584,826 active members*
5,170 visitors online*
Register for free
Login
Page 1 of 2 12
Results 1 to 20 of 38
  1. #1
    Join Date
    Jun 2018
    Posts
    84

    Horrible Surface Finish

    Good Day to All! Looking for some help from the resident wizards on how to improve surface finish on circles, arcs and angular cuts. First some background. My vertical mill is a PM-932 converted to CNC about five years ago. The mechanicals are 1600 oz in steppers on X and Y and a 4200 oz in stepper on Z. Ball screws on X and Y are 1605 and the Z is a 2005. All axis have double ball nuts. Gibs are adjusted properly and backlash is no more than .001" inch. I checked cutter runout and it is less than .0008". Spindle bearing preload is set and deflection checked with a DTI. Electronics consist of PMDX-126 breakout board and ESS. Spindle is controlled by a PMDX-107 control board. Control software is Mach3 version 062. I use CamBam and Fusion360 for my gcode generation both with the same surface finish results. X and Y drivers are set to 3200 steps. X and Y velocity is set to 120 and Acceleration is set to 30. Both very conservative settings. Step and Dir pulse is set to 5. Let me know if any further info is needed.

    Now onto the surface finish issue at hand. My pictures aren't the greatest but they do show the how bad the surface finish is. The first picture is a screen shot of my stock and part orientation. I'm showing this because machining in X or Y produces a very smooth finish. When combining X and Y for circles, arcs or angular cuts the surface finish is terrible. The second picture is illustrates the smooth surface finish in X and the third picture shows the surface finish on an angular cut. The faceting/scallops are about .030" apart. This cut was made with a 1/4" 4 flute end mill running at 4000 rpm and 20 IPM in 1018 mild steel at .188" DOC. Changing the feedrate and DOC had no effect.

    I have played with some of the Mach3 configuration settings and CamBam settings but nothing has changed related to the surface finish. Any insight as to what is causing this and pointers on how to correct it is appreciated.

  2. #2
    Join Date
    Dec 2013
    Posts
    5717

    Re: Horrible Surface Finish

    Can you post your .cb file. Add a .txt extension to be able to attach it to a post.
    Jim Dawson
    Sandy, Oregon, USA

  3. #3
    Join Date
    Dec 2005
    Posts
    436

    Re: Horrible Surface Finish

    From the info you provided it seems that you have identified he two most probable sources, machine loosesness or software.

    I would look into the software setting for how small the linear segments are to interpolate an arc. If the linear segments are too large you might get arcs that are not smooth. That would mean your software that generates the g code, does the g code have line segments that correspond to the surface finish?
    You mention arcs have bad surface finish, what about straight segments that are not parallel to the axis of the machine?
    And on circles, is the surface finish uniformly bad around the circumference or is there a correlation between changes in finish and the axis of the machine? ( ie possibly good when movement parralel to X and Y )

  4. #4

    Re: Horrible Surface Finish

    It could be a Mach3 motion setting. Maybe post a screen shot of those settings.

  5. #5
    Join Date
    Jan 2018
    Posts
    1516

    Re: Horrible Surface Finish

    The 3rd one with all the lines all over it looks like what my little X2 machine now produces.
    I've rebuilt the whole thing to no avail.
    I've narrowed mine down (I think) to the head gib slideways. Been like it since I badly crashed it and it looks like my head spread thus resulting in massive head vibrations. Looks, set and feels tight with no vibrations from table or column, just the head.

  6. #6
    Join Date
    Jun 2018
    Posts
    84

    Re: Horrible Surface Finish

    Quote Originally Posted by Jim Dawson View Post
    Can you post your .cb file. Add a .txt extension to be able to attach it to a post.
    Jim - here is my CamBam file.
    Attached Files Attached Files

  7. #7
    Join Date
    Jun 2018
    Posts
    84

    Re: Horrible Surface Finish

    Quote Originally Posted by cncuser1 View Post
    From the info you provided it seems that you have identified he two most probable sources, machine loosesness or software.

    I would look into the software setting for how small the linear segments are to interpolate an arc. If the linear segments are too large you might get arcs that are not smooth. That would mean your software that generates the g code, does the g code have line segments that correspond to the surface finish?
    You mention arcs have bad surface finish, what about straight segments that are not parallel to the axis of the machine?
    And on circles, is the surface finish uniformly bad around the circumference or is there a correlation between changes in finish and the axis of the machine? ( ie possibly good when movement parralel to X and Y )
    Thanks for responding. I looked through the CamBam configuration settings and found Arc Fit Tolerance which is set to .001" and Auto Arc Fitting which is active. In the Mach3 post processor Arc Center Mode is set to Incremental CP-1 (options are Default, Absolute, Incremental P1-C, C-P2, and P2-C), Arc Output is set to Normal (options are Convert to Lines and Helix Convert to Lines), and Arc to Lines Tolerance is .01". I have listed all of the settings related to arcs.

    Straight segment cuts not parallel with X or Y are represented in picture 3. The surface finish is uniformly bad on circles and arcs.

    The gibs are a tight as I can get them without losing steps. I use the "Bridgeport" method where for X you center the table on the saddle and mount an indicator on the saddle and place the DI on one end of the table. Pushing and pulling on the table the goal is to get no more than .001" of movement after letting go of the table. The same method is used for Y except the DI is mounted on the mill base. I've done this and the best I got was about .0015". Not too bad for a Asian machine.

  8. #8
    Join Date
    Jun 2018
    Posts
    84

    Re: Horrible Surface Finish

    Quote Originally Posted by CL_MotoTech View Post
    It could be a Mach3 motion setting. Maybe post a screen shot of those settings.
    I will do that in the morning when I get in the shop. Specifically what settings are you interested in seeing. I'd like to get you everything in one shot.

    Thanks

  9. #9
    Join Date
    Jun 2018
    Posts
    84

    Re: Horrible Surface Finish

    Quote Originally Posted by dazp1976 View Post
    The 3rd one with all the lines all over it looks like what my little X2 machine now produces.
    I've rebuilt the whole thing to no avail.
    I've narrowed mine down (I think) to the head gib slideways. Been like it since I badly crashed it and it looks like my head spread thus resulting in massive head vibrations. Looks, set and feels tight with no vibrations from table or column, just the head.
    I'd like to say I never crashed my machine but that wouldn't be the truth. The mill is quiet and doesn't vibrate and straight line cuts are absolutely smooth. I'm thinking it's software related.

  10. #10
    Join Date
    Dec 2013
    Posts
    5717

    Re: Horrible Surface Finish

    I don't see a problem with the CB file, all the settings and tool paths look normal. What you might try is edit the post processor, set Arc Output to Convert to Lines, and set Arc to Lines Tolerance to 0.001. I creates a huge G code file, but it might smooth things out. If that fails, then it is either a mechanical problem or a setting in Mach3. I can't help much with Mach3, I don't know much about the settings. With that much scollop, you should be able to feel the machine jerking as it cuts.
    Jim Dawson
    Sandy, Oregon, USA

  11. #11
    Join Date
    Jun 2018
    Posts
    84

    Re: Horrible Surface Finish

    Quote Originally Posted by Jim Dawson View Post
    I don't see a problem with the CB file, all the settings and tool paths look normal. What you might try is edit the post processor, set Arc Output to Convert to Lines, and set Arc to Lines Tolerance to 0.001. I creates a huge G code file, but it might smooth things out. If that fails, then it is either a mechanical problem or a setting in Mach3. I can't help much with Mach3, I don't know much about the settings. With that much scollop, you should be able to feel the machine jerking as it cuts.
    Thanks Jim. I'll try your suggestions tomorrow.

  12. #12
    Join Date
    Dec 2005
    Posts
    436

    Re: Horrible Surface Finish

    "The surface finish is uniformly bad on circles and arcs." if this means that the phenomena is the same around the whole 360 degrees of a circle then to my mind this is a software issue.

    I suggest three things
    1- use a simple example, to simplify debugging: A circle.
    2-Machine a simple circle and post the gcode and picture in this thread.
    3- Try a whole new software to generate the gcode, see if that recreates the problem, this will eliminate or focus the source of the error.

  13. #13

    Re: Horrible Surface Finish

    Have you verified the ballscrew is running true . Also , have you tried to conventional mill your finish pass

  14. #14
    Join Date
    Jun 2018
    Posts
    84

    Re: Horrible Surface Finish

    I ran several diagnostic tests this morning using CamBam and Fusion360 to generate gcode.

    For the first test I adjusted Arc Output to Convert to Lines and set Arc to Lines Tolerance to 0.001. The first two pictures are the results using CamBam code. The X and Y axis straight cuts are pristine. The angle cuts are OK but not as good as the straight line cuts. The radius cuts are worse showing more severe scalloping than the angle cuts.

    The second test used F360 code. Surprisingly the results were not as good as CamBam. See pictures 3 and 4. The straight cuts were very close to the CamBam results while the angle and radius surfaces were somewhat worse. For info I ran two tests using F360. The first test Radius Arcs was checked as "No" while the second test it was checked as "Yes". There was no discernible difference in the surface finish between the two tests.

    For info the tests were run using a new 3/8" 4-flute carbide end mill running at 450 sfm and .0018" CLPT. DOC was .500".

    I still believe it's a software issue but not sure where to go from here.

  15. #15
    Join Date
    Dec 2013
    Posts
    5717

    Re: Horrible Surface Finish

    I think we need a bit more information about your machine. Model? Size? I assume you are using stepper motors? If so, maybe adjusting the micro-stepping a bit finer would help.
    Jim Dawson
    Sandy, Oregon, USA

  16. #16
    Join Date
    Dec 2005
    Posts
    436

    Re: Horrible Surface Finish

    Can you redo your test with a shallower depth of cut? keep EVERYTHING elese the same?

  17. #17
    Join Date
    Jun 2018
    Posts
    84

    Re: Horrible Surface Finish

    Quote Originally Posted by metalmayhem View Post
    Have you verified the ballscrew is running true . Also , have you tried to conventional mill your finish pass
    I have not done a runout check on the ball screws nor tried conventional milling. Another test for tomorrow!

  18. #18
    Join Date
    Jun 2018
    Posts
    84

    Re: Horrible Surface Finish

    Quote Originally Posted by Jim Dawson View Post
    I think we need a bit more information about your machine. Model? Size? I assume you are using stepper motors? If so, maybe adjusting the micro-stepping a bit finer would help.
    My mill is a PM-932 conversion. It's a medium size benchtop mill weighing about 850 lbs. Table is 9 x 32". Yes, steppers on X and Y are 1600 oz in and Z is 4200 oz in. Current microstepping is set to 3200. I can set it to 6400 and see if that has an effect.

    - - - Updated - - -

    Quote Originally Posted by cncuser1 View Post
    Can you redo your test with a shallower depth of cut? keep EVERYTHING elese the same?
    Certainly can. I'll test it in the morning.

  19. #19
    Join Date
    Jan 2005
    Posts
    1943

    Re: Horrible Surface Finish

    I would just hand write a G-code with a G3 to cut a circle and see what it looks like. If the faceting is still there then you know it is either the machine or the controller. and not a CAM problem.

  20. #20
    Join Date
    May 2008
    Posts
    1185

    Re: Horrible Surface Finish

    I'm thinking bent ball screw or the end work is not centered.

    I know it sounds strange but loosen the bolts on the bearing support side of a screw and see if it gets better.

    It could be either end.

    I had a problem like that a few years ago. The end work was off center and the hole table would rock around and make waves in the cut. It got worse as you moved to the area with the bad end work.

    I did a video of it.

    https://www.youtube.com/watch?v=d3Gg7qDeFCY
    youtube videos of the G0704 under the name arizonavideo99

Page 1 of 2 12

Similar Threads

  1. Boring 6061 - horrible surface finish!
    By Kreftmoto in forum Material Machining Solutions
    Replies: 5
    Last Post: 11-04-2015, 08:06 PM
  2. Horrible surface finish from endmill
    By skray775 in forum MetalWork Discussion
    Replies: 11
    Last Post: 11-03-2012, 05:42 AM
  3. Surface Finish
    By pgf545 in forum PTC Pro/Manufacture
    Replies: 14
    Last Post: 01-24-2012, 04:03 PM
  4. Surface Finish
    By dlange in forum MetalWork Discussion
    Replies: 6
    Last Post: 09-21-2010, 07:02 PM
  5. Surface finish
    By d.a.v.e in forum Mechanical Calculations/Engineering Design
    Replies: 1
    Last Post: 11-10-2006, 08:35 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
  •