dustboy wrote: ↑Sat Dec 12, 2020 12:13 am
Still in the testing phase so it’s kind of a mess. Don’t judge.
EFC1720E-B10B-428C-9D6B-762204DF2723.jpeg
The issue is, your "testing phase" wiring may well be causing some of your problems...
If I were called to work on that, I would strip it out, and start from square one with getting Acorn on the bench and getting one motor turning, then connecting the other motors and getting them turning. Then I would wire the VFD/spindle properly and test.
You have a MAZE of what appear to be ground wires running all over.
I am only offering you my suggestion and opinion...I would start back at square one.
Marty
Reminder, for support please follow this post: viewtopic.php?f=20&t=383
We can't "SEE" what you see...
Mesa, AZ
I know it looks like a rat’s nest but it’s likely not the issue as I have a friend who is building an almost identical machine. Only difference is he started with a new cabinet and wired the thing from scratch on a bench like you suggest, his cabinet is so neat and organized you could run a comb through the wires. He has the same problem.
That maze of ground wires is all home runs from each component to the ground bus.
I now have the Meanwell PS hooked up 220V and eliminated any other 110V connections that were using ground as a neutral. No change.
I hooked my oscilloscope to the step output and I can see a ton of noise when I switch on the VFD.
The 1k pull-up resistors seem to be necessary, as I was unable to get the machine to function reliably (disgregarding VFD issue) without them. The poster in this thread https://www.cnczone.com/forums/centroid ... 4-cnc.html had luck with 10k resistors, not sure if that would make any difference.
I am working on the same conversion and I have the same problem. When I bench tested the unit, all worked fine, I had all of my wiring done so all that I had to do was pull the step and direction wires thru the case and boom, it should have worked. I did use longer wires to connect my Y and Z. My X works fine. My Y and Z are the axis that crap out when I turn on the spindle, just like Dustboy. This makes me think that I needed the 10k resistor trick. Well, I tried that and nothing. I am a tad frustrated to say the least. Attached is the picture of my controller for your enjoyment. Any and all help is greatly appreciated.
I would move the VFD as far away from the Acorn and logic supply as possible, and make the spindle motor power cable leave the cabinet as directly as possible from the bottom of the VFD. Outside the cabinet, keep the spindle motor power cable separated from any signal cables.
Do the step-and-direction signal cables to your Y and Z ClearPath motors travel through the same cable carriers that your spindle motor power cable does?
Yes, unfortunately there is only one cable carrier for each axis, so the VFD cables must travel within inches of the Y and Z control cables. This is very typical of CNC routers so it must be possible to achieve, if not a st/dir signal then there would be an encoder signal.
This guy shows some quantifiable EMI reduction using a line filter and ferrite rings.
I will move the vfd and all that apply out of the cabinet tomorrow. I guess what still throws me into a state of confusion is why does one servo work, granted it has the shortest run, and the others don't? Does the length of the controller wires make that big of a difference? Should the 10k ohm resistor trick works? What say you? Im planning on getting another vfd for my vacuum table so this will give me the opportunity to reimagine the “vfd case” that is in a separate area.
Had previously installed a shielded VFD cable from Automation Direct, grounded at both ends. Unfortunately it didn’t solve the problem. I then installed the filters on the VFD per the YouTube video in my post above. This made a big difference, the spindle will now run without causing an error with the servos. Also did some work on the wiring to physically separate signal and power.
I am still getting intermittent faults even without the spindle running. I have noticed that there is 0 ohms between the Acorn ground and the COM, it appears that there is no isolation between the two. In my experience in the A/V world, this is typically a cause of ground loop noise. Is it possible that is happening here?
