584,829 active members*
5,330 visitors online*
Register for free
Login
IndustryArena Forum > WoodWorking Machines > DIY CNC Router Table Machines > Avid CNC > Machine accuracy issues - I need some guidance please
Page 1 of 3 123
Results 1 to 20 of 55
  1. #1
    Join Date
    Jul 2013
    Posts
    608

    Unhappy Machine accuracy issues - I need some guidance please

    My machine is running nema 34 motors with Gecko 203 drivers. Sold by CNC Router Parts.
    USB UC100 interface

    I have recently installed a precision collet .

    The machine cuts very well when it comes to larger items, but it seems to have a problem when it comes to smaller cuts.

    Take a look at images below:

    The first image shows a 0.5" x 0.5" rectangle cut close to the X / Y zero.



    This next image is about 10" from the rectangle above, notice how badly it cut the shape. I even tried to slow down the fed rate.



    Holes regardless of spiral ramping are not true, and it this case, the holes were drawn to be a specific distance from the edge, but they were cut at an offset.



    Here is a circle.. notice how the top is deformed.



    I am wondering if anyone has any idea on what could be wrong.
    I have been told in the past that my machine could be loosing steps?

    I have some questions:

    1) are the nema 34 not adequate for small detailing work?
    2) what causes circles to be deformed?

    3) how can I fix this? / if I am unable to make these cut, these machine is useless to me and I need to look else where.

    Thank you

  2. #2
    Join Date
    Mar 2003
    Posts
    35538
    Looks like you have some backlash, or something is loose somewhere.
    Gerry

    UCCNC 2017 Screenset
    http://www.thecncwoodworker.com/2017.html

    Mach3 2010 Screenset
    http://www.thecncwoodworker.com/2010.html

    JointCAM - CNC Dovetails & Box Joints
    http://www.g-forcecnc.com/jointcam.html

    (Note: The opinions expressed in this post are my own and are not necessarily those of CNCzone and its management)

  3. #3
    Join Date
    Apr 2010
    Posts
    82
    I don't think the size of the motors should have any effect. As for losing steps - certainly not at those feed rates I would think.

    Circles will be deformed if the steps/linear distance values are incorrect in the motor tuning in Mach3. What values are you using?

    I have not yet calibrated my system - I'm just using the values provided by CNCRP, but I do intend to do a calibration when I get done with playing and start using the machine more seriously.

  4. #4
    Join Date
    Jul 2013
    Posts
    608
    I looked over the machine and nothing is loose.
    I decided to plane the cutting surface just to see if that would have an effect on anything and while I was able to plan it, I noticed the the edge that corresponds to my X (grantry) do not cut a straight line.

    It start the line at one point, and it curves up and comes back down.
    My machine has the steel rails (the "pro linear kit") I wonder if the steel is cambered at the point where I am getting the error, but that would very hard to detect.

    The gantry is squared. I double checked by taking measurements.

    I really wish this machine worked right !! - its been nothing by challenges since I placed the order.
    Its a beautiful machine, but if there is no accuracy.... it really is a big problem.

    I am open to ideas. I think ultimately this could be why the shaped above are being cut correctly.






    I wanted to add, this pictures are at opposite ends over a 36" span

  5. #5
    Join Date
    Apr 2010
    Posts
    82
    Check that the steel rails are fully set into the vcon clamps all the way along the axis.

  6. #6
    Join Date
    Jul 2013
    Posts
    608
    I just took apart the entire gantry to see..
    I noticed that in the portion where the abnormality takes place, the carrier moves.
    The rails are all pressed into their v cons. I even took them out and flipped them. But the have such camber that is consistent w the fact is a little loose in the center.

    If I properly tighten the carrier at the center, then it gets tight for the carrier when it moves towards the ends.

    I wonder that if as long as it does not completely bind, if the nema 34 will still move w out any problems and not miss steps.

    I am also unable to get the res front carrier plate to be perpendicular to my table top. I guess that will be ab adjustment that I will have to compensate at the spindle.

  7. #7
    Join Date
    Apr 2010
    Posts
    82
    I looked back to your original thread to see what your machine looks like. I note that you have 4 v-con clamps across the gantry which is the long axis of your machine. I have 5 v-con clamps across my gantry which is half the size of yours. I may have been sent extra by mistake, but I used what was in the kit. I also have 5 v-cons along my long axis.

    When you tightened the vcons down, did you have clamps across the 2 rails is several places?

    Perhaps a different sequence of tightening the v-cons (with the carriage in place) would help? In that way you could perhaps get a consistent clearance.

  8. #8
    Join Date
    Feb 2014
    Posts
    13
    When I built mine I tried to make it so the resistance as you move the gantry across was fairly consistent. It's probably ok to have it feeling a little tight. Those Nema 34 motors are quite powerful. If you have it tighter near the ends and loose in the middle I would loosen up the screws in the middle and use a screwdriver to barely flex the rail out slightly and then tighten it back. Then test the gantry again. I spent over 2 hours doing this on my gantry till I was happy with the consistency of movement.

    I'm not sure why your X and Y axis are having issues but for Z axis problems you need to check a couple things if your using the ballscrew assembly. The file from cncrp might have had 4000 steps for the Z axis but it would need to be 10160 if you have the Pro ballscrew assembly. But if that is set wrong it should be pretty obvious because you will have to tell it to go 2.4" to get 1". The step numbers provided by cncrp are calculated values based on how they run the motors. If you are using different electronics they might need to be changed. But I would expect your X and Y steps to be the same value...mine are 2037.

    If you saw my other posts you can also have Z axis problems in the pro ballscrew assembly from the dust cover plate. Remove the thin metal plate in back and run your tests again. It won't fix X and Y but should help the Z axis positioning.

    I found out accidentally a couple times just how powerful those Nema 34 motors are. When I first started and was trying to figure out the homing I had a bit in the router that cleared the table just fine but didn't clear the vcon blocks on the side. When I tried to do a homing procedure it rammed the bit into the aluminum block while moving the gantry over. After I retracted everything and started testing things again it took a while for me to notice the wheels had skipped off the v rail on one side. So I disengaged the motor and realized I could not move the gantry by hand...even pulling it pretty darn hard. But the motor was moving it along just fine and I didn't even know it wasn't on track until I eyeballed it. The second thing I did was last night. I was using g code to try to surface my spoil board. I was converting code I found here from mm to inches and I messed up one number. The original homing move had Z-2 when I really needed Z2. I had a 1/2" milling bit in the router and when I hit go it rammed it through the 3/4" board at full speed.

  9. #9
    Join Date
    Jul 2013
    Posts
    608
    Gilrock,

    Thank you for your post. it is very useful to know that I am not the only know who spend 2 hours trying to adjust and properly setup the carrier. Last night spend over 3 hours taking it apart, tweaking the clamps and steel and putting it back together. In my opinion I this process should have been much easier and less time consuming. Let this be a heads up to those about to build - spend sometime on your gantry.

    I added some more v-con clamps. I now have 10 on the front. I am not sure it made things better, but after all the time I spend I got the carrier to be pretty consistent.

    I set it up so that it had a bit of resistance consistently, I can still move it by hand, but it stops fairly quickly is pushed. (this just the carries without any weight).
    One of the issues I was having is that carrier would move in plan and elevation (looking from the top and front) in ways that I shouldn't - I am going to attribute this to the bad cuts.

    today I was able to make some cuts: a few lines in X and Y and a final test part of what I cut this weekend.
    The X and Y line seem straight(er) I really need to resurface my table to erase the bad lines and really see.

    CNCRP has reached out to me via e-mail after seeing this thread. I hope they are able to help me further.

    regardless Kjk2002, thank you for pointing out the v-cons. I think everything helped.
    The machine is cutting much better (apparently)

    I need to resurface my top quite a bit since its not flat, or use a thinker pieces of MDF. All I had this weekend was 1/2"
    maybe I need to do 1" thick?

    Here is a pictures of todays trial, looks better already:




    Feed rate: 20 IPM
    Plunge Rate: 8 IPM
    Clearing method : offset

  10. #10
    Join Date
    Apr 2010
    Posts
    82
    Looks like you have your problems on the run.

  11. #11
    Join Date
    Jul 2013
    Posts
    608
    What do you mean by that ?

  12. #12
    Join Date
    Apr 2010
    Posts
    82
    You are close to a solution to the accuracy issues.

  13. #13
    Join Date
    Jul 2013
    Posts
    608
    Quote Originally Posted by kjk2002 View Post
    You are close to a solution to the accuracy issues.
    haha, yes. thankfully - I hope

  14. #14
    Join Date
    Jul 2013
    Posts
    608
    I was exchanging e-mails today with a friend, and he told me this:

    I slow my feed rate for small cutting tools, not small cuts.
    He also tells me that he cuts at 250 ipms and that I should use exact stop mode...
    http://www.machsupport.com/wp-conten...ettings_v2.pdf

    Exact Stop – This mode accelerates and decelerates to each “point” in the gcode. Mach-3
    only sees one move at a time and usually machines run somewhat rough and very slowly
    in this mode. Exact stop should only be used where a machine must not round any
    corners (inside or outside). However, remember that most CAM software will output
    many tiny G01 moves to form arcs. In exact stop mode this type of movement will leave
    very bad surface finishes and can be hard on tooling and machine components.
    This is all in response to my problem here.

    I run my 1/8" bit at depth of 1/16" at a time to make the pockets but my and also slow down the feed for the smalls cuts.
    if I cut longer straight runs, I will do 60 ipm with the same 1/8" but with 1/8" depth steps.

    Can I get some input from you guys?

  15. #15
    Join Date
    Jul 2013
    Posts
    608

    Question um.. I need some second thought here

    one of things I did originally before taking anything apart was to try to use mach 3 to calibrate the motor steps.
    this gave me a number very close to what I found on the CNCRP XML but not quite the same numbers they provided. With the setting I have I was able to make the cut I showed today.

    The other question that I have related to feedrate,
    I am lowering my feedrate to 20 ipm to cut with 1/8" bit as shown, but I am being told that I should not need to lower my feedrate to make small cuts.
    I also cut at 60 ipm this is all with the same 1/8" bit.

    I am being told I can run the machine at 500 ipm and not have to slow it down to make small cuts. What are your thoughts on this?
    Additionally, It was recommended I switch to EXACT STOP mode. However from the mach 3 description below it seems like that mode makes machine run erratic and would produce worse results

    Exact Stop – This mode accelerates and decelerates to each “point” in the gcode. Mach-3
    only sees one move at a time and usually machines run somewhat rough and very slowly
    in this mode. Exact stop should only be used where a machine must not round any
    corners (inside or outside). However, remember that most CAM software will output
    many tiny G01 moves to form arcs. In exact stop mode this type of movement will leave
    very bad surface finishes and can be hard on tooling and machine components.
    I did find the look ahead setting, mine is set to 20. I am thinking of going to 100 - would that make things run any better?

  16. #16
    Join Date
    Mar 2003
    Posts
    35538
    You really don't want to cut in Exact Stop mode, as it can give very jerky motion. My recommendation would be to set lookahead to 200, CV on, Stop CV on Angles to 89, and all other CV options off, in both General Config and the Settings page.
    Be aware that CV mode can cause rounded corners at high speed, which can be worse with slower acceleration settings. But Stop CV on angles at 89 should get you square 90° corners.
    I really don't know of anyone using Exact Stop mode on a router with Mach3.

    Feedrate depends a bit on rpm, and at slower speeds, you should be lowering your rpm. I'd try running between 200ipm @ 15,000 to 300ipm @ 18,000 rpm. Depth of cut at 1/8". You can plunge at 50 ipm or more, but ideally, you want to ramp into the cut. You can ramp in at full cut speed, and it's much better on the bit. Machine rigidity comes into play at higher feedrates, and can affect the cut quality. If you're not getting an acceptable quality of cut, then slow down a bit. Or increase rpm.
    Gerry

    UCCNC 2017 Screenset
    http://www.thecncwoodworker.com/2017.html

    Mach3 2010 Screenset
    http://www.thecncwoodworker.com/2010.html

    JointCAM - CNC Dovetails & Box Joints
    http://www.g-forcecnc.com/jointcam.html

    (Note: The opinions expressed in this post are my own and are not necessarily those of CNCzone and its management)

  17. #17
    Join Date
    Jul 2013
    Posts
    608
    Gerry, so at 40 ipm, or even 80 ipm for straight cuts at 20,000 rpm I am running my machine very very slow.
    Thank you

  18. #18
    Join Date
    Mar 2003
    Posts
    35538
    Yes. I almost never run slower than 100-125, and when I do, it's mainly because my machine isn't very rigid.
    Gerry

    UCCNC 2017 Screenset
    http://www.thecncwoodworker.com/2017.html

    Mach3 2010 Screenset
    http://www.thecncwoodworker.com/2010.html

    JointCAM - CNC Dovetails & Box Joints
    http://www.g-forcecnc.com/jointcam.html

    (Note: The opinions expressed in this post are my own and are not necessarily those of CNCzone and its management)

  19. #19
    Join Date
    Jul 2013
    Posts
    608
    thank you

  20. #20
    Join Date
    Sep 2007
    Posts
    108
    I build guitars, I use exact stop mode on some of my finish passes on pockets and such. But I do cut in CV mode for almost everything else.

    Sent from my SCH-R720 using Tapatalk 2

Page 1 of 3 123

Similar Threads

  1. accuracy issues
    By lloydinz in forum DIY CNC Router Table Machines
    Replies: 2
    Last Post: 05-24-2013, 07:06 PM
  2. Accuracy issues
    By runninwild400 in forum Torchmate
    Replies: 7
    Last Post: 03-19-2009, 11:27 PM
  3. x2 accuracy issues
    By BobMorside in forum Syil Products
    Replies: 12
    Last Post: 01-03-2009, 02:29 AM
  4. new setup, accuracy issues
    By DogWood in forum Benchtop Machines
    Replies: 4
    Last Post: 06-20-2008, 04:05 AM
  5. Accuracy issues
    By itsme in forum Benchtop Machines
    Replies: 16
    Last Post: 10-29-2006, 08:17 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
  •