Nice illustration video. I’m an ex K & T Service Tech who installed many types of 3-4-5 axis CNC equipment in Aerospace back in the day. We verified alignments with a 14” precision Granite on most, sometimes a 24”. We also ran laser yaw, pitch and roll on our ways before the assembly. Finally lost motion correction on each ball screw and rotary table was performed. Keep up the good work.
I am really glad you made this video. I am building my own mill, and was wondering how to check z axis motion alignment with the spindle axis. I have not seen anyone cover that so far, and that too so elegantly. Thank you so much for this
Nice video - very clearly explained. (I'm not convinced that you're measuring much when checking the XY plane, but overall you've got the bases covered :) )
It's best to build each axis as it's own actuator instead of mounting rails directly to the machine base. That way you can add adjustment screws for fine tuning. It's far easier and better than using shims.
It's important for the two rails on an axis to be perfectly parallel, so after you spend a lot of time dialing that in you don't want to loosen them again.
Hi. Awesome build! Do you have any feedback re the linear rails. I am building a very similar machine and am wondering if I should go with genuine HIWIN rails or if the ebay rails do a good enough job. Thanks for your content. I have watched your vids far too many times!
The cheapo ebay rails have worked well enough for me so far. If you're going for a hobby-level machine (only run occasionally, don't need extreme accuracy) I think they're fine. Sometimes the bearings come very dirty, but you can clean them out to make them run smoother. The rails sometimes aren't perfectly straight, but you can measure it and use the mounting bolts to straighten them.
@@imnoexpertbut thanks a lot for your reply. I really appreciate it. I’m using the machine to prototype some parts from forged carbon fibre plates. Not as demanding as metal but there are other considerations like dust control using liquid. I’m looking for accuracy to be honest. I’m looking for repeatable accuracy of 0.03mm to 0.08mm (0.001” to 0.003”). I believe you have already reached tolerances close or better to this before squaring up your mill. Really interested to know what results you are getting now in terms of repeatable accuracy. Thanks again for all of the valuable content and comments.
One thing that you should be able to do with your CNC software is make an error table, the software should know which way things are out of alignment and correct for that.
I love informative videos, what is the torque of your servos, are you able to machine steel accurately, doesn’t matter how fast it is im just curious about accuracy
Just my 2 cents, but I am not sure you are testing what you think you are testing. As an example lets take your picture at 3:40. Assume the X axis is perfect, except for the tilt. You'll never measure any tilt with the setup you are showing (moving in the x direction). You would have to move the Y axis for that. Even worse, if the tilt were around the Y axis direction, neither X nor Y movement would show in the dial indicator. Edit: I think I should clarify that the any tilt of the X table around the Y axis is not really a problem because it can be corrected entirely by tramming the spindle.
I know it's confusing, and it's also hard to talk about, but that picture at 3:40 is showing the X and Y axis being perfect, and the *table* is tilted relative to the X and Y axes. That's why it will show up on the indicator. If the table was perfectly aligned with the X/Y but the axis itself was tilted, then you are correct and the indicator dial wouldn't move when the axis moved. You can measure how far off a surface is from its direction of travel by putting an indicator on it and moving it along that direction of travel. You can't measure how far off an *axis* is by putting an indicator on it and moving it along the direction of travel. This old tony has a Maho CNC video where he demonstrates that. Anyway, all this stuff is surprisingly hard to think about from first principles, which is why I wanted to make a video about it.
I can buy large pieces of stock online for cheap with free shipping. 6x20 inches 1 inch thick aluminum for around $50 shipped. I couldn't find prices like that online for steel, and there aren't places around me that will sell in small quantities. I think epoxy granite would be a great option, but I haven't tried it yet.
Nice illustration video. I’m an ex K & T Service Tech who installed many types of 3-4-5 axis CNC equipment in Aerospace back in the day. We verified alignments with a 14” precision Granite on most, sometimes a 24”. We also ran laser yaw, pitch and roll on our ways before the assembly. Finally lost motion correction on each ball screw and rotary table was performed. Keep up the good work.
I am really glad you made this video. I am building my own mill, and was wondering how to check z axis motion alignment with the spindle axis. I have not seen anyone cover that so far, and that too so elegantly. Thank you so much for this
Nice video - very clearly explained. (I'm not convinced that you're measuring much when checking the XY plane, but overall you've got the bases covered :) )
Cool video .. very informative.. but are those chips on you lead screw when you were facing the table ... I have to call the cops on you
Thanks You .Rails on your x axes should have lateral support(all 3 axes must have)
This is awesome! I’ll need this info for my CNC Mill lol
Very helpful video. thanks for posting! :D
It's best to build each axis as it's own actuator instead of mounting rails directly to the machine base. That way you can add adjustment screws for fine tuning. It's far easier and better than using shims.
Great video! I have an error in my Taig Z and have it all apart... traced it to the column not being square with the base.
Great video! Where can I get this "plate with holes" you are making calibration with?
Can you do a video of how you wired your arduino to the tb6600 off that screw block I’m doing same design but lil diffrent would really appreciate it
Great video! How much of alignemnt can be adjusted via the linear rails themselves? Thx!
It's important for the two rails on an axis to be perfectly parallel, so after you spend a lot of time dialing that in you don't want to loosen them again.
@@imnoexpertbut I see! Makes sense... So we set that up first and don't touch it again while we go on to tramming the other axes.
Can you please do a video of how to set up the electronics I have the same electrical layout but need help setting it up
Well done dude!!! I thank so much for this informative video. ❤🌷🙌
Whoa! Did your Y-axis grow or was it always that size?
Great stuff - Thanks for the update! I needed some motivation to finish my OB Minimill mods.
Yep, always that size. Moving it to the edge of the workbench reminded me just how heavy it is!
Hi. Awesome build! Do you have any feedback re the linear rails. I am building a very similar machine and am wondering if I should go with genuine HIWIN rails or if the ebay rails do a good enough job. Thanks for your content. I have watched your vids far too many times!
The cheapo ebay rails have worked well enough for me so far. If you're going for a hobby-level machine (only run occasionally, don't need extreme accuracy) I think they're fine. Sometimes the bearings come very dirty, but you can clean them out to make them run smoother. The rails sometimes aren't perfectly straight, but you can measure it and use the mounting bolts to straighten them.
@@imnoexpertbut thanks a lot for your reply. I really appreciate it. I’m using the machine to prototype some parts from forged carbon fibre plates. Not as demanding as metal but there are other considerations like dust control using liquid. I’m looking for accuracy to be honest. I’m looking for repeatable accuracy of 0.03mm to 0.08mm (0.001” to 0.003”). I believe you have already reached tolerances close or better to this before squaring up your mill. Really interested to know what results you are getting now in terms of repeatable accuracy. Thanks again for all of the valuable content and comments.
One thing that you should be able to do with your CNC software is make an error table, the software should know which way things are out of alignment and correct for that.
I love informative videos, what is the torque of your servos, are you able to machine steel accurately, doesn’t matter how fast it is im just curious about accuracy
I haven't tried to mill steel yet! The torque of the steppers I use is 269oz.in/1.9Nm.
And what kind of engines did you use stepper or servos, and can give a link to them or their name
Steppers. It's all here: ua-cam.com/video/JAylwj41W8k/v-deo.html
Just my 2 cents, but I am not sure you are testing what you think you are testing. As an example lets take your picture at 3:40. Assume the X axis is perfect, except for the tilt. You'll never measure any tilt with the setup you are showing (moving in the x direction). You would have to move the Y axis for that. Even worse, if the tilt were around the Y axis direction, neither X nor Y movement would show in the dial indicator.
Edit: I think I should clarify that the any tilt of the X table around the Y axis is not really a problem because it can be corrected entirely by tramming the spindle.
I know it's confusing, and it's also hard to talk about, but that picture at 3:40 is showing the X and Y axis being perfect, and the *table* is tilted relative to the X and Y axes. That's why it will show up on the indicator. If the table was perfectly aligned with the X/Y but the axis itself was tilted, then you are correct and the indicator dial wouldn't move when the axis moved. You can measure how far off a surface is from its direction of travel by putting an indicator on it and moving it along that direction of travel. You can't measure how far off an *axis* is by putting an indicator on it and moving it along the direction of travel. This old tony has a Maho CNC video where he demonstrates that. Anyway, all this stuff is surprisingly hard to think about from first principles, which is why I wanted to make a video about it.
@@imnoexpertbut Ah, I see what you mean now. Thanks!
why did you use aluminium?
I can buy large pieces of stock online for cheap with free shipping. 6x20 inches 1 inch thick aluminum for around $50 shipped. I couldn't find prices like that online for steel, and there aren't places around me that will sell in small quantities. I think epoxy granite would be a great option, but I haven't tried it yet.
@@imnoexpertbut ah ok thanks, I thought aluminium had some other properties.
90 degrees is overrated 😁👍 Imagine how much better the world would be f x, y, and z axis were at 84 degrees to each other.
Great video. Sad that you are using imperial unit system.
There is no system in Imperial... :D