cumulative X, Y, and Z errors - <resolved>

[JPT]

smoothing on ? turn it off. (another classic )

Not sure with a rectangle, but if you cut a simple circle in few depth steps with smoothing you get something similar.

Smoothing is off.
.

[JPT]

(have no idea how to read those report's)

The machine was turned on from a cold start.

The mill was homed prior to all tests.

The first test in this XL sheet is a test to determine if the X axis is moving the distance in the CNC12 DRO display of the Acorn controller.

The purpose of this test was to determine if the settings for distance moved per ballscrew turn and stepper motor steps were set correctly.

The X axis was tested in both - and + directions over a range of 10mm.

The actual distances were determined using a Mitutoyo digital dial indicator attached to the spindle motor mount and measuring a vise attached to the table's fixture plate.
.

.
I considered the resulting measurement deltas to be reasonable for this class of mill.

On a pass/fail basis, I judged this test to be a Pass.

Note: In addition to this test, I also did an eyeball max X axis range test using a 300mm ruler/scale during the initial setup of the Acorn to confirm that the indicated DRO distance matched the actual distance moved.
.
.
The second test was to determine if the cumulative error stackup seen in the programs air-cut tests and the test parts could be duplicated by manually moving the X axis 10 full cycles through its complete range of motion.

I used the MPG to manually cycle the X axis to max - and max + 10 times.

The measurement was made by a Haimer 3D Taster probe against the vise fixture face.
.

.
Again, I considered the measurement delta to be reasonable for this class of mill.

On a pass/fail basis, I judged this test to be a Pass.
.

[cnckeith]

change the steps per rev to 2000 on the driver and in the Wizard, write settings and reboot.

related reading.
https://centroidcncforum.com/viewtopic.php?f=63&t=1801

[cnckeith]

4 turns per inch?

[JPT]

(have no idea how to read those report's)

The machine was started from a cold start.

The mill was homed prior to the tests.

The measurements were taken with a Haimer 3D probe mounted in the spindle.

The Haimer probe was calibrated for concentricity in the mill spindle prior to the tests.
.
.
These tests were air cuts using dedicated test nc programs produced by Fusion 360 and the standard Centroid mill post processor.

The purpose of these tests were to test the amount of X, Y, and Z cumulative error stackup.

The measurements were taken via a Hiamer 3D probe against surfaces of the vice which was attached to the table's fixture plate.
.

.
.
Three test programs were used: v8, v9, and v10.

Each program had more steps on the X & Y and then the X, Y, and Z axes.

The v8 program is a single pass, single cut.

The v9 program adds seven steps down in Z.

The v10 program has seven steps down in Z and six steps in on X and Y.
.
.
The cumulative error amounts from the air cuts were significant.

The v10 test yielded 1.02mm error in X, 0.83mm error in Y, and 0.06mm error in Z.

On a pass/fail basis, I considered these results a fail.
.

[JPT]

4 turns per inch?

The machine is set up in metric.

It's 4mm per turn.

I tested this by manually turning the X axis ballscrew and measuring with a mm ruler/scale and then confirming with a dial indicator.
.

mm-per-turn.jpg (7.76 KiB) Viewed 804 times

[cnckeith]

ok great, why is the drive at 10,000 these yakos have poor torque performance at that level.

[JPT]

why is the drive at 10,000 these yakos have poor torque performance at that level

Because that was drive setting used with the prior controller.

The prior controller produced solid and reliable mill operation using those drive settings, so I left them "as is" to reduce the number of variables with the retrofit.
.
.

change the steps per rev to 2000 on the driver and in the Wizard, write settings and reboot.

related reading.
https://centroidcncforum.com/viewtopic.php?f=63&t=1801

Thanks for the specific suggestion and the link.

That link is a very useful and informative post and is what I used to determine the Acorn settings with the "as is" drive settings.

I will change the dip switches on the drives and the corresponding Wizard settings and rerun the air cut tests.

[cnckeith]

be sure to set the step rate freq. to 100,000 as well.

[H.C]

meant the cnc12 report

coming also from another controller, had to crank up the steps on the drivers to get the same smooth movement on the centroid