Calibration question


Page 1 of 2 12 LastLast
Results 1 to 20 of 30

Thread: Calibration question

  1. #1
    Community Moderator difalkner's Avatar
    Join Date
    Nov 2014
    Location
    United States
    Posts
    730
    Downloads
    0
    Uploads
    0

    Default Calibration question

    My calibration still isn’t as close as I know it can be so I created a file today for some testing and calibration. This is a busy piece but served its purpose. Here’s what I cut with a 1/4" two-flute downcut spiral bit running 18k rpm and 125 ipm. The depth of cut is 0.1875" and it is cut with one pass leaving 0.005" on the side walls for a final clean up cut to get the fuzz off. The depth needed to be just deep enough to allow my dial calipers in for good measurement.

    Calibration question-calibration-test-8-11-18-jpg

    I started on the bottom left, ‘Before’, and made my first cut. You can see it was out +0.005” on Y and -0.010” on X. I changed the settings in Mach4 and then ran the top left. Ignore the diagonals, those are to ensure I didn’t get too close to the screws. The circled dimensions are exactly as they’re supposed to be. I then ran the top right and it’s exact on Y and out +0.005” on X. The middle right was next and you can see it is exact on Y but out -0.002” on X. Bottom right is exact on X and +0.005” on Y (I wrote it wrong on the piece). The circle is very close, as well.

    However, all of the inside circles are 0.015” to 0.025” undersize and I don’t get that. The outside squares, rectangles, and circles are close enough to be acceptable. I think the variance is due to the material which is a sort of soft Purebond plywood from Home Depot. If I was cutting hardwood I think the tolerances would be much closer.

    So how do I get the inside circles to cut accurately? This makes it very difficult when I’m doing inlays and inserts, fitting dowels or bolts in holes, etc. I can’t change settings now or the other measurements will be off and I don’t want that. Is it something not set correctly in Fusion 360? Something in Mach4? It's not mechanical; the machine is a Saturn 2x4 and is very rigid and tight and if it was mechanical the squares, rectangles, and outside circles would be off, as well.

    Thanks!

    David

    Similar Threads:
    David
    Romans 3:23
    CurlyWoodShop - www.etsy.com/shop/CurlyWoodShop
    David Falkner - www.youtube.com/user/difalkner
    difalkner - www.instagram.com/difalkner


  2. #2
    Member ger21's Avatar
    Join Date
    Mar 2003
    Location
    Shelby Township
    Posts
    35538
    Downloads
    1
    Uploads
    0

    Default Re: Calibration question

    So how do I get the inside circles to cut accurately?
    Try cutting them at 30 ipm, and see what size they are.


    It's not mechanical; the machine is a Saturn 2x4 and is very rigid and tight
    Don't be so sure. Ultimately, your rigidity depends on a spring holding the pinion into the rack. How rigid is a spring?

    Gerry

    UCCNC 2017 Screenset
    [URL]http://www.thecncwoodworker.com/2017.html[/URL]

    Mach3 2010 Screenset
    [URL]http://www.thecncwoodworker.com/2010.html[/URL]

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

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


  3. #3
    Community Moderator difalkner's Avatar
    Join Date
    Nov 2014
    Location
    United States
    Posts
    730
    Downloads
    0
    Uploads
    0

    Default Re: Calibration question

    I can cut it that slow, assuming it won't burn. Actually, I could do the roughing pass at 125 ipm and the finish 0.005" at 30 ipm and that should do it, I would think.

    The springs are pretty tight but if it was the spring it would have to affect the squares, rectangles, and outside circles and those are pretty close. Whatever the issue it's only showing on inside circles and bores.

    Thanks, Gerry.

    David

    David
    Romans 3:23
    CurlyWoodShop - www.etsy.com/shop/CurlyWoodShop
    David Falkner - www.youtube.com/user/difalkner
    difalkner - www.instagram.com/difalkner


  4. #4
    Member
    Join Date
    Nov 2017
    Location
    United States
    Posts
    35
    Downloads
    0
    Uploads
    0

    Default Re: Calibration question

    The .005" you posted as stock left. Was that in the stock to leave section in the can profile? If it was not there but in the section up top for finishing passes, then you might have stock to leave checked, which is fudging the numbers. I know stock in my setup of fusion 360, it is .1mm, which is roughly .004". And since that is radially, double that number and you get .008". Add .005" to that and you get .013, close enough to the .015" to look into.

    I could be completely off. And maybe look at the circles in model view to see if the circle is actually setup right, making sure decimal places out in settings is set to 3 0's past the decimal.



  5. #5
    Community Moderator difalkner's Avatar
    Join Date
    Nov 2014
    Location
    United States
    Posts
    730
    Downloads
    0
    Uploads
    0

    Default Re: Calibration question

    Quote Originally Posted by firepong View Post
    The .005" you posted as stock left. Was that in the stock to leave section in the can profile? If it was not there but in the section up top for finishing passes, then you might have stock to leave checked, which is fudging the numbers. I know stock in my setup of fusion 360, it is .1mm, which is roughly .004". And since that is radially, double that number and you get .008". Add .005" to that and you get .013, close enough to the .015" to look into.

    I could be completely off. And maybe look at the circles in model view to see if the circle is actually setup right, making sure decimal places out in settings is set to 3 0's past the decimal.
    The 0.005" was for the rough cut. For the finish cut I'm cutting full depth with no 'Stock to leave'. Yesterday I invested some time in recalibrating all axes. Z and X were spot on but for some reason Y was out 1/16" in 44". So I corrected that and ran the files again with conventional cuts for final and got everything far, far closer. The inside cuts are still out 0.007" to 0.009" but that's close enough for me to do a negative value on 'Stock to leave'. I didn't want to compensate 0.025", though.

    I haven't been taking 0's out to three decimal places, figuring that 0.75" is the same as 0.75000..." so I have to ask - is it, as far as Fusion 360 and Mach4 is concerned?

    David

    David
    Romans 3:23
    CurlyWoodShop - www.etsy.com/shop/CurlyWoodShop
    David Falkner - www.youtube.com/user/difalkner
    difalkner - www.instagram.com/difalkner


  6. #6
    Member ger21's Avatar
    Join Date
    Mar 2003
    Location
    Shelby Township
    Posts
    35538
    Downloads
    1
    Uploads
    0

    Default Re: Calibration question

    Yes, .75 = .75000

    Gerry

    UCCNC 2017 Screenset
    [URL]http://www.thecncwoodworker.com/2017.html[/URL]

    Mach3 2010 Screenset
    [URL]http://www.thecncwoodworker.com/2010.html[/URL]

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

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


  7. #7
    Community Moderator difalkner's Avatar
    Join Date
    Nov 2014
    Location
    United States
    Posts
    730
    Downloads
    0
    Uploads
    0

    Default Re: Calibration question

    Here's an update that I couldn't post with the forum being down for a few days -

    Using a 60° V-bit into the spoilboard I engraved a rectangle - 23" x 46" - so I could measure diagonals. They are a tiny bit under 1/16" off on the diagonals.

    Fusion 360 says the diagonal should be 51.43" and I'm measuring 51 7/16" (51.4375") - close enough to 7/16". All I have to measure that distance is a tape measure so this is going to have to be close enough. The other diagonal is measuring 51 1/2" or a tiny bit under but very slight. So for all practical purposes I can call it 1/16" greater distance.

    So here's my question - is this enough to correct (it is to me, btw), is it best fixed by mechanical means, and/or if it's fixable in software how do I accomplish that? I rarely cut anything that large but if I can get it closer in a fashion that doesn't require major mods then I'd like to get it closer.

    I'm running Mach4 with ESS, proximity sensors on Y are on the X- side only but I do have two steppers on Y.

    This graphic may help -
    Calibration question-calibration-rectangle-8-18-18-jpg

    David

    PS - I did cut the 18" tall stripes plaque and the inserts fit perfectly so we've made a great deal of progress with y'all helping - thank you! It's not yet sanded or finished but they are glued in place.
    Calibration question-038-sfc-stripe-yet-sanded-jpg

    David
    Romans 3:23
    CurlyWoodShop - www.etsy.com/shop/CurlyWoodShop
    David Falkner - www.youtube.com/user/difalkner
    difalkner - www.instagram.com/difalkner


  8. #8
    Member ger21's Avatar
    Join Date
    Mar 2003
    Location
    Shelby Township
    Posts
    35538
    Downloads
    1
    Uploads
    0

    Default Re: Calibration question

    Best fix would be mechanical.

    You might also be able to correct it by moving one of the home switches on the gantry, so it homes itself into square. The ESS may also have a setting to do that in software.

    Of course, if the gantry is too stiff to move the small amount required, neither may work.

    Gerry

    UCCNC 2017 Screenset
    [URL]http://www.thecncwoodworker.com/2017.html[/URL]

    Mach3 2010 Screenset
    [URL]http://www.thecncwoodworker.com/2010.html[/URL]

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

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


  9. #9
    Member
    Join Date
    Sep 2005
    Location
    United States
    Posts
    1740
    Downloads
    0
    Uploads
    0

    Default Re: Calibration question

    My Saturn 2, 4x4 has two motors and two switches on the Y axis. so in theory it would be self correcting.

    1000x750 Workbee CNC - Mach4 - PMDX USB - Windows 10 Pro


  10. #10
    Member ger21's Avatar
    Join Date
    Mar 2003
    Location
    Shelby Township
    Posts
    35538
    Downloads
    1
    Uploads
    0

    Default Re: Calibration question

    Yes, but the correction relies on the location of the switches. They'd likely need some adjusting to get a perfectly square gantry.

    Gerry

    UCCNC 2017 Screenset
    [URL]http://www.thecncwoodworker.com/2017.html[/URL]

    Mach3 2010 Screenset
    [URL]http://www.thecncwoodworker.com/2010.html[/URL]

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

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


  11. #11
    Member
    Join Date
    Sep 2005
    Location
    United States
    Posts
    1740
    Downloads
    0
    Uploads
    0

    Default Re: Calibration question

    Quote Originally Posted by ger21 View Post
    Yes, but the correction relies on the location of the switches. They'd likely need some adjusting to get a perfectly square gantry.
    That's the issue Gerry they are good grade of proximity switches and from what I can tell from measuring with a digital caliper they are very close on the same trigger point. My time has been taken up by a painting project and still waiting on my "special order" 5x8 spoil board.

    1000x750 Workbee CNC - Mach4 - PMDX USB - Windows 10 Pro


  12. #12
    Community Moderator difalkner's Avatar
    Join Date
    Nov 2014
    Location
    United States
    Posts
    730
    Downloads
    0
    Uploads
    0

    Default Re: Calibration question

    I have two sensors on Y but one is on Y+ and one on Y-, both on the X- or left side. When I was building it Nate said I didn't need two sensors on Y-, that doing the way I have it is acceptable.

    David

    David
    Romans 3:23
    CurlyWoodShop - www.etsy.com/shop/CurlyWoodShop
    David Falkner - www.youtube.com/user/difalkner
    difalkner - www.instagram.com/difalkner


  13. #13
    Member ger21's Avatar
    Join Date
    Mar 2003
    Location
    Shelby Township
    Posts
    35538
    Downloads
    1
    Uploads
    0

    Default Re: Calibration question

    that doing the way I have it is acceptable.
    Assuming the gantry is already perpendicular to the other axis....

    Gerry

    UCCNC 2017 Screenset
    [URL]http://www.thecncwoodworker.com/2017.html[/URL]

    Mach3 2010 Screenset
    [URL]http://www.thecncwoodworker.com/2010.html[/URL]

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

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


  14. #14
    Member
    Join Date
    Sep 2005
    Location
    United States
    Posts
    1740
    Downloads
    0
    Uploads
    0

    Default Re: Calibration question

    Quote Originally Posted by ger21 View Post
    Assuming the gantry is already perpendicular to the other axis....
    It must be a pretty common thing if Mach3 has it as a feature. Both ends of my over 4 ft gantry have motors, and sensors. Y and A or Slave. Last night I noticed that my A or Slave side belt was loose and not aligned correctly. So I going to remove today and check the pulley setscrews and do what is needed. Then I will check out the auto squaring function.

    1000x750 Workbee CNC - Mach4 - PMDX USB - Windows 10 Pro


  15. #15
    Member
    Join Date
    Oct 2016
    Location
    United States
    Posts
    236
    Downloads
    0
    Uploads
    0

    Default Re: Calibration question

    Quote Originally Posted by ger21 View Post
    Best fix would be mechanical.

    You might also be able to correct it by moving one of the home switches on the gantry, so it homes itself into square. The ESS may also have a setting to do that in software.

    Of course, if the gantry is too stiff to move the small amount required, neither may work.
    I just mentioned this on another thread, but my Saturn 4x2 is too stiff to home itself into square. I think the best bet is to loosen up the gantry bolts. It requires some iteration. I use dowel pins in holes and measure the diagonals over the pins using a framing square in combination with my calipers. See https://www.cnczone.com/forums/finel...ml#post2209222 (Post #46) Any other long fixed-length material with one long straight edge and an internal corner would work though.

    I think I managed to get my diagonals to about 0.01" measured on about a 16" square after about 4-5 iterations.

    I'm not certain this is the cause of your original inaccuracies, though. For what it's worth, I have been able to interpolate 1.25" holes in aluminum to about 0.003" with a long 1/8" bit, but they came out oval across a diagonal. I still need to figure out why that is happening, but the general consensus is flex somewhere in the system. The bit was probably too small, long and flexible to begin with, but I think there is something else going on, too.

    -Robert



  16. #16
    Community Moderator difalkner's Avatar
    Join Date
    Nov 2014
    Location
    United States
    Posts
    730
    Downloads
    0
    Uploads
    0

    Default Re: Calibration question

    I'm in the process of upgrading the drive spindles to the newer style and then I'll check mine again. Your diagonal is a little over 22" long and my diagonal was 52 7/16" but I'm eager to run it again and see what I've got when this upgrade is completed.

    David

    David
    Romans 3:23
    CurlyWoodShop - www.etsy.com/shop/CurlyWoodShop
    David Falkner - www.youtube.com/user/difalkner
    difalkner - www.instagram.com/difalkner


  17. #17
    Member
    Join Date
    Apr 2010
    Location
    USA
    Posts
    109
    Downloads
    0
    Uploads
    0

    Default Re: Calibration question

    Hi David,
    The new drive spindles helped me out a lot. I haven't had a chance to really go to town on them, but the lateral movement in the originals was causing me all sorts of issues. Nate sent me new belts along with the new spindles also.
    Erik



  18. #18
    Community Moderator difalkner's Avatar
    Join Date
    Nov 2014
    Location
    United States
    Posts
    730
    Downloads
    0
    Uploads
    0

    Default Re: Calibration question

    I got the new spindles on today and could immediately hear the rack & pinion meshing. I never heard that before. It was likely because the old pinions were softer and they wore to fit the gear racks but the new pinions are hardened. So that obviously tells me something is out of alignment. I loosened the necessary bolts and the gantry immediately shifted forward on the right side (X+) . That tells me it has been in tension this whole time. Anyway, I got with Nate on his procedure to square the gantry and am in the middle of doing that now. I'm hopeful this will also correct my calibration square being long on that side. We'll see...

    David

    David
    Romans 3:23
    CurlyWoodShop - www.etsy.com/shop/CurlyWoodShop
    David Falkner - www.youtube.com/user/difalkner
    difalkner - www.instagram.com/difalkner


  19. #19
    Member
    Join Date
    Apr 2016
    Location
    United States
    Posts
    841
    Downloads
    0
    Uploads
    0

    Default Re: Calibration question

    Hello David,


    I am thinking about upgrading from my 80/20 4X4 (NEMA 34s, 3” x 6” gantry, 2.2kw water cooled spindle) to a Saturn 2. I sent Nate an email with some questions a couple of days ago, including asking about the method he uses to square the gantry. Unfortunately, he hasn't replied yet. Any chance you could share the technique he passed on to you?


    We haven't met previously, but I've followed your build and projects. I have thoroughly enjoyed your videos. Like you, I get real fussy about dialing in my equipment. I have to admit that I get obsessive about it.


    I managed to get the gantry on my current machine to within about .002 over about 58". I couldn't get any closer to spot on, because of flex and the like inherent in a bolt together machine. To get there took some time, but the method I used is pretty much guaranteed to get you as close as humanly possible. It’s called the 5-Cut Method. It was designed to make precisely square fences on table saw crosscut sleds. I've also used it to get the blade on my chop saw within less than .001" of perpendicular to the saw's fence over its 12” cut length. Attached is a link to the definitive video on the technique: . Needless to say, to make use of it on a CNC takes some minor adaptation. Here's how I did it:


    1. Mount a fence along, and parallel to, the Y axis. I put mine on the left (X--) side. I put a 1/4" steel dowel in the spindle's chuck and ran it to X4 Y10. I pushed a 48" piece of 80/20 extrusion up against the left side of the pin. On the left side the extrusion, I fastened a pointed piece of aluminum up against the extrusion, trapping the extrusion between the pin and pointed piece. This allows the extrusion to pivot to either the right or left. Pivot the extrusion a bit to the left (toward X--) to clear a path to the next step.


    2. Using the MDI in the control software (I use Mach4) enter g90 x4 y40 (or whatever suits your Y machine length). This moves the spindle down the spoilboard in a straight line an additional 30 inches. Note: I have my Mach4 set up in inches, not mm. Rotate the extrusion up against the dowel in the spindle and fix the extrusion in place fore and aft. Clamps should do the trick. Note that I routed a rabbit down the left side of my spoilboard, and have T-track mounted in it. All I have to do is mount blocks up against the edge of rabbit and I'm good to go. The method I used with the extrusion was the one I used before I found myself working way too hard to get stock located properly on the table.


    3. The fence described above serves a temporary fixed reference point for cutting. Use MDF for your test cutting material. I repeat. Use MDF for your test cutting material. Do not use plywood or anything else. MDF cuts more cleanly than about anything else out there even remotely near MDF’s price point.


    4. Make the cuts described in the video along the X axis. The MDF needs to be held tightly against the fence. Once you make the cut, you rotate the stock clockwise, so the freshly cut edge is now against the fence. Make the cut, rotate clockwise, make the cut, etc. Be sure the cut leaves material on the offcut side. This allows some support on both sides of the cut, and minimizes side to side bit deflection. I was able to use hold downs to keep the offcut captive, but adding tabs should work as well. I avoided them, because cutting them loose just added delay in an already tedious process.


    5. I was able to adjust my gantry within around .030" in about 58 or so inches before I was unable to adjust it any further. From that point, I used my homing sensors to force it the rest of the way square. With the linear rails, forcing is probably not an option. To minimize racking, I took the total error and divided it in half. I then moved one sensor one way, and the other sensor the opposite way, e.g., one sensor was moved in the Y-- direction and the other in the Y++ direction.


    6. To keep from chasing my tail, I used 1 2 3 blocks to make sensor adjustments. On the sensor moving toward the Y-- direction, I held a feeler gauge against the sensor, and slid a 1 2 3 block up against the feeler gauge. Then, I clamped the 1 2 3 block in place, removed the feeler gauge, and adjusted the sensor until it touched block. On the other side, I clamped the 1 2 3 block directly against the sensor. Then Loosened the sensor, inserted the feeler gauge, and tightened the sensor to it.


    In my never ending search for gantry squaring techniques, I've never seen the 5 Cut Method mentioned or described. The video I linked to is about 37 minutes long, but there is a lot packed into it. The technique requires a bit of math, but nothing more than simple add, subtract, multiply and divide. For those who are reluctant to invest the time watching it, the method multiplies the error amount by a factor of 4. The first cut reflects 1X amount of error. The second cut is error times 2. The third is error times 3 and the forth is error times 4. So. when you get to making adjustments, it's necessary to divide the error amount by 4. Also, you have to be aware that the error is over the length of the stock you are cutting, not the error over the width of the CNC. So, to get to amount of adjustment necessary for your machine, you take the error derived from the cuts, divide it by the length of the 4th offcut, and then multiple that result by the distance between your sensors. You then divide that result by 4 to get back to the 1x error. Then, you know the amount of error over the width of your machine. I believe that no matter what system you use, you have to extend the error to the distance between sensors.


    I’ve noted that while measuring the diagonals between 4 points will tell you whether you have a square or parallelogram (whether an absolute distance, or just the amount its out as between the cross angle measurements) as described previously, there is one part of that no one talks about. No one talks about how to translate the measured error (think 2 hypotenuses) into a usable numerical error for purposes of actually adjusting the gantry. I'm guessing that one might get some workable numbers by employing trigonometry, but I'm unwilling to relearn what I haven't used in almost 50 years just to find out I guessed wrong.


    One final note. I really like my embedded t-tracks. I see that most folks just cut the dado and screw in the tracks. I came across an alternate installation that I have really enjoyed using. A fellow in New Zealand came up with it. He goes by CNC Nutz on You Tube. His take on it sounded good to me, and I’ve found that it works well in practice. It took a bit more time putting together my spoilboard, but with the clarity of 20/20 hindsight, it was darned well worth it. Here’s a link to his video: . He’s very prolific. Many are very good.

    Regards,

    Gary Edwards




  20. #20
    Community Moderator difalkner's Avatar
    Join Date
    Nov 2014
    Location
    United States
    Posts
    730
    Downloads
    0
    Uploads
    0

    Default Re: Calibration question

    Quote Originally Posted by ger21 View Post
    Assuming the gantry is already perpendicular to the other axis....
    This was the key, Gerry. I loosened and/or removed the 24 fasteners required to square the gantry at least 5 times before I got it dead perfect. You can see in the spoilboard how many rectangles I cut so I could measure the diagonals. I cut multiple rectangles of 23x46, 22x44, and 20x40 and kept sneaking up on 'perfect' until I finally got it to where no matter what size rectangle or square I cut the diagonals are identical. At least, they're as identical as I can see with 3.0 reading glasses and getting right down on the tape to compare each measurement.

    At one point the diagonals different by 3/16" on the 23x46 rectangle. And I don't have to be concerned with temp and humidity because the shop is climate controlled 24/7 but that aside, neither temp or humidity will have enough effect on MDF in the span of a few minutes to cause a change. I'm comfortable enough with my measurements and methodology at this point to say this is as close in calibration as I can get this CNC machine.

    You can see here how many times I cut for these tests (patience is one of my strong points - LOL!) -
    Calibration question-002-calibration-rectangles-jpg

    David

    David
    Romans 3:23
    CurlyWoodShop - www.etsy.com/shop/CurlyWoodShop
    David Falkner - www.youtube.com/user/difalkner
    difalkner - www.instagram.com/difalkner


Page 1 of 2 12 LastLast

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
  •  


About CNCzone.com

    We are the largest and most active discussion forum for manufacturing industry. The site is 100% free to join and use, so join today!

Follow us on


Our Brands

Calibration question

Calibration question