Centroid Community CNC Support Forum

First post; probably should have posted sooner as I've been chasing the same problem for two weeks and am becoming very frustrated. This is also my first "from scratch" CNC build (I've done an OpenBuilds kit in the past with success, but it's far simpler). Searched the forum for anything that may help and watched many, many YouTube videos. Can't seem to figure out what's happening with my new build.

I'm using an Acorn board, DYN4 drivers, DMM servos (two 400W, one 750W) with the DMM DB25 DYN4 AC Servo Drive to Centroid Acorn Cables [CA-CUACRN-M4-1]. It's wired up as prescribed in S14989.dwg and power wiring is in accordance with DMM guidance (line reactor, main breaker, main emi filter, 24VDC contactor connected to estop, individual emi filters for each driver). Power is separated from signal lines and the enclosure is larger than needed at 24"Tx20"Wx12"D.

I'm using a Dell Optiplex 7020 with Intel i7 well above the 1500 benchmark for milling Centroid specifies. I've set up the CNC PC per Centroid's directions and CNC12 communicates with the Acorn board with no obvious issues there (0% packet loss, tested several times).

When using the USB cable directly connected to the DYN4 in RS232 mode, the servos turn the direction and speed commanded perfectly and smoothly.

When in step and direction mode and connected to the Acorn the servos will spin the direction and speed commanded, but when not being jogged the servos move very slowly (drift, one direction) and erratically (almost like a popping sound).

I've tried no load tuning the servos and it didn't help.

In the latest configuration, the x-axis still has the drift and popping, but will spin "forward" perfectly and not spin "reverse" at all. The y-axis and z-axis spin "forward" perfectly and spin in "reverse" slightly slower -- no drift or popping on the y-axis and z-axis. Super confused how the axes now have inconsistent symptoms.

Any help or nudges in the right direction would be greatly appreciated. Happy to provide whatever evidence is needed to enable troubleshooting.

When you first set this up did you do a bench test of the Acorn as instructed? Then did you wire up ONE drive at a time and get it working outside the cabinet? Or did you toss everything into the cabinet and wire it up? Please post a fresh report and some screen shots of your DMM software settings with each drive so we can verify settings and see how things are configured. We can also look at your error log for noise or wiring issues.

Thanks for the quick response and general guidance on providing more useful troubleshooting information. I hope this can move us further towards resolution.

Yes, bench tested. Heartbeat always present within a couple seconds of booting up. Attached one Dyn4 and 400W servo for bench test -- worked fine. I did not, however, attach multiple drives and servos before laying out and mounting in the enclosure.

Requested files attached.

Start-up sequence:
Turn on power to 24VDC supply, wait until Acorn heartbeat presents (drive indicators along header on Acorn board are GREEN at this time). Then turn on drives and servos through releasing E-Stop. Then boot up CNC12 on the PC. Release electronic E-stop in CNC12. Hit cycle start (drive indicators go RED but unconnected 4th axis remains GREEN). Now I can command the servos, but get the following results:

x-axis: turns forward, not reverse, drifts and "pops" lightly but worst of three axes
y-axis: turns forward, turns reverse (but slightly slower), drifts and "pops" but barely perceptible
z-axis: turns forward and reverse at same speed, least amount of drifting, but still present

I have the EMI filters for each drive CPU grounded (all grounds go back to a single ground bus fed direct from service entrance). Based on John Manura's build (basically the same build where he said he had to isolate the filters to get the servos to turn), I tested with these grounded and isolated and had no difference in result.

Could you post a copy of your drive settings in cnc 12? It has been a while since I did mine. I can compare it to my settings.

John

I have the line number set to 500 where you have 2048. Can you move in both directions using DDMDRV app?

The line number parameter is only relevant if you are using the encoder output on the drive and I don’t see anything plugged into your encoder output. For most systems this number is irrelevant.

Pics of CNC12 settings for configuration. Apologies, couldn't figure out how to get the new touchscreen to do print screen function.

Yes, in DMM DRV RS232 mode they move perfectly in both directions with no "popping" or drift.

Update: minor win achieved. Adjusted the power supply to provide 24VDC (it was a low -- 22-23VDC) and the termination for the red and brown wires coming off the DYN4 to Acorn DB25 lead and now have all three axes moving forward and reverse at the prescribed speed. X-axis no longer only goes forward; y-axis no longer has slight speed reduction in reverse compared to forward. Unfortunately drift/popping when not commanded is worse. Seems the increased voltage to those two (six) wires amplifies the drift/popping issue but corrects the inconsistency across the axes I was experiencing.

I'd suggest a re-tune of the motors with the new voltage.

Conducted a retune. During the retune, it was flawless -- no jittering. As soon as I updated the gains and went back to pulse and direction mode, the jitter came back on all three servos. Edited to add a scope readout of what's happening. First is current, second is position. Meanwhile, Acorn DRO indicates no movement whatsoever.