Centroid Community CNC Support Forum

It happened again- everything was shut down so I started up with everything in a shut down state. I ran a very simple program as you can see (a bolt circle and a pocket in the middle). You can see that the z depth on the DRO shows about 1.2, but when I measured with a tape measure the tip of the center drill was about 1" from the top of the part- so this would have plunged in about .20" immediately without the slow approach. I have attached a couple of photos and a fresh report.

Marty, in response to your questions- yes this is my first CNC conversion, not even sure if I am using angular contact bearings. I already watched that video and did help somewhat on axis direction. Maybe im wrong but it seems using angular contact bearings and axis calibration are very important and believe I have calibrated the axis, becasue when I MDI an axis it moves the distance I ask it to -verified by a dial indicator. So not sure how if I am off slightly on some of these items that it should throw off the z axis by almost 1/2" at times, but then sometimes the z axis works perfect?

swissi, i do rehome everytime after I restart just to make sure. Also I have never noticed my z axis falling 1/2" or even less than that. From what I can tell it stays right where I shut down. I never notice an interuption of power and the axis always seems to move very smoothly.

Marty, the z axis ball screw is larger than the x or y axis. The x and y axis have the same pitch.

I would think If there was an issue with the z axis that it would show up when I do an MDI command to move. I have done this several times where I would make a command in the - and + direction and the DRO and the mechanical measurment would line right up. A 1/4- 1/2" of error should be showing up. Its only when I run a program?

Is the way I set up my tools wrong? Thought I did it right

I set parameter 3 to 2- which sets z ref to home. Then i began to set each tool- I currently have 6 tools setup. Also do I need to use a different WCS everytime I have a different part? I would not think so- but not 100% sure. It seems that if i rezero it should overwrite what was in the WCS previously

Marty, any reason why pitches would matter? Each axis is independent and as long as they are calibrated accordingly it should not matter- correct?

milling wrote: ↑Tue Jan 05, 2021 11:30 am I would think If there was an issue with the z axis that it would show up when I do an MDI command to move. I have done this several times where I would make a command in the - and + direction and the DRO and the mechanical measurment would line right up. A 1/4- 1/2" of error should be showing up. Its only when I run a program?

If you are loosing steps on Z, it will most likely occur during rapid moves. When you are doing your MDI commanded moves, what feedrate are you using?

milling wrote: ↑Tue Jan 05, 2021 11:49 am Marty, any reason why pitches would matter? Each axis is independent and as long as they are calibrated accordingly it should not matter- correct?

Your revs per inch do not look right. They are quite a bit off from each other. I would think if the screws were the same pitch, the revs/in would be MUCH closer.

Do you own a dial indicator? Do you own a test indicator? Gage blocks or at least 1-2-3 blocks?

I think you should go back and rewatch the Centroid Mill training video and I think you should read up the Mill Operator's Manual

Axis Label Motor Encoder Lash Comp. Limit Home Dir Screw
revs/in counts/rev. (inches) - + - + Rev Comp
1 X 6.4068 4000 0.0000 0 0 1 0 N N
2 Y 7.9802 4000 0.0000 0 0 0 1 Y N
3 Z 5.0787 4000 0.0000 0 0 0 1 Y N


If I were called to look at that machine, these are the things I would check:
1. Make sure the stepper drivers are indeed set to 4000 steps per motor revolution
2. I would make sure the axis are moving in the correct direction
3. I would calibrate each axis. If I call a move it better move exactly what I call. I would calibrate with a known dimension standard. You can use a 1-2-3 block if you mic it first.
4. I would check backlash, this checks the mechanical aspect of each axis of the machine. Marc (cncsnw) has a great way to measure backlash: http://www.cncsnw.com/Lashhowto.htm
5. I would inspect and test each home switch. Make sure they are configured in the Wizard correctly. I would test and set the soft limits appropriately and test them.

5. THEN I would set a Z reference position, set up a tool library, and measure the tool offsets.

Marty

Ill give it a try

I was a bit confused on the x and y as well- but thats what I got. It is confusing to me that the x is far off from the y. any reason why the y would be so off? I definitely counted 6 threads per inch and dont' know why it ended up at almost 8?

How should i set a z-ref position? Should I take one of my tool holders and just use that with say a rod installed?

yes, i do own a dial indiactor and 123 blocks. I used both to set the turns ratio as well as setting up my tools