Centroid Community CNC Support Forum

bilorn wrote: ↑Wed Jun 06, 2018 8:06 pm All AC is star grounded. When scoping the y axis step output noted that when wires were moved the y would move slightly.
Reterminated all step direction and enable signals.
While wiggling wires again the same thing happened and once the y axis started running, had to power down acorn to stop it.
I am suspecting a cold solder joint on the Acorn connector.
Next I will remove Acorn and inspect, possibly reflow the connector pins.
All of this happened with the vfd not powered.
It isn't looking good.

But you said the axis moved with Acorn POWERED down previously?
You could move the axis to the 4 axis on the Acorn board and tell it to be Y and try that for diagnostic purposes....I don't see a reason why it wouldn't work?

Good luck chasing down that gremlin

I scoped step outs with servo not powered, noted waveform at 60 hz with hi freq riding it for 3 ms. not a sine wave.
Tried MPG and found it working! Powered servo and MPG continued to work! I inverted step in wizard and 60 hz wave was gone.
Ran peg-a program to completion. Problem isn't gone but much better, each hole got about .005 deeper than previous one, total
error about .060. Not feeling good with weird problem going without finding a cause. I think error is caused by vfd now.
I have no clue as to what made machine go wild, have trust problems now!
Thanks again
Bill

This morning MPG not working again.
I think this may be a windows problem, without the cnc12 running the MPG still displays DRO data,
data goes from Acorn to windows via ethernet to usb puck to MPG, I think.

Ran Peg-a program and still had .020 down drift on z for the 12 pockets.
If it is noise why the error in one direction only?
Could it be that the drive is getting a free signal momentarily?

Bill,
I would check the connections on all your wiring,
I had an issue with my setup one axis doing weird stuff...
it ended up being a loose connection to that drive motor..
then I went through and check the rest of the wire connections
haven't had an issue since the wires seemed to loosen up a bit
after running the machine awhile in disclosure I am running steppers not servos.

Charlie

bilorn wrote: ↑Thu Jun 07, 2018 7:23 pm This morning MPG not working again.
I think this may be a windows problem, without the cnc12 running the MPG still displays DRO data,
data goes from Acorn to windows via ethernet to usb puck to MPG, I think.

Ran Peg-a program and still had .020 down drift on z for the 12 pockets.
If it is noise why the error in one direction only?
Could it be that the drive is getting a free signal momentarily?

Did you try to move the axis from Virtual Control Panel? Try using a mouse to press the virtual buttons.

I did reterminate the xyz step dir and enable points on the Acorn, this helped the situation but didn't cure completely.
Inverting the step output helped out, on the scope it was hard to see step pulses thru the fuzzy 60hz wave with them
not inverted, it's a wonder the drives could pick out the steps. I am reasonably sure the z drift problem is due to the
vfd to spindle cable being run next to z axis power and encoder cables. I was thinking just analog 400 hz and less
wouldn't require shielded cable, wasn't thinking this 3 phase power was also pwm-ed. I should have caught this after
my first run of the hole program. It was with a 1/4" HS end mill at 4800 rpm. No problem until I switched to a carbide
router bit and raised the rpm to 20000.


I use the mouse on the virtual panel most of the time, less prone to fat finger, especially z- and start button close
together. I am currently using the mouse pad on a toshiba laptop, but I also used a wireless mouse without problem.
Ordered shielded cable for spindle.

bilorn wrote: ↑Sat Jun 09, 2018 12:10 am I did reterminate the xyz step dir and enable points on the Acorn, this helped the situation but didn't cure completely.
Inverting the step output helped out, on the scope it was hard to see step pulses thru the fuzzy 60hz wave with them
not inverted, it's a wonder the drives could pick out the steps. I am reasonably sure the z drift problem is due to the
vfd to spindle cable being run next to z axis power and encoder cables. I was thinking just analog 400 hz and less
wouldn't require shielded cable, wasn't thinking this 3 phase power was also pwm-ed. I should have caught this after
my first run of the hole program. It was with a 1/4" HS end mill at 4800 rpm. No problem until I switched to a carbide
router bit and raised the rpm to 20000.


I use the mouse on the virtual panel most of the time, less prone to fat finger, especially z- and start button close
together. I am currently using the mouse pad on a toshiba laptop, but I also used a wireless mouse without problem.
Ordered shielded cable for spindle.

Are you saying air cuts with no spindle running and VFD off works ok? Wireless MPG works without VFD off?

The test program with multiple 1 inch z moves was done with an indicator mounted in the spindle, no problem with
z on this test. Dry runs of the peg-a program with vfd off had varying amounts of error on the z axis, but always
in the downward ( gravity? ) direction. I believe two problem exist, 1. unshielded vfd cable. 2. A connection problem,
noted when step direction enable cable is is jiggled at Acorn connection. Even after the retermination, jiggling the
cable would cause the y axis issue sound but doesn't appear to move. The cable I am using is 2 pair 20awg teflon
with copper braid shield and teflon jacket with shield tied to d-shell body at drive.
While waiting for the vfd cable I will remove Acorn and take a look at connector pins solder connections.
In the past I've seen problems like this with loosened or cold solder joints, not likely but I gotta look!

The MPG problem however is not related to any of the above, it might be a windows problem, I don't know my way
around win10 very well and don't look forward to learning it. XP was better!

What kind of drives and motors do you have?
Marty

DMM DYN4 x and z are 400w y is a 750w.