    Default

    Thanks, Al

    I was .1 off on the offset between the large header and small header - it looked to me like it was lined up between pins, so I marked it as .55 but it was actually .45; thanks for the catch.

    I figured the spacing between pins would round to .1 so I left it 1.400.

    Attached Thumbnails Attached Thumbnails Speed Control For SX3 ?-sx3-daughterboard-jpg  
    Attached Files Attached Files


  20. #100
    Community Moderator Al_The_Man's Avatar
    Join Date
    Dec 2003
    Location
    Canada
    Posts
    24216
    Downloads
    0
    Uploads
    0

    Default

    Most likely the original was in metric?
    The distance between the two headers is not super critical due to the length of the pin stand-off.
    Al.

    CNC, Mechatronics Integration and Custom Machine Design

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


Page 5 of 7 FirstFirst ... 234567 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

Speed Control For SX3 ?

Speed Control For SX3 ?