Page 1 of 11

Cuts not repeatable, losing steps(FIXED) but will continue thread

Posted: Sun Nov 14, 2021 11:02 pm
by lavrgs
I am trying to make 10 features that are the same and possibly due to some clamping issues (part movement?) the first ring was not concentric with the first circular pocket.
20211114_175540_HDR.jpg
All the circular pockets were cut first then the the rings. After that, I located the center of each circular pocket to cut the ring around it with a simplified Intercon program... In the picture the third ring from the right was cut and recut to reduce the outer diameter of the boss. The second instance was cut, after finding the center, with the same program and the size was considerably larger. After re-cutting using the same program again the size was smaller than the third feature... in the order of 0.010 on the diameter. How can the same program cut features that are that different sizes...? I'm wondering if its a machine problem but I've done individual sample cuts of the feature with repeatable results.
I essentially did the same procedure to two features with different results...
The center pocket diameters are consistent...

Any suggestion would be apreciated

Re: Cuts not repeatable

Posted: Mon Nov 15, 2021 9:54 am
by CNCMaryland
Let me start with this question. Why not CAM this up and do all processes at the same time so you can eliminate error due to human interaction? How you are refinding center could be the problem.

Re: Cuts not repeatable

Posted: Mon Nov 15, 2021 12:37 pm
by lavrgs
CNCMaryland wrote: Mon Nov 15, 2021 9:54 am Let me start with this question. Why not CAM this up and do all processes at the same time so you can eliminate error due to human interaction? How you are refinding center could be the problem.
65861903300__1728B0B3-A083-4771-A40F-38EE25D59495.jpg
Gage used to find center of holes
Originally it was completely programmed. In hindsight I would cut all the features in each position but instead I cut 10 hole pockets and went back to cut the grooves. When the first groove was cut in was not concentric so I stopped the program and carried on by using a program just to cut the grooves. I have all the axes calibrated and backlash removed but over a distance of 24 inches there must be some position errors being introduced.
This error must be why the same program used to recut at the same position takes off extra material and makes the feature that was originally oversize become undersize

Re: Cuts not repeatable

Posted: Mon Nov 15, 2021 1:44 pm
by cnckeith
thanks for posting the report.zip.
it shows a overall turns ratio for x of 10.7998. how did you determine this value?
what is the pitch of the ballscrew on X?
how is the X axis motor connected to the the X axis ball screw?

Re: Cuts not repeatable> updated pullys & ratio

Posted: Mon Nov 15, 2021 1:52 pm
by lavrgs
cnckeith wrote: Mon Nov 15, 2021 1:44 pm thanks for posting the report.zip.
it shows a overall turns ratio for x of 10.7998. how did you determine this value?
what is the pitch of the ballscrew on X?
how is the X axis motor connected to the the X axis ball screw?
My servo motors are connected via pulleys and timing belts. My ball screws are 5turns per inch. I set the turns ratio using Marty’s method of using a base of known length and calculating the turns ratio via his spreadsheet. This was done with compensation off and then turned on after the commanded motion was verified.
The ratio of the pulleys is 1:2on Y (15:30)and 1:216666 on X (12:26)

Re: Cuts not repeatable

Posted: Mon Nov 15, 2021 2:08 pm
by cnckeith
with a 10 turns per inch screw and x axis motor belted 2:1. then the value of 10.7998 for the X axis overall turns ratio is impossible.
are you sure you don't mean that the screw is 5 turns per inch?

on the x axis, is the big pulley on the ballscrew or the motor? how many teeth are on each pulley?

what make model of servo motors are being used?

Re: Cuts not repeatable

Posted: Mon Nov 15, 2021 2:37 pm
by lavrgs
cnckeith wrote: Mon Nov 15, 2021 2:08 pm with a 10 turns per inch screw and x axis motor belted 2:1. then the value of 10.7998 for the X axis overall turns ratio is impossible.
are you sure you don't mean that the screw is 5 turns per inch?

on the x axis, is the big pulley on the ballscrew or the motor? how many teeth are on each pulley?

what make model of servo motors are being used?
You are correct 5 turns per inch. I’m not in the shop and I don’t want to commit to the teeth on the pulleys. I’m GUESSING the x is 12 on the motor and 26 on the ball screw. Y is 15/30… but I will verify. The motors are DMM .75kW
Are there any errors that can be causing the inaccuracies? For example the last frame in the intercon program is calling an outside circle 1.518 and I measured it at 1.525 and 1.510 after recutting without moving the table.

Re: Cuts not repeatable

Posted: Mon Nov 15, 2021 2:44 pm
by cnckeith
(2/1= 2) x 5 turns per inch = 10 turns of the motor yields 1" of travel. set the x axis turns ratio to 10. reboot the controller.

now test commanded distance value compared to actual movement over the longest distance possible. a machinist ruler and a fine pointer is good to do this "rough' test. after confirming that commanded movement does indeed match the actual moved distance you can move onto the "fine dro adjustment"

fyi.. typical turns ratio numbers after a very accurate measurement (fine dro adjustment as marty likes to call it) are just a few tenths or maybe a thousandths off if you have pulley ratios like 2:1 and 5 turn per inch ballscrews... so for the X axis on your machine a more realistic value after performing a the fine adjustment would be something like 10.001 or 9.999

Re: Cuts not repeatable Pulley Confirm

Posted: Mon Nov 15, 2021 4:26 pm
by lavrgs
Pulley confirmation; X 12 on Motor 26 on Ball screw Y 15 on Motor 30 on Ball Screw. So the X axis 10.7998 makes it bit of sense; 2.1667x5=10.8333. I will set my X turns ratio to 10.833333333.... The belt on the X axis was a bit tight so I loosened it so it wasn't in tension.
I will remove the backlash comp and do the test with engineering scale. With X Turns ratio set to 10.833333 it moved 12 inches based on a commanded X12. So by eyeball its very close....
IMG_2255.jpg

Re: Cuts not repeatable

Posted: Mon Nov 15, 2021 6:18 pm
by cnckeith
ok..great!