Trying to troubleshoot some jerky motion in cuts

[cnckeith]

wow nice machine.. this machine is crying out for some real servos! who would run such a big heavy machine with stepper motors?

[cnckeith]

look for loose stuff. have you checked for lost motion in the mechanical system?

the climb cut you mention will 'bring out' any mechanical problems and amplify them.

[centroid467]

1-Check the actual voltage supply to the drives if you are reusing the oem power supply. They might have several taps to match different incoming voltage.

1- Just went to triple check and looks like I was incorrect. (Disclaimer- I am not good with electronics so its best to assume I don't know what I'm talking about) The voltage from the transformer (machine is 220 3phase) to the Power board (bottom left label 50vdc) is reading 35vdc. Each wire, grey and blue reads 35vdc when I probe one terminal at a time to ground (if this is the correct way)

Now at the top- where the wires go from the power board to the stepper drivers reads 40-70vdc. But, when I test them with a multimeter is it reading 83vdc. Sounds like this could be the culprit? I have no idea how this powerboard works but i assume its suppose to take 50vdc in and output 40-70vdc? But its taking in less voltage and putting out more than it should?

From the silkscreen labels on the power board: one would expect 50 VAC (likely RMS) to the bottom terminal. This can be measured by using the multimeter in volts AC mode across the gray and the blue wires. Be very careful to not short them - it is best to disconnect power and solidly attach leads from the meter to the wires using properly rated clips or terminals before turning power back on for the measurement.

You should expect to read an average DC output of ~45 VDC assuming you have 50 VAC RMS coming from the toroid to the power board. This is calculated from:
Peak DC voltage: 50 Vrms * sqrt(2) = 70.7 VDC
Average DC voltage: ( 2 * 70.7 VDC) / pi = 45.0 VDC

The driver Terminal reads the following voltage:

(24-80vdc) VCC+ 82-32VDC
GND- 0VDC

What is the DC voltage from the VCC+ terminal to the GND- terminal on the drives? It looks like you measured it to an appropriate GND but I'm just confirming that what you are reporting makes sense.

[muibubbles]

wow nice machine.. this machine is crying out for some real servos! who would run such a big heavy machine with stepper motors?

Thank you, and AXYZ the company would. I'm seeing sergei's post in the FB group with his machine flying and it definitely makes me want to buy some clearpath servos... I'm just stuck in the middle if its worth throwing more money and investing more time into this machine if it can't produce the results I need...

[muibubbles]

look for loose stuff. have you checked for lost motion in the mechanical system?

the climb cut you mention will 'bring out' any mechanical problems and amplify them.

I've checked the belts
Disconnected the belts and checked for shaft play in the transmission (large pulley to pinion)
I don't really know what I was intending with this video but the machine "feels" pretty tight to me?
https://youtube.com/shorts/G2iMSoIGEfI

Any specific tests you have in mind for lost motion? I'm running out of ideas what to check at this point... Like I'm pretty sure my rack is worn on the y-axis (gantry) but I don't have a definitive way to verify it. When i Mic out the rack, everything falls between +/- .003-.004" so I assume that's within range?

[muibubbles]

1-Check the actual voltage supply to the drives if you are reusing the oem power supply. They might have several taps to match different incoming voltage.

1- Just went to triple check and looks like I was incorrect. (Disclaimer- I am not good with electronics so its best to assume I don't know what I'm talking about) The voltage from the transformer (machine is 220 3phase) to the Power board (bottom left label 50vdc) is reading 35vdc. Each wire, grey and blue reads 35vdc when I probe one terminal at a time to ground (if this is the correct way)

Now at the top- where the wires go from the power board to the stepper drivers reads 40-70vdc. But, when I test them with a multimeter is it reading 83vdc. Sounds like this could be the culprit? I have no idea how this powerboard works but i assume its suppose to take 50vdc in and output 40-70vdc? But its taking in less voltage and putting out more than it should?

From the silkscreen labels on the power board: one would expect 50 VAC (likely RMS) to the bottom terminal. This can be measured by using the multimeter in volts AC mode across the gray and the blue wires. Be very careful to not short them - it is best to disconnect power and solidly attach leads from the meter to the wires using properly rated clips or terminals before turning power back on for the measurement.

You should expect to read an average DC output of ~45 VDC assuming you have 50 VAC RMS coming from the toroid to the power board. This is calculated from:
Peak DC voltage: 50 Vrms * sqrt(2) = 70.7 VDC
Average DC voltage: ( 2 * 70.7 VDC) / pi = 45.0 VDC

The driver Terminal reads the following voltage:

(24-80vdc) VCC+ 82-32VDC
GND- 0VDC

What is the DC voltage from the VCC+ terminal to the GND- terminal on the drives? It looks like you measured it to an appropriate GND but I'm just confirming that what you are reporting makes sense.

I completely missed that, I didn't even realize it was in VAC! Uploading a few more pics. Toroid to power board is reading ~62VAC
(24-80vdc) VCC+ 82-32VDC Sorry that was a typo it was reading 82-83VDC but now its reading 80VDC This reading is consistent from the power board to the terminals at the driver

[Ken Rychlik]

The tag above the lower right torrid coil shows what color to use with what incoming voltage. I changed to a different wire to get my dc voltage down. I forgot the colors and can't see your tag good. Mine was wired for 208v input and I have almost 250 so I changed it to the color for 240. Mine was putting out to much voltage and a simple wire change fixed it. The big wires going to the green board are not changed. You change where you apply the input voltage to that coil.

[muibubbles]

The tag above the lower right torrid coil shows what color to use with what incoming voltage. I changed to a different wire to get my dc voltage down. I forgot the colors and can't see your tag good. Mine was wired for 208v input and I have almost 250 so I changed it to the color for 240. Mine was putting out to much voltage and a simple wire change fixed it. The big wires going to the green board are not changed. You change where you apply the input voltage to that coil.

I think you nailed it!!! The previous owner had it wired for 208VAC and I just tested my line coming in and I have 240VAC, to be exact My legs are 254, 254 and 246 so I am getting even higher voltage to my shop. So I swapped the orange wire for the white and things are now within spec (for the most part)!

The power board is receiving ~53-54VAC vs ~62VAC Prior
The power board is now putting out ~70VDC vs 80-83VDC as well as the power going into the drivers.
Terminal ports on the drivers range from 32VDC-36VDC vs all of them getting 41VDC prior.

Here is the new clip calibrating turns per rev, the motors sound better and if you watch the needle it has a nice fluid sweep motion vs prior it was jerky.



I also ran a test cut using intercon and the results look very promising. MUCH better and more like what I'd expect from a cnc router. I dont want to get ahead of myself before I make a real cut but I'd be happy with these results. The rest would be fine tuning my feeds/speeds and strategies.

New issue: when I was calibrating turns per rev, I was trying to calibrate my Slave axis (tied to my other X-axis motor). I assume there is less wear along the A-axis gear rack because when I command a 1" move, it moves 1.008". After about 10 minutes of constantly calculating and changing the turns per rev and nothing happening, i realized that it was paired and taking the X-Axis calibration. What would be my solution to this?

Thanks again for everyones help, I will post back tomorrow to confirm if the problem has been solved!

[Ken Rychlik]

The slaved axis is paired and assumed to be everything exactly the same. As long as you have the pro license, axis squaring stays in the system. Mine had a tag inside the control box that listed the steps per inch at 1600. It took a little math but wasn't hard. Just divide that number by 1600 to get your turns if you have that tag. With a simple ratio, you can tell it to move 20" and measure how far it went. Adjust your turns to correct. I do it at full length of the machine after I am close and it has been helpful.

[muibubbles]

The slaved axis is paired and assumed to be everything exactly the same. As long as you have the pro license, axis squaring stays in the system. Mine had a tag inside the control box that listed the steps per inch at 1600. It took a little math but wasn't hard. Just divide that number by 1600 to get your turns if you have that tag. With a simple ratio, you can tell it to move 20" and measure how far it went. Adjust your turns to correct. I do it at full length of the machine after I am close and it has been helpful.

Correct, I've always had them paired with the assumption that they are exactly the same. However as I'm re-tuning everything, I though for the first time I should just check out what the slave axis is reading. With the same values as the X-Axis, its traveling 1.01" when commanded 1" So i'm wondering if there is a way to override the turns per rev since there is a difference, or is it negligible and just run it?

Curious, what kind of cnc machine are you working on?