Welcome to Our Community

Some features disabled for guests. Register Today.

Problem engraving PCB on Ox

Discussion in 'CNC Mills/Routers' started by micksulley, Oct 3, 2018.

  1. micksulley

    Builder

    Joined:
    Jul 11, 2017
    Messages:
    16
    Likes Received:
    0
    I am trying to engrave a PCB on my Ox using a 0.5mm cutting bit, and consistently breaking bits. I am cutting at 200 mm/min. The reason seems to be that it judders.
    I have checked the pulleys, all tight. Checked the belt tension, seems OK. I have cleaned the wheels, belts, pulleys, etc.
    I have changed microsteps from 8 to 32, slight improvement but still a definite judder.
    If I increase the feed rate to around 2000 mm/min it seems better, but still judders, and that is much too fast for a 0.5 mm bit.
    The judder occurs on all 3 axis, worst on Y, not quite as bad on X and on Z I can tell that there is a judder but so slight that I wouldn't bother about it.

    I am running it from bCNC on a Raspberry Pi.

    Any suggestions?
     
  2. GrayUK

    GrayUK Openbuilds Team Elder
    Staff Member Moderator Builder

    Joined:
    May 5, 2014
    Messages:
    1,668
    Likes Received:
    627
    A few questions. :rolleyes:
    1 - Does it, and has it, always juddered, no matter what?
    2 - Does it judder if you cut something more substantial?
    3 - Does it judder if you Aircut?
    4 - Does it judder the whole length of an axis, or just in an area?
    5 - What is your general depth of cut when getting judder?
     
  3. micksulley

    Builder

    Joined:
    Jul 11, 2017
    Messages:
    16
    Likes Received:
    0
    1- Not sure. In the early days I did cut a PCB successfully, so the judder either wasn't there or wasn't as bad, but I made so many mistakes it is difficult to remember exactly how it went.
    2-5 If I cut wood or MDF I can use a higher feed rate and yes it still judders but not enough to cause a problem. If I lower the feed rate to 200 it judders the same as PCB. Judder in air as well and for the whole of the available cut area. Depth of cut makes no difference to the judder.

    Looking at the drive pulleys it seems that it is the stepped motors that are causing it, if it was a mechanical issue I would expect to see the pulley rotating smoothly and the gantry jerking, but it does look like the pulley rotation is jerky.
     
  4. GrayUK

    GrayUK Openbuilds Team Elder
    Staff Member Moderator Builder

    Joined:
    May 5, 2014
    Messages:
    1,668
    Likes Received:
    627
    6 - Oh by the way. Are your steppers silent when they run, or noisy? i.e. Grating or grinding? :rolleyes:

    Personally, I would loosen and slip the belt off one of the steppers. So you have a No Load situation.
    If necessary I'd put a bit of tape on the pinion to help with observation.
    Next, I would double check the number of Amps you are sending to the steppers.
    What steppers are you using and from where?
    A video could be really helpful if you can manage it. :)
     
  5. micksulley

    Builder

    Joined:
    Jul 11, 2017
    Messages:
    16
    Likes Received:
    0
    OK here is a video, there is road noise in the background I'm afraid, but you can still hear the sound of the stepper motor.
     

    Attached Files:

  6. GrayUK

    GrayUK Openbuilds Team Elder
    Staff Member Moderator Builder

    Joined:
    May 5, 2014
    Messages:
    1,668
    Likes Received:
    627
    My Apologies.
    Sorry, you misunderstood me. I meant a video of it working normally. Running up and down the Y and X axis under normal load.
    Although a video of what you have at the moment, but running the steppers as you would normally may help.

    What settings did you have set on that video?
     
  7. Kevon Ritter

    Kevon Ritter Veteran
    Builder

    Joined:
    Apr 30, 2015
    Messages:
    597
    Likes Received:
    294
    None of that stuff matters nearly as much as runount does. What are using to cut? small endmills are guaranteed to break without even moving through the material to be cut. If you want to engrave without a higher pecision spindle, stick with v engravers.
     
  8. micksulley

    Builder

    Joined:
    Jul 11, 2017
    Messages:
    16
    Likes Received:
    0
    Here is a video of it cutting my PCB in air, the volume is a bit low but if you turn it up you can hear the motors. It is running at 75 mm/min.
    How do I measure the current? Do I measure the current in one of the wires to the motor or the 24v supply to the board?
     

    Attached Files:

  9. GrayUK

    GrayUK Openbuilds Team Elder
    Staff Member Moderator Builder

    Joined:
    May 5, 2014
    Messages:
    1,668
    Likes Received:
    627
    Research your steppers and find what Amps they should be getting.
    Check your control board and see what Amps you are sending to them.
    Check your drivers and see what they are set at.
    I have never milled PCB so am not fully aware of the speeds normally used for this project.
    I'm hoping someone with experience will step in and advise.
     
    Mark Carew likes this.
  10. joe williams

    joe williams Well-Known
    Builder

    Joined:
    Nov 9, 2015
    Messages:
    100
    Likes Received:
    59
    Tell us the brand and model # of your steppers as well as the brand and model # of the drivers...you may have individual drivers for each motor or you may have all the drivers together on a single board...this will help troubleshoot the problem.
     
    Mark Carew likes this.
  11. micksulley

    Builder

    Joined:
    Jul 11, 2017
    Messages:
    16
    Likes Received:
    0
    I bought it all as a kit from Oozenest. The motors are Motech Motor MT2303HS280AW-C 1.8 degree/step, the controller is Spark Concepts xPro V3 running
    grbl 1.1f, with all 4 drivers on the single board. I'm not able to check the current at the moment, I will do that tomorrow.
    Thanks for your help.
     
  12. MaryD

    MaryD OpenBuilds Team

    Joined:
    Apr 13, 2016
    Messages:
    649
    Likes Received:
    639
    Perhaps @Ryan Lock would have some ideas as well.
     
    joe williams likes this.
  13. Peter Van Der Walt

    Peter Van Der Walt OpenBuilds Team
    Staff Member Moderator Builder Resident Builder

    Joined:
    Mar 1, 2017
    Messages:
    15,061
    Likes Received:
    4,313
    The background high-pitched noise going up and down in frequency, along with the jump, jump jump, looks to me like the classic overheating DRV8825 (Set too much current). When a DRV8825 overheats it shuts itself down, cools down a little, re-enables itself: repeat... Feel if the drivers are too hot, then either add some forced air cooling, or turn the current down just a little to keep them cool. Or upgrade Hooking Up External Drivers to the CNC xPRO V3
     
    GrayUK, joe williams and MaryD like this.
  14. micksulley

    Builder

    Joined:
    Jul 11, 2017
    Messages:
    16
    Likes Received:
    0
    I measured the current, put my meter in series with one of the motor wires. When moving it jumps around between about +2.1A and -2.1A. I took a video so you can see. Is that how you wanted me to measure the current?
    The specs say the driver is 2.5A and the motors 2.8A
    I couldn't easily get in to feel the drivers so measured the temperatures with an infra red probe, they are all around 18-20 degrees C at the moment so I guess that is not the problem.
     

    Attached Files:

  15. micksulley

    Builder

    Joined:
    Jul 11, 2017
    Messages:
    16
    Likes Received:
    0
    I just connected an oscilloscope between the motor lead and earth, I have uploaded 3 pictures showing the trace, all running a move at 50 mm/minute. Do these traces look normal? I was expecting to see something resembling a sine wave. IMG_20181006_135314.jpg IMG_20181006_135326.jpg IMG_20181006_135335.jpg
     
  16. JustinTime

    JustinTime Veteran
    Builder

    Joined:
    Dec 18, 2013
    Messages:
    784
    Likes Received:
    259
    These are normal. The word 'chopper' sometimes comes up when talking about stepper motors and their drivers. This is the reason for that.
     
  17. joe williams

    joe williams Well-Known
    Builder

    Joined:
    Nov 9, 2015
    Messages:
    100
    Likes Received:
    59
    Your graphs are showing pulses...this is how stepper motors work. There is a little bit of 'ringing' on the beginnings and endings of the pulses, but it doesn't appear to be significant. A question or two...are you using shielded cable to your steppers?, do you have any gear spewing out high frequency noise? (chinese spindle, radio transmitter, wifi router close by,etc..) How is your system grounded?(Pi3, xpro, motors, power supply, mechanical)...just stabbing in the dark to see if we hit something?!
     
  18. micksulley

    Builder

    Joined:
    Jul 11, 2017
    Messages:
    16
    Likes Received:
    0
    Stepper cables are not shielded.
    Nothing generating hf noise, nearest wireless router is 15 metres away or more.
    The 24volt power supply is connected to the 240v mains and earthed, although the 24v output is floating, neither side connected to earth. The aluminium framework of the Ox is not earthed.

    I have looked for info on shielded cables for steppers and the general view seems to be that it would protect other stuff from noise generated by the steppers, but it doesn't help the stepper. Do you think that earthing would help?

    I have done some more testing today with different micro steps. Obviously with full step the pulley 'ticks' round 1.8 degrees at a time, sounds and feels horrible. With greater micro steps the pulley movement is jerky, as you saw on the video I posted a few days ago, but the jerkiness is in 1.8 degree increments, so within the 1.8 degrees it seems to move part way and then click to the end of the 1.8. degrees. I don't know if that helps to diagnose it.
     
  19. David the swarfer

    David the swarfer OpenBuilds Team
    Staff Member Moderator Builder Resident Builder

    Joined:
    Aug 6, 2013
    Messages:
    3,463
    Likes Received:
    1,915
    oh, an 8825 driver is never going to do 2.5 amps reliably. try 1.25
     
    joe williams likes this.
  20. micksulley

    Builder

    Joined:
    Jul 11, 2017
    Messages:
    16
    Likes Received:
    0
    I've just tried turning the current down and it makes the jerkiness more pronounced. Out of interest, the spec for the DRV8825 says that it will drive 2.5 amps, why do you say that it won't? Is that from experience?
     
  21. David the swarfer

    David the swarfer OpenBuilds Team
    Staff Member Moderator Builder Resident Builder

    Joined:
    Aug 6, 2013
    Messages:
    3,463
    Likes Received:
    1,915
    yes, experience, they just get too hot. unless 'actively cooled' which means a heatsink and a good fan for airflow.

    the spark website says
    "Drive 4 motors with DRV8825 Stepper Drivers - 2.5A (peak) with 1.75A (RMS) with up to 1/32 microstepping".
    so, please set to 1.5amps and 8x microsteps and test (including checking that the PSU is stable).

    in electronics when they say 'xAmps peak' they mean 'under ideal, contrived, laboratory conditions, maybe water cooled, with a fan, with everything else exactly right' (-:
    the rest of us get maybe 80% of that under good conditions, or the smoke gets out.
     
  22. micksulley

    Builder

    Joined:
    Jul 11, 2017
    Messages:
    16
    Likes Received:
    0
    OK, yes I agree that spec vs reality are often different.
    Mine do have heat-sinks and fans, don't know if that is how Spark supply the board or if Oozenst add them.

    I have set microstepping back to x8 and the current is 2A, so should be OK. When I reduce the current the jerkiness does get worse. The power supply is solid at 24v, tried increasing it to 27v and it didn't seem to make any difference.

    Just looking at the spec sheet for DRV8825 it shows oscilloscope traces for the current. I have a decent scope but my knowledge is not great, if I put a low value resistor in series with one winding and put the probe across the resistor will that give me a similar trace? And will that tell me anything?
     
  23. micksulley

    Builder

    Joined:
    Jul 11, 2017
    Messages:
    16
    Likes Received:
    0
    Just taking a step back here, the problem as I see it is that the movement is jerky at low speeds and the jerk it a full step intervals. As can be seen in first video I posted, given that it is 8 microsteps to move the 1.8 degrees, for each 1.8 degrees it seems to move maybe 4 or 5 microsteps smoothly then jump the last 3 or 4.

    Can anyone confirm that the Ox on 1/8th microsteps should run smoothly at 100 mm/min (2600 steps/min), or am I chasing the impossible?
     
  24. Giarc

    Giarc OpenBuilds Team
    Staff Member Moderator Builder Resident Builder

    Joined:
    Jan 24, 2015
    Messages:
    3,017
    Likes Received:
    1,682
    What happens if you double it? I was milling a PCB with a 30 degree v bit at 500 mm/min with no problems. I drilled the holes much slower, though. I have accidentally started a v-carve using F-engrave forgetting to change the default speed of 127mm/min (exactly 10 rpms for my machine). Holy crap was it jerky and noisy! I use lead screws so I probably have to spin my motors more than twice as fast as your belt drive so I couldn't imagine how slow yours must be turning at 100 mm/min. Maybe, try milling the pcb at a faster speed. Or is this jerky at all speeds?
     
  25. micksulley

    Builder

    Joined:
    Jul 11, 2017
    Messages:
    16
    Likes Received:
    0
    It is jerky at all speeds, however as the speed increases it becomes less noticeable. I have just taken videos at various feed rates, then slowed down the higher feed rates and you can see that it is pretty much the same as at the lower rates.
     
  26. David the swarfer

    David the swarfer OpenBuilds Team
    Staff Member Moderator Builder Resident Builder

    Joined:
    Aug 6, 2013
    Messages:
    3,463
    Likes Received:
    1,915
    looks like this is a known problem with 8825 dirvers....
    possible solution is to set fast decay mode (maybe that pin is not soldered correctly?).
    DRV8825 in 32 micro step mode - Motor Drivers Forum - Motor Drivers - TI E2E Community

    I am also wondering if maybe one of your microstep selection jumpers is not making contact and you are getting full steps.
    can you confirm that when you change microstep settings the distance moved for 200 steps actually changes?
     
  27. Ryan Lock

    Ryan Lock Veteran
    Builder

    Joined:
    Oct 12, 2014
    Messages:
    530
    Likes Received:
    293
    Hi,

    If you need to constantly run at low speeds, i would recommend putting the board into fast decay mode using the solder jumpers on the back. I think this will help a lot for your use case.

    Best Regards

    Ryan Lock
     
  28. micksulley

    Builder

    Joined:
    Jul 11, 2017
    Messages:
    16
    Likes Received:
    0
    I tried setting the solder jumpers to fast decay and it made the problem very much worse. A bit more research told me that some xPro boards are marked incorrectly, so I suspect that mine is one of those and I had actually set it slow decay, hence the problem got worse. Unfortunately the solder jumper pads are so small and fragile that in removing them with a solder sucker I removed some of the pads, so there is no way for me to try the jumpers the other way.

    I have also tried using 30 degree V bits and I am having more success with those. I will continue to experiment.
     
  29. GrayUK

    GrayUK Openbuilds Team Elder
    Staff Member Moderator Builder

    Joined:
    May 5, 2014
    Messages:
    1,668
    Likes Received:
    627
    micksulley.
    You might try this thread:-
    CNC xPRO Driver
    michael is the guy who made the Xpro, and I'm sure he'd be able to help you one way or another.
    If you don't get a response, try getting in touch directly with the PM service on the Envelope up top.
     

Share This Page

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice