Welcome to Our Community

Some features disabled for guests. Register Today.

OpenBuilds LEAD CNC

Discussion in 'CNC Mills/Routers' started by MaryD, Nov 20, 2018.

  1. jamin35008

    jamin35008 Well-Known
    Builder

    Joined:
    Feb 7, 2019
    Messages:
    136
    Likes Received:
    44
    Are you using the cnc4newbies z or doing your first run with the stock lead?
     
  2. ljvb

    ljvb Well-Known
    Builder

    Joined:
    Feb 14, 2019
    Messages:
    276
    Likes Received:
    87
    I'm using the stock lead at the moment.
     
  3. WR_Rider

    WR_Rider New
    Builder

    Joined:
    Apr 3, 2019
    Messages:
    63
    Likes Received:
    3
    Hello, first post, hopefully this is a good spot for a couple questions.

    I'm trying to decide between this "lead" build or a "workbee" build. To me they generally look the same except a few wheels and the side plates. I guess my biggest question would be is one more "rigid" then the other? For general use would one be any better then the other? I was leaning towards the lead but I found a workbee (non amazon) for $150 less so I'm not sure which way to go.

    Thx
     
  4. wojak

    wojak New
    Builder

    Joined:
    Jul 5, 2018
    Messages:
    20
    Likes Received:
    30
    I didnt had opportunity to test workbee, however amount of changes I've made in lead makes me think workbee is solving few of them.
    IMHO lead cnc is like workbee without custom plates.


    About z-axis. My concept seems to be fail :( It is bending, what is annoying I am unable to find where.
    Looks like it is sum of problems, a little bending on X axis, a little bending on x axis carriage, and a little on z-axis (5mm alu + 5mm stainless steel support, with 15mm supports on each side).
    I can literally bend it with finger :( What is annoying I've considered lever effect when designing.


    [​IMG]

    EDIT!
    NVM seems everything is ok, fixed few screws, mounted spindle and it works great (except bit is not straight after screwing...)

     
    #454 wojak, Apr 9, 2019
    Last edited: Apr 9, 2019
  5. SpringBob

    SpringBob New
    Builder

    Joined:
    Nov 9, 2017
    Messages:
    16
    Likes Received:
    5

    What are you using to generate the Trochoidal milling g-code?
     
  6. wojak

    wojak New
    Builder

    Joined:
    Jul 5, 2018
    Messages:
    20
    Likes Received:
    30
    estlcam
     
    SpringBob likes this.
  7. ljvb

    ljvb Well-Known
    Builder

    Joined:
    Feb 14, 2019
    Messages:
    276
    Likes Received:
    87
    How are you guys sizing the spoilerboard.. I had to tweak the soft limits x y dimensions in the firmware as the stock values the moving parts kept hitting things. so after a little trial and error I changed the values to avoid hitting things.

    That said.. when I designed my spoilboard in Fusion, I used the work area numbers identified in the Lead specs. That appears to fall well outside the limits though. So what values are people using for width and length.

    This is my current build. spoilboardv3.png
     
  8. Alex Chambers

    Alex Chambers Master
    Moderator Builder

    Joined:
    Nov 1, 2018
    Messages:
    2,680
    Likes Received:
    1,321
    Can't help you with actual numbers, but had an identical situation with my workbee. Just to be on the safe side and make sure it didn't reach its soft limits I designed my spoiler board surfacing 5mm less than the machine dimensions I had to set.
    Alex.
     
  9. SpringBob

    SpringBob New
    Builder

    Joined:
    Nov 9, 2017
    Messages:
    16
    Likes Received:
    5
    That is what I'm using as well.
     
  10. ljvb

    ljvb Well-Known
    Builder

    Joined:
    Feb 14, 2019
    Messages:
    276
    Likes Received:
    87
    I'm having an odd problem.. Hoping someone came across it and there is a solution before I start pulling the machine wiring apart.

    When I turn off my router (the Dewalt one) the limit switch trips. Not a deal killer for now, as I have not made run any large jobs yet, small test runs. I'm more concerned that there is a grounding issue or a short somewhere that might trigger in the middle of a larger job and screw things up.
     
  11. Alex Chambers

    Alex Chambers Master
    Moderator Builder

    Joined:
    Nov 1, 2018
    Messages:
    2,680
    Likes Received:
    1,321
    I assume your mains cable to the router is well away from stepper and limit switch wiring? When you switch off your router there will be a large back emf. The low voltage wiring can act as an arial and affect the electronics.
    Alex.
     
  12. Dmhaes

    Dmhaes New
    Builder

    Joined:
    Jan 3, 2019
    Messages:
    40
    Likes Received:
    20
    Mine does the same thing, I’m probably going to reroute the power cable for the router overhead with the dust collection instead of along the drag chain.
     
  13. Alex Chambers

    Alex Chambers Master
    Moderator Builder

    Joined:
    Nov 1, 2018
    Messages:
    2,680
    Likes Received:
    1,321
    Do it now @Dmhaes, the voltage spikes from switching off your router can be several 100's of volts. The transmission through other wiring acting as aerials is not efficient but your electronics could be permanently damaged - its not just an inconvenience.
    Alex.
     
    SpringBob likes this.
  14. ljvb

    ljvb Well-Known
    Builder

    Joined:
    Feb 14, 2019
    Messages:
    276
    Likes Received:
    87
    I'm familiar with the issue (my degree is Electrical Eng :) ) All of the low voltage wires go to the rear, high voltage to the front. They are are numerous feet apart. The only real connection is the power supply and router are plugged into the same power strip.
     
  15. Alex Chambers

    Alex Chambers Master
    Moderator Builder

    Joined:
    Nov 1, 2018
    Messages:
    2,680
    Likes Received:
    1,321
    That's good practice @ljvb, not everyone knows that (see above). The symptoms (as I am sure you are aware) suggest voltage spikes getting through to the low voltage side - a decent power supply should stop interference via your power strip. The only suggestions I have are checking all the connectors on the wiring and possibly shielding the limit switch wiring.
    Good luck, Alex.
     
  16. syl

    syl New
    Builder

    Joined:
    Jan 7, 2019
    Messages:
    49
    Likes Received:
    12
    Is it possible to configure the width and length of the lead cnc?
     
  17. ljvb

    ljvb Well-Known
    Builder

    Joined:
    Feb 14, 2019
    Messages:
    276
    Likes Received:
    87
    Yes, in the configuration options of the openbuilds software. You can set the limits and then upload to the firmware. Takes a little trial and error to get the exact values.
     
  18. ljvb

    ljvb Well-Known
    Builder

    Joined:
    Feb 14, 2019
    Messages:
    276
    Likes Received:
    87
    And for an update.. my laptop crapped out (the one I was using to connect to the Lead. So I picked up a Raspberry Pi 3 b+ the other day and the POE hat. (I have a rather hefty POE.. Power over Ethernet switch) . I'll be running cncjs and connecting to it remotely with another laptop to control the lead. I have not picked up a screen yet, the options are annoying and cheap.. or really good and expensive.
     
  19. gregers05

    gregers05 Well-Known
    Builder

    Joined:
    Feb 15, 2019
    Messages:
    131
    Likes Received:
    33
    Would be curious to get more details on that setup once you get it figured out. I bought a cheapo chinese tablet to run mine on to start out so idk how long ill keep that around
     
  20. ljvb

    ljvb Well-Known
    Builder

    Joined:
    Feb 14, 2019
    Messages:
    276
    Likes Received:
    87
    It's pretty simple really. I was using CNCJS on my laptop. I'll still be using it, just on the Pi. the POE hat is just so I can power it from the switch, so I don't need to add another power block to the strip. cncjs is web based, and I have a folder mounted that when I drop a file into it it is loaded automatically. then I just use my tablet or laptop (other laptop, I have a few) to connect to the web interface, and manage it. Similar in nature to openbuilds website.
     
  21. Mark Carew

    Mark Carew OpenBuilds Team
    Staff Member Moderator Builder Resident Builder

    Joined:
    Jun 28, 2013
    Messages:
    2,742
    Likes Received:
    2,407
    Hey guys quick LEAD Build update:
    Thanks to the hard work of the OpenBuilds Team - The videos in the build section have been updated to now reflect the BlackBox wiring! :thumbsup:
    All the videos related to the Build in that section have been cleaned up and look really sharp and are easy to use now.
    Hope this helps, Thank you
    Mark
    Capture.JPG
     
    gregers05 and sharmstr like this.
  22. gregers05

    gregers05 Well-Known
    Builder

    Joined:
    Feb 15, 2019
    Messages:
    131
    Likes Received:
    33
    Perfect! Been waiting on my black box and worong kit that were on back order and they came in today!
     
    Mark Carew and sharmstr like this.
  23. Mason Dixon

    Builder

    Joined:
    Apr 19, 2018
    Messages:
    4
    Likes Received:
    0
    Do the 2 leadscrew nuts for each leadscrew eliminate backlash? I haven't had time to go through all 16 pages for this design, thanks.
     
  24. Mark Carew

    Mark Carew OpenBuilds Team
    Staff Member Moderator Builder Resident Builder

    Joined:
    Jun 28, 2013
    Messages:
    2,742
    Likes Received:
    2,407
    As you know there really is no such thing as completely zero backlash in a system as everything bends twist and moves even if it's just a little. But having the double nut does help in adding a preload to engage more of the threads and the resulting resolution is very impressive.
     
    sharmstr likes this.
  25. jbhurst

    jbhurst New
    Builder

    Joined:
    Jun 19, 2017
    Messages:
    19
    Likes Received:
    3
    I'm not sure if this should be its own thread, but I'm in limit switch hell with my 1010, so I thought I'd try here first.

    The machine seemed to run fine for a while and successfully milled some acrylic and routed some PCB's, but mostly was doing light duty and calibration work.

    Once I switched to routing fully copper clad 18"x24" x062" PCB's I would get false limit switch triggers. They were repeatable and seemed to happen at about the same point of the cut on the work piece. I know this is a fairly common issue, and I've done some searching for solutions and tried some things that I'll list below, but here are the weirder points:
    • The problem with false alarms on copper surfaces would consistently go away when I would clamp the work piece to ground. I don't like this solution because the route goes all the way around the work piece and the operator has to clamp it, then move the clamp to the other side of the spindle once it passes.
    • Since the Alarm message doesn't say which axis had a limit trigger (WHY?!?) I tried running the test, disconnecting each limit switch form the BB in turn and then all of them. The problem happened even with all of the switches physically disconnected from the BlackBox. Huh?
    • After even more testing, I finally ran into a case where the alarm would trigger even with the work piece grounded. In this case, I was trying to mill about 0.015" into a hard G10 table at the same time and it made it 3/4 of the way through (instead of about 1/4) and then stopped on the return path but showed no error (bCNC) and stayed in a run state. Looking at the log though, it showed an alarm1 error. I'm thinking this trigger may be a different problem from the others and may be related to a current spike form the steppers and/or spindle.

    I've already done some searching on the issue and am working my way through the items in this thread: Electronics 101.


    I'm starting to think I may have two problems: EMI/static buildup on a large conductive work surface and EMI caused by current spikes on the spindle or steppers.

    My build is a LEAD1010 with a BlackBox controller. The high torque steppers weren't available when I ordered, so I used these NEMA 23's: https://www.amazon.com/gp/product/B00PNEPW4C/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1
    • 3Nm(425oz.in) holding torque
    • NEMA 23 bipolar 2.24"x2.24"x4.45" 4-wire
    • 1.8 deg. step angle(200 steps/rev)
    • Rated current 4.2A & resistance 0.9ohms
    The spindle is a Dewalt 611. I also have 2 usb cameras, a USB microscope, and an additional light and 12v power line to the gantry. The system acts as a CNC mill/router for large panel PCB's and also does some inspection and work and is a test bed for some computer vision experiments I'm doing. The spindle and a shop vac are plugged into the normally off ports on one of the new AC relays OB started selling. They are driven by the coolant output because the relay output on my BB isn't working. Maybe that indicates other possible problems on the BB that could be related to the false triggers? My BB was from the very first batch, I think.

    I have no background in EE or really in building physical things in general. Most of my work has been in software, and I took this on as a learning project and to build a platform for some software experiments I want to do. It has definitely been a learning experience, and some of it has been fun, but I was so frustrated last night I was considering scrapping the whole thing and getting something off the shelf. I have a big sunk cost problem though, and I really do like using an open hardware and software platform. For its PCB cutting production role, however, it has to be reliable and I can't afford to spend a lot more time on it and certainly can't continue to fuss with it on and off for months to come. It needs to be able to reliably cut rectangles in big circuit board panels for about 2 hours a day without error or I need to find another solution. If this machine is always going to work more like a "hobby project" with various errors to ignore and workarounds that prevent a consistent repeatable process, then I need to quit spending time on it.

    Stuff I've tried:

    Started moving the system to a star ground setup with a common terminal point on the frame with a wire that goes from there to building ground. I haven't yet done everything in the electronics 101 post above. I really thought the kit would work by following the video and using the provided parts without having to do all that extra work.

    I started with using the limit switches that came with the kit, but upgraded the Y and Z to beefier (but still pretty cheap) ones like these https://smile.amazon.com/gp/product/B06WRN7FQB/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1.

    I grounded the vacuum hose in more places.

    I re-routed the spindle power down the vacuum line instead of along the gantry (but then moved it back when it didn't help since it was ugly).

    I removed the microscope and its light to reduce weight, power, and a possible source of EMI.

    I did the test mentioned above removing the limit switches one at a time and still triggered the alarm with none plugged in. (Except the probe - that can't do it if not probing, right?)

    Stuff I'm looking into trying:

    Moving to NC instead of NO for the limits. This was next on my list to try, but after I was getting the errors without the limits even attached, I decided to post here first.

    I'd rather not re-wire the whole thing, so I'm trying here before replacing all of the kit wiring with more expensive shielded wiring like this: https://smile.amazon.com/50-FEET-18...stepper+cable&qid=1557321843&s=gateway&sr=8-1 and a 2 wire version for the limits.

    I'm also considering buying a 3rd round of limit switches and was considering these https://smile.amazon.com/dp/B00FDT7...olid=3AO4XSHI98ODL&psc=0&ref_=lv_ov_lig_dp_it but would like suggestions.

    Modifying my table with routed channels for ground wire to make permanent contact with tooling pins. Tricky because the pins are in plastic plug tooling and come in and out.

    Adding a conductive, grounded wasteboard with plated holes that contact the pins (this wouldn't be that hard for me to do, but I'm worried it won't fix the problem due to the alarms triggering even with the switches off).

    Adjusting the power on the Nema 23's. If I max the dial on the BB out, they fail with the "!" warning. If I back off about 1/4 turn, they don't seem to fail unless they are doing so and somehow triggering an Alarm1. They also don't seem to ever get hot.

    Thanks to anyone that took the time to read all this! I really like what OpenBuilds is trying to do and how helpful the community has been.

    -Brady
     
  26. Alex Chambers

    Alex Chambers Master
    Moderator Builder

    Joined:
    Nov 1, 2018
    Messages:
    2,680
    Likes Received:
    1,321
    I won't comment on all your post - there are other people on here much more familiar with the black box than me who would be better able to advise.
    You said "still triggered the alarm with none plugged in. (Except the probe - that can't do it if not probing, right?)" - well no - if it's picking up EMI then its acting as an aerial and could be creating a false signal. Unplug it and try again - won't necessarily solve your problem, but will eliminate another possibility.
    Good luck!
    Alex.
     
  27. jbhurst

    jbhurst New
    Builder

    Joined:
    Jun 19, 2017
    Messages:
    19
    Likes Received:
    3
    Ok, I'll try that. I meant though, that even if it did trigger, it shouldn't cause an alarm state unless a probe operation was in progress, right? I thought probe signals were ignored unless being polled during a probe command, but maybe that's not right.
     
  28. sharmstr

    sharmstr OpenBuilds Team
    Staff Member Moderator Builder Resident Builder

    Joined:
    Mar 23, 2018
    Messages:
    2,030
    Likes Received:
    1,428
    FWIW I attach my alligator clip to my touch plate while not in use. So, I believe that you are right in thinking that the probe signal is ignored unless a G38 is running.
     
  29. Alex Chambers

    Alex Chambers Master
    Moderator Builder

    Joined:
    Nov 1, 2018
    Messages:
    2,680
    Likes Received:
    1,321
    I don't know - I don't have a black box so I don't know how it handles signals from a probe. On my workbee with a Duet controller to probe is paralleled to the Z limit switch, which is why I suggested trying that. I was going to tag some other team members to attract their attention to your post, but I see @sharmstr has already responded - I'll just add @Peter Van Der Walt in case he can help.
    Alex.
     
  30. Peter Van Der Walt

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

    Joined:
    Mar 1, 2017
    Messages:
    13,749
    Likes Received:
    4,070
    The Probe shouldnt cause an alarm.

    While unlikely, its not impossible for "all unplugged" to get a false trigger, the PCB tracks can still act like antennas, but this would require a massively abnormal amount of EMI to be a problem. The onboard circuit has a RC filter built in.

    Let me try and find some highlights from your questions, to answer:


    Because Grbl doesnt tell us: gnea/grbl
    And Grbl doesnt know because: (from Which limit switch is tripping · Issue #239 · gnea/grbl)
    "The AVR processor only provides which port experienced a pin change. It does not specify or record which pin on the port triggered."


    Whats the chance of the copper dust causing shorts? Remote, but not impossible


    Not too bad on the EMI front... way better than say a VFD spindle... Shouldnt be causing a lot of EMI


    Now this seems like a more likely source of EMI
    Plastic hose, fast moving airstream and particles... = tonnes of static electricity...
    Ever vacuum a carpet barefoot? Shocking results usually ensue right? Swopping in one of those aluminum shopvac hoses and grounding it, will help

    Did we ever work on this issue with you?
    If not:
    - Jumper in place? On which?
    - If Jumper is on M3/5 are you sure you are sending M3 S1000 (M3 alone doesnt turn on, M3 S1000 = 100% PWM = solid on)
    - If Jumper is on M8/9 then M8 should turn it on
    Revert with feedback on this, even if just to eliminate.
    NB have 24v on and BlackBox switched on for these tests.

    Nothing related to these issues changed in next versions... mostly cosmetic, and one extra diode to stop fan running off USB.


    :) usually not needed either! You can do this! It may just take a little tinkering. 99.9% of users need not even tinker, but environments, use cases and setups differ... Simplify (throw out the cameras, power cables, shopvacs etc and run it as shown in the video to build some trust)


    Shouldnt need that much grounding at all. Nor all the extra work. I'd strip all that out as it might be making it worse! (Setups not as tested in our labs...)

    With those: NC wiring would help: See gnea/grbl

    Or you could try ours: Xtension Limit Switch Kit - designed to work with the BlackBox
    Has pullups on the signal wire all the way to the switch... (;

    The ! comes on when the chip reaches overtemperature... The slightly different motors may be a cause, or environmental (higher ambient temp for example. Fan and heatsinks can only pull out so much heat.
    Your motors are 4.2A motors. Blackbox is 3.2A RMS. Not really perfectly suited motors...



    Long story short... I think you are overthinking EMI.
    Apart from the vacuum, there really doesnt seem to be huge sources of EMI in your setup.

    If you want to increase reliability a lot, disable the HARD Limits so you can move forward and get some work done :)
    Then on a weekend when you arent as frustrated, jump back in and eliminate some of the problems from above and see if any of that helps.

    Limits are great for a safety feature, but a little sanity checking before running the job can avoid crashes too.
    Homing isnt as useful as setting Zero either. Personally I find little use for switches... So its its a source of headache, remove it from the game
     
    Alex Chambers likes this.

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