Hello Guest it is March 29, 2024, 03:45:46 AM

Author Topic: Tolerance puzzle.....  (Read 4105 times)

0 Members and 1 Guest are viewing this topic.

Tolerance puzzle.....
« on: February 20, 2013, 08:43:43 AM »
I don't get to do pocketing often, but when I do I am puzzled to see that the difference between my drawing dimensions and actual pocketed dimensions can vary by as much as .020". I use Mastercam and compensate for tool wear; there's no noticeable backlash that I can see. Have I forgotten something in Mach3?
In my case, I simply have a rectangle measuring .406 wide, and the actual cut pocket width measures .393". I am using a .125 end mill compensated for .002 wear.
thanks
tc

Offline Fastest1

*
  •  920 920
  • Houston, TX
    • View Profile
Tolerance puzzle.....
« Reply #1 on: February 20, 2013, 08:45:28 AM »
Which axis is showing the variance or is it both? If it is just 1 axis. I would think it is an axis calibration or backlash compensation value error.
I want to die in my sleep like my grandfather, not like the passengers in the car! :-)

Offline Fastest1

*
  •  920 920
  • Houston, TX
    • View Profile
Tolerance puzzle.....
« Reply #2 on: February 20, 2013, 08:46:29 AM »
Climb or conventional cut?
I want to die in my sleep like my grandfather, not like the passengers in the car! :-)
Re: Tolerance puzzle.....
« Reply #3 on: February 20, 2013, 09:05:24 AM »
It's a climb cut, and ( I am embarassed) I can't take a very accurate measurement of the Y axis, given that my caliper wont go that far. It's a good observation, however, and I'll figure out a way to measure the Y cut....
tc

Offline BR549

*
  •  6,965 6,965
    • View Profile
Re: Tolerance puzzle.....
« Reply #4 on: February 20, 2013, 12:13:19 PM »
Try doing it as a conv cut and see IF the error goes the other way.

(;-) TP

Offline Fastest1

*
  •  920 920
  • Houston, TX
    • View Profile
Re: Tolerance puzzle.....
« Reply #5 on: February 20, 2013, 02:47:52 PM »
Why be embarrassed? I would use a dial indicator myself.
I want to die in my sleep like my grandfather, not like the passengers in the car! :-)
Re: Tolerance puzzle.....
« Reply #6 on: February 20, 2013, 03:08:18 PM »
Why be embarrassed? I would use a dial indicator myself.

I failed to mention that the pocket is only .100 deep. I pocketed a small rectangle in order to use my caliper for accurate x and y measuring and I found that a .250 x 2" drawing cuts at .228 x 1.991.
tc

Offline BR549

*
  •  6,965 6,965
    • View Profile
Re: Tolerance puzzle.....
« Reply #7 on: February 20, 2013, 04:35:04 PM »
IS masterCam doing the tool comp or MACH3.  Where are you appling the tool wear value ?

Have you verified the steps per values are correct in config ??

Was that a climb cut or Conv?

(;-) TP
« Last Edit: February 20, 2013, 04:36:58 PM by BR549 »
Re: Tolerance puzzle.....
« Reply #8 on: February 21, 2013, 10:42:01 AM »
I've just checked for backlash, and there's none to speak of( less than .001).  The MC tool comp WAS set to computer. I changed it to wear  and still no change. In fact I entered a wear of .010 for the tool in the tool offset and still no change. I am using wood for testing purposes. The only change that I can think of is that the pc went out during the storm, and then I restarted Mach3. Here's the code which results in a 0.015 difference in width.  The sample pocket geometry is .300 and the actual cut is .285.
%
O0000(CALIBRATION)
%
O0000(CALIBRATION)
(DATE=DD-MM-YY - 21-02-13 TIME=HH:MM - 10:47)
(MCX FILE - D:\MY DOCUMENTS\MY MCAMX5\MCX\THROTTLE\THROTTLE_SECOND VERSION\ABS TEMPLATESFROMINSKAPE WITH FIRST ONE OFF.MCX-5)
(NC FILE - D:\MY DOCUMENTS\MY MCAMX5\MILL\NC\THROTTLE+SECOND VERSION\CALIBRATION.NC)
(MATERIAL - WOOD INCH)
( T232 |  1/8 FLAT ENDMILL | H232 )
N100 G20
N102 G0 G17 G40 G49 G80 G90
N104 T232 M6
N106 G0 G90 G54 X.376 Y11.148 A0. S4278 M3
N108 G43 H232 Z2.
N110 Z.2
N112 G1 Z-.05 F16.16
N114 X2.1572 F26.16
N116 Y11.2292
N118 X.376
N120 Y11.148
N122 X.3291 Y11.1011
N124 X2.2041
N126 Y11.2761
N128 X.3291
N130 Y11.1011
N132 F16.16
N134 G41 D232 X2.2041 F26.16
N136 Y11.2761
N138 X.3291
N140 G40 Y11.1011
N142 G0 Z2.
N144 M5
N146 G91 G28 Z0.
N148 G28 X0. Y0. A0.
N150 M30
%
« Last Edit: February 21, 2013, 10:47:54 AM by screwhead »
Re: Tolerance puzzle.....
« Reply #9 on: February 21, 2013, 08:55:14 PM »
Well, after changing the bearings, making adjustments to Mach3 backlash, lowering the steppers' top velocity values, I finally was able to control the cut.
It appears that the speed at which I was cutting a "calibration square in wood", 56ipm, was too fast.  At this speed, a .300 wide square .060 deep, was undercut by .010 before the backlash compensation. Cutting such a small square subjected the steppers to very frequent go no go motion.  But even with this adjustment, I could barely get the correct cut dimension. When I lowered the feed speed to 15ipm, the cut ended up at .305.  Finally some real change!
It appears that this is due mostly to either the feed speed being too fast or the stepper motor values settings being too high.
What's strange is that I've made many cuts in the past and never had such big discrepancies.  Then again, my cuts were for larger dimensions.
I'll need to keep checking . Now I've lost all confidence in the machine.
tc