KP3 Error or Mill? <Solved, Probe issue>

[RJS100]

I have been using a PM833 benchtop mill converted to CNC for a little over a year. The x,y and z axis use closed loops steppers (Nothing high end). The ball screws are double nut ground screws.

I recently probed the top surface of my vise repeatedly. I set the WCS to zero for the z axis on the first probe cycle. See image. The figures I got were:
-.0003
.0008
.0000
.0004
-.0002

The document that came with the probe noted:
Z repeatability: .0002 Results: .00014

The above figures vary .0011

The question is which is not working correctly, probe or mill?

I do not own any sophisticated metrology equipment, so I lowered the spindle holding a spot drill over a precision ground cylinder. I marked a spot on the cylinder and rolled it under the spot drill. After careful setup, I raised and lowered the spindle .0002 and -.0002 a dozen times. Each time the cylinder rolled clear of the spot drill when raised and hit the cylinder when lowered (see image). This tells me that the mill has a resolution of at least .0004

Does this mean the probe is not working properly? I did check that the stylus is tight.

Richard

[WesM]

How is your stepper coupled to the ballscrew? I would begin with the drivetrain, since you are relying on the steppers encoder for position, I would take a close look at everything after the stepper.

Is your Z axis counterweighted? Have you done backlash comp?

Not knowing anything about your particular setup, my instinct tells me to look at how your Z is coupled. A cheap coupler, a poorly selected coupler or even a little wear can throw things off pretty quickly.

[RJS100]

Thanks for your thoughts. The z axis is not counter weighted, but it is coupled with an oversized Oldham coupling (high quality from Ruland). Backlash compensation is set at less than thou and when I test it, it is immeasurable. As a quick sanity check, I used a 1" dial test indicator, ran 900/1000 in each direction and it returns exactly to the starting position (see attched).

[suntravel]

First test is with a dial indicator attached to the milling head if it moves with probe speed always to the same position in -Z direction.

Move the head to a similar position like with the probe. Set the dialindicator to 0 set Z to zero
MDI:
G1 Z0.1 F5
G1 Z0 F5

Repeat several times.

Uwe

[RJS100]

Not sure what I was thinking. If I wanted to test the repeatability of the z axis, of course... use a test indicator

I performed the test with the test indicator as noted and the z axis returned exactly to zero each time. I repeated this test at least a dozen times.

Just for fun, I repeated the test again using a dial test indicator with .0005 accuracy (I know these don't measure linear movement, but they will measure repeatability). The dial test indicator also returned exactly to zero while running a similar test.

Looks like the probe may be the culprit. Anything else I should try?

[suntravel]

In this case I would send the probe in for service.

Uwe

[RJS100]

Thanks. That is what I had in mind. Kind of funny. When I first got the probe, I only once crashed it because I didn't fully seat the electrical connector. It broke the tip... but I have never crashed it since. I always touch the tip looking for the led to light every time I use it.

This will never be a cheap hobby... but certainly fun!

[RJS100]

I received my probe back today. It was returned with a note indicating the specific repairs that were made and a set of test data showing that the probe is now working with impressive accuracy:
X Center Range: .00003
Y Center Range: .00003
Z Repeatability: .00017


I Installed a new stylus and made the appropriate adjustments to ensure that the stylus was concentric. I then ran the standard probe calibration using the Swissi App and a calibration ring. The actual bore of my test ring is: 25.4020mm or 1.00008 inches

The results from this standard calibration were:
Measured Bore: 1.0006
Error = .00052
X = .99735
Y= 1.00060
Error in X & Y = -.00325


The error of .00052 and the error in X & Y of -.00325 seemed a bit higher than I expected. Am I looking at this the wrong way? One thought I have is that when I was adjusting the concentricity of the probe, when I backed out one of the three adjusting screws, perhaps I was to delicate with them and did not tighten them enough. I will re-check the concentricity tomorrow to see if the stylus is still concentric.

Richard

[suntravel]

Hi Richard,

a typical trip map of kinetic probes is a bit like a triangular shape, but since most measurements are XY direction turn the probe a bit and find out the position with equal deviation in X and Y. Mark this position on probe and spindle to use it later also.
Do the calibration in this position.

Uwe

[RJS100]

Thanks... Really appreciate that. That will be tomorrow's homework.