Very Exciting Boring Machine Conversion

[dustboy]

I'm totally stumped here guys. Still getting "step timing error" from any of the 3 servos intermittently. VFD is completely powered off. I don't think the problem is with the Clearpaths. Do other servos have a problem with the ST/DIR signals?

Switched from 1k to 10k pull-up resistors on all ST/DIR outputs. Gave the Acorn board a direct ground to the bus bar. 220VAC supply to the Meanwell.

Revised As-built attached. Can an engineer from Centroid please help?

SPEEDY AS BUILT.pdf

(3.48 MiB) Downloaded 196 times

[martyscncgarage]

I don't have much to add, other than try and run the frequency setting down in the Advanced Tab as a test. Not sure where it is set to now.
I assume you guys with Clearpaths are using their cables? I assume they are shielded and you have the shield grounded at the Acorn end?

I have never really used clear paths, but a friend has them on a lathe and they seem to work ok. Many people have used them successfully with Acorn.
CNC4PC now has an interface board for Acorn that allows directly plugging in the other end of the CP cable to it.

Hope you guys find out what the issue is...and post the solution. Have you been back in touch with Teknic about the issue?

Marty

[dustboy]

Thanks Marty. I’ve tried running the frequency at all 3 settings. The most frustrating part is that after over 2 years of building this machine, I’m just one little step from making chips. The Clearpaths are really a great product, and so is the Acorn. I will say that the Clearpath cables are nothing special, no twisted pair or shielding and way overpriced. Which is why I kept the existing shielded wire in the machine, now I’m rethinking that plan. Teknic is out of ideas, they have been helpful but if the signal reaching their product is bad there’s only so much they can do.

Here are a couple things I plan to try:
1. Run stranded CAT6 for the signals, and preserve +/- pairing throughout the run. Right now there is a single conductor that carries 24v+ to the servo for ST/DIR/EN.
2. Mount Acorn to the inside of the cabinet door for easier access and simpler wire routing away from noise sources.
3. Thought about trying the DB25 5v push/pull logic outputs. Moving to a lower voltage only puts the signal closer to the noise floor.

That CNC4PC board is interesting.

[martyscncgarage]

https://cnc4pc.com/c86accp-clearpath-co ... oller.html

Not sure if it will make a difference but it might....

[dustboy]

I think I found the solution...replaced the machine’s original signal wiring with stranded CAT6, maintaining the +/- signal pairs within each twisted pair. The Centroid schematic appears to show a single 24v conductor running to each servo, then splitting to A+, B+ and EN+ inputs. This doesn’t take advantage of the EMI suppressing properties of a twisted pair cable. Furthermore, Teknic’s own signal cables are to be avoided as they are not twisted pair nor are they shielded.

I ran the machine with spindle powered for a total of about 30 minutes with no faults. FWIW the pull-up resistors are also not installed.

[jasonGO]

Dustboy- just confirming you followed the schematic as shown previously (removing the orange lines where you had the resistors) and used cat6.
I am having the same issues you are describing after going from a router to a spindle- I have shielded data cable going to my clear paths but NOT twisted pairs and shielded.
I might try to upgrade to the twisted pair shielded cat6!
This sounds like it remedied your headaches.

Before I do I was unclear if you had issues when you had the resistors attached and if so why go to the cat6 data wires?

Cheers!

[dustboy]

I was still having trouble even with the resistors, and I am not using them now. I had shielded cable before but had only a single 24v+ wire to the servo that was jumpered to the other + positions on the servo 8-pin connector.

[jasonGO]

That is my set up now except I am not jumping at the 8 pin connector, I have them all tied together at a bus bar located in the enclosure.

Looking into stranded Cat6.
Thank you.

[carboncymbal]

I think I found the solution...replaced the machine’s original signal wiring with stranded CAT6, maintaining the +/- signal pairs within each twisted pair. The Centroid schematic appears to show a single 24v conductor running to each servo, then splitting to A+, B+ and EN+ inputs. This doesn’t take advantage of the EMI suppressing properties of a twisted pair cable. Furthermore, Teknic’s own signal cables are to be avoided as they are not twisted pair nor are they shielded.

I ran the machine with spindle powered for a total of about 30 minutes with no faults. FWIW the pull-up resistors are also not installed.

I am having the same clear path faults on my. Is your your twisted pair solution still working for you? Have you had to make any other changes?

[Krlerma]

The twisted pair idea worked well. Also make sure u ground your shielding.