584,842 active members*
4,184 visitors online*
Register for free
Login
Results 1 to 19 of 19
  1. #1
    Join Date
    Sep 2004
    Posts
    107

    I think I have an encoder wiring issue

    I've got CNC10 LInux with a DC3IO. The XYZ axes are fine.

    I'm in the process of hooking up a 4th axis with the DC Single add-on drive after having the 4th axis stuff "resting" for a long time. I have a new F14 Dynapar encoder on the servo motor.

    I did the wiring in the cabinet today after finishing the cable and started checking things with the motor clamped to the mill table (not to the rotab yet). I'm getting a 410 axis position error.

    I've changed the motor spin direction in the config area, and that makes no difference. The motor always turns in the same direction no matter which direction I turn the MPG dial or if the direction is reversed or not in the config page. The count in PID keeps decreasing, always showing a negative number that gets ever smaller (more digits in the number, so think of it as a negative increase).

    As best I can tell I've got the encoder cable/plugs/DB9 wired correctly, but I suspect I'm not getting a signal from the A or B channel or maybe have something miswired (probably the latter). I get counts, it just can't seem to tell anything about what direction it is going in.

    So since I am getting counts, does that mean that either the A or the B channel is fine and I messed the wiring on the other pair? My limited understanding is that direction is told from the A leading or following the B.

    How do I troubleshoot this? Should I take an DCV reading between the A and A/ and then the B and B/ wires to make sure I got them in the right position. Is the Z and Z/ the index signal, which sounds like it should be showing up once per rev as a voltage change?

    I am not an electronics person and I don't have an O-scope. But I've done 2 Centroid mill conversions and they've both worked so I suspect I've got an "oops" buried in the DB9 cable to the PC/motion control board. I'm hoping at worst I can change locations and resolder on a couple of wires in the AMP connector from the conduit to the rotab, and not need to rip everything apart to pull the encoder cable out of the conduit.

    The home/clamp ack signals are fine and show up in the ALT-I screen, the motor spins, so it seems like it would have to be an encoder issue. But that's just my non-electrician guess.

    thanks,
    Michael

  2. #2
    Join Date
    Sep 2004
    Posts
    107

    Re: I think I have an encoder wiring issue

    I'll add that while the +/- jog buttons/MPG always increment in the same direction, manually turning the motor shaft does have the DRO reading go up and down.

  3. #3
    Join Date
    Dec 2008
    Posts
    44

    Re: I think I have an encoder wiring issue

    If your encoder counts smoothly both positive and negative when you rotate the shaft with your hand., you encoder is probably ok. With the shaft pointing at while your looking at the motor faceplate, the motor should count more positive when you rotate the shaft CCW. If your counts opposite, you can swap the polarity of you motor power wires for that axis.

    Typically a motor that always takes off in 1 direction indicates a problem with the drive. You can isolate the problem by connecting your motor, cables and encoder to a known good encoder connection and motor power connection for another axis

  4. #4
    Join Date
    Sep 2004
    Posts
    107

    Re: I think I have an encoder wiring issue

    Thanks for the suggestion.

    I disconnected the encoder cable and checked connectivity from the black flat 9 pin Molex plug (supplied with the encoder) to the DB9. Everything seems to check out -- connectivity only where I expect to see it as I move from pin to pin at each end.

    I plugged the 4th axis cable and motor wires into X and set the machine to 3 axis.

    I get a "410 X axis(1) positioning error". Changing motor direction in config changes the direction I get from the jog panel/MPG. The motor does spin in a different direction when I move to a different direction on the panel/MPG.

    The X DRO moves in both directions.

    Switching back to the DC single and 4 axes I again get spin in only one direction, no matter if reversed or not in config or if the jog/MPG is commanding a different direction. The DRO for B only gets larger this time. The 410 now shows the 4th axis in the error message. Moving the motor by hand has the DRO track in both directions.

    This cable is made using fresh shielded power and encoder cable from Ajax and the shields are grounded.

    Could I have the fiber connections backwards or would that keep it from working at all?

    While I've had the DC Single for a long time (I think I bought it with the rest of the control or shortly thereafter) these past few days are the first time it has had power connected to it.

    Is there anything in the pages of numeric parameters that might not have been set correctly that could cause this? If you tell me what param #s to check I'll do that and report what they are. I set them according to

    CNC Services Northwest - Rotary Fourth Axis Tips

    Can you suggest anything else to try?

    ETA: there is no "runaway" activity. One incremental jog button press or click on the MPG triggers a spin that then coasts to a stop.

    thanks,
    Michael

  5. #5
    Join Date
    Mar 2016
    Posts
    42

    Re: I think I have an encoder wiring issue

    If you swapped the 4th axis encoder cable as well as the 4th axis motor power cable with the X axis and are getting Position Error on the X axis when attempting to move the X axis then it means that you have an issue either with the motor, encoder, or the cabling.

    You didn't mention anything about connecting the X axis encoder cable and the X axis motor power cable to the 4th axis. Have you tried doing that and commanding the 4th axis to see if you still get it to fault? If you do, then it means that there is an issue with the DCSINGLE.

    Have you also attempting to reverse the motor power wires on the drive rather than in the Motor Configuration menu?

    Please generate a current report from your system following the attached tech bulletin and email it to [email protected].

    If the fiber optic cables were not installed correctly, you would be getting drive communication messages. You can easily verify that they are installed correctly by verifying that the DF LED on the DCSINGLE is lit.

  6. #6
    Join Date
    Sep 2004
    Posts
    107

    Re: I think I have an encoder wiring issue

    It appears that swapping the motor cables at the drive fixed it.

    Mr. Murphy always finds a visit to my garage to be productive, and with his help I of course hooked up the two motor wires (which since they were not labeled at the motor gave me no clue which was + and which was -) backwards. Since I changed the rotation in config, which I thought would work just like physically switching the wires, and that made no difference I spent days not looking at the motor wires.

    Can you give me a dumbed-down explanation of why physically changing the wires fixed this but reversing rotation in the control did not? I'd like to learn something from this exercise.

    It was nice to see that my first attempt at soldering up a big Amphenol connector was not the problem.

    The Centroid stuff so far has always been pretty straightforward for me to get running, so having these never before seen errors had me very concerned.

    thanks for the advice, I appreciate it.

    cheers,
    Michael

  7. #7
    Join Date
    Jul 2007
    Posts
    236

    Re: I think I have an encoder wiring issue

    the simple reason is... the motor direction and the encoder direction must match for a servo motor to be controlled in closed loop operation....

    your motor and encoder direction did not match..so...when you flipped the motor power wires.. positive and negative.. you changed the motor direction to match the encoder direction. (you have a 50-50 chance of getting it right the first time if you don't follow the encoder install instructions..)

    interesting to note: you could also achieve the same result by flipping two channels of the encoder wiring (which changes the way the encoder counts up or down with motor CW/CCW movement , which is more hassle ..so its always easiest to swap the motor power to get the encoder and the motor direction going the same way.

    then once you have motor and encoder direction going the same way... you can use the software 'motor direction' to flip the direction of the servo motor to make the axis positive and negative movement match the Cartesian coordinate system that cnc is based on. the software flip is there so that once you have the motor and encoder hardware working in harmony.. you don't have to rewire both the encoder and the servo motor to changed direction.. ( the software flip..is not intended to match up encoder running one way and the motor the other, it is intended to swap axis direction convention. )

    post some photos of your install, we'd all like to see it!

  8. #8
    Join Date
    Sep 2004
    Posts
    107

    Re: I think I have an encoder wiring issue

    Thanks for the info!

    I don't think my camera has a wide enough lens to get a shot inside the (now rather messy) cabinet since the mill is backed into a corner with just enough room to open the doors on the cabinet. Here's a photo of the Tree 425 at a friend's shop before it moved home.

    http://www.eurospares.com/graphics/m...ee425PJ_02.jpg

    I removed the touch screen and panel from that console and have the 17" LCD monitor and keyboard there. There were some very small (450W) Yaskawa AC servos on it, it seems the deceased EE owner was trying to roll his own control and software. I put three new Centroid/Glentek 40 in/pound servos on it.

    The 170mm new old stock NEWS(Yuasa) rotab was originally purchased as a spare for a VMC that was worn out and scrapped before the table was ever used. The guy who bought the table when it was surplussed never used it before he sold it to me.

    http://www.eurospares.com/graphics/m...quenocover.jpg

    http://www.eurospares.com/graphics/m...anualcover.jpg

    Here's the test certificate that came with the table:

    http://www.eurospares.com/graphics/m...testreport.jpg

    My first conversion was a Tree 325 that had a Dynapath control On that one I used the OEM Baldor motors. The Centroid let me eliminate two of the three cabinets on that. That went to a friend when the apparently never used-but-stripped 425 became available. I put another basic 3 axis control in the 325 and moved the DC3IO with all the software unlocks over to the 425.

    My 425 is actually a TOS FV30 made in the Czech Republic and rebadged by Tree. The weight is approx 6000 pounds. It is CAT40 with 6" of quill, and 30/15" X and Y, a 12x51" table and 6000 RPM.

    cheers,
    Michael

  9. #9
    Join Date
    Jul 2007
    Posts
    236

    Re: I think I have an encoder wiring issue

    nice heavy duty machine..well worth the retrofit! I always liked those tree knee mills with the rigid ram...
    looking to the future.. if you take on any new dc servo motor based cnc retrofits ...the "allin1dc cnc control board" is basically the DC3io and the CNC10 cpu combined into one board..(new cpu is named the mpu11 )

    and.. fyi...the Oak board can run those old yaskawas that were removed as well..sigma i,and ii's run great with the oak board..
    here is a video talking about the new oak cnc control board for ac brushless motors.
    https://youtu.be/goIE7KRxXcI

  10. #10
    Join Date
    Sep 2004
    Posts
    107

    Re: I think I have an encoder wiring issue

    Keith, as a retired hobbyist machinist/racer I'm hoping that once I get a full 4 axis mill going I'll never need to do another mill conversion!

    I have to admit that a small CNC lathe sounds appealing, but I'm trying to resist that as I've got a very nice Mori Seiki manual lathe that is probably more than I really need.

    http://www.eurospares.com/graphics/m.../mymoribig.jpg

    But if I could find a small yet nice additional lathe that would fit in the shop, especially if it could use those small Yaskawa AC servos and drives,

    http://www.eurospares.com/graphics/m...Tree425_22.jpg

    a 2 axis conversion might happen. I wouldn't need a CNC lathe for peeling off .2" per pass, the Mori does that kind of work. But detail/radii/threading could make a CNC lathe somewhat justifiable. What does Centroid/Ajax offer there?

    I may post back in the not too distant future for tips on servo tuning on the mill. It has always seemed very abrupt and "bang-y", even when doing digitizing with a slow feed. That seems like it is maybe trying to accel/decelerate too harshly. I've run the autotune but it doesn't seem to make any detectable difference. I saw a video of some servo tuning of a Centroid system on the bench (possibly posted by you or Scott) but I lack the training to make much sense of it (along with no o-scope as used in the video).

    cheers,
    Michael

  11. #11
    Join Date
    Jul 2007
    Posts
    236

    Re: I think I have an encoder wiring issue

    send me a copy of the system report file.. i can make suggestions to the pid parameters to smooth things out.

  12. #12
    Join Date
    Sep 2004
    Posts
    107

    Re: I think I have an encoder wiring issue

    Keith, I just generated a report and mailed it to the support email address. Thanks for the help.

  13. #13
    Join Date
    Jul 2007
    Posts
    236

    Re: I think I have an encoder wiring issue

    michael, hello. please send me a copy as well. Keith

  14. #14
    Join Date
    Jul 2007
    Posts
    236

    Re: I think I have an encoder wiring issue

    thanks for sending the system report.. from that information i can see that the axis acceleration times are very low.. (fast accell) this is due to the fact that the machine axis overall turns ratios are large 10:1 or higher..
    since the machine has such hi turns ratios.. the motor has a ton of mechanical advantage over the axis.. so autotune is tuning the servos out to the max accell rate that it can stay in control... well. that's great.. but thats a bit too quick.. so its a simple thing to slow down the accel rates for each axis...... change all of these in image below to .5 (seconds) . fyi.. servo motor accel time is the time it takes to accelerate to maximum velocity...
    Attachment 329910

  15. #15
    Join Date
    Dec 2008
    Posts
    44

    Re: I think I have an encoder wiring issue

    Quote Originally Posted by cnckeith View Post
    thanks for sending the system report.. from that information i can see that the axis acceleration times are very low.. (fast accell) this is due to the fact that the machine axis overall turns ratios are large 10:1 or higher..
    since the machine has such hi turns ratios.. the motor has a ton of mechanical advantage over the axis.. so autotune is tuning the servos out to the max accell rate that it can stay in control... well. that's great.. but thats a bit too quick.. so its a simple thing to slow down the accel rates for each axis...... change all of these in image below to .5 (seconds) . fyi.. servo motor accel time is the time it takes to accelerate to maximum velocity...
    Attachment 329910
    I would also set Kd for the 4th axis to 3

  16. #16
    Join Date
    Sep 2004
    Posts
    107

    Re: I think I have an encoder wiring issue

    FWIW, tonight finished the great CNC rotab commissioning project.

    All the hardware stuff is done -- cover plates, mounting the Amphenol connector, the clamp is cleaned up/unstuck from sitting for too many years since it left the factory, wiring sorted etc etc etc.

    Anything left should be software stuff. I'll run an autotune on the table and see what that looks like after I set Kd for the 4th axis to 3 as suggested above.

    Next up is a major picking up of all the tools, wiring, misc parts that have been dragged out and scattered around the house and garage the last month. I also need to figure out what kind of tooling to buy (or build) to make the table usable -- 5C/3J/ER collet chuck, 3 or 4 jaw chuck, trunnion table/angle plates, etc.

    Thanks to everyone for the advice, I appreciate the help.

    cheers,
    Michael

  17. #17
    Join Date
    Jul 2007
    Posts
    236

    Re: I think I have an encoder wiring issue

    after auto tune.. re adjust accel rate to .5.. as autotune will set them to the quickest possible.

  18. #18
    Join Date
    Sep 2004
    Posts
    107

    Re: I think I have an encoder wiring issue

    Keith, in TB045 (rev1) it says that the reciprocal of the change to accel time should be applied to the Ka number. Is that applicable to my machine? That document seems to be for an earlier version than CNC10. It also mentions TB044 but I don't have a copy of that.

    eg: for X of Ka of 26 and accel of .17, should Ka be changed to 8.85 when changing accel to .5?

  19. #19
    Join Date
    Jul 2007
    Posts
    236

    Re: I think I have an encoder wiring issue

    yeah, its a good idea.. think of ka like when you step on the gas to pass a car at 45 mph..its how hard you step on it..

Similar Threads

  1. Encoder wiring
    By kck63 in forum MetalWork Discussion
    Replies: 0
    Last Post: 01-21-2016, 05:12 AM
  2. Pic encoder wiring A B i and T
    By gurvy in forum PIC Programing / Design
    Replies: 3
    Last Post: 08-16-2007, 02:03 PM
  3. Encoder Wiring Help
    By amitim in forum CNC Machine Related Electronics
    Replies: 2
    Last Post: 05-02-2006, 10:31 PM
  4. Help with selector switch wiring issue (***actually a motor issue***)
    By BEDFORD in forum Charter Oak Automation Support Forum
    Replies: 7
    Last Post: 04-07-2006, 09:19 PM
  5. Encoder wiring
    By balsaman in forum Gecko Drives
    Replies: 15
    Last Post: 12-30-2003, 06:01 PM

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
  •