You could try doing a usb to rs485 convertor then getting an arduino or modifying the existing one to accept an rs485 connection. You have to figure out first if its a power spike/loss resetting the cpu of the control board, the host controller getting zapped, or the data just getting garbled enough it looses connection.ĭoes the host usb device disappear and re-appear?

Usb isn't fully differential signaling so it's not a great choice for noisy machine control, it needs the ground reference to work correctly and it also doesn't handle corruption well. so improving that with shielding and isolation is what you need too do. So you either have a very noisy spindle that is upsetting the power, or signal integrity. :::EDIT::: I also have this problem with two different USB cables, one of which has a ferrite cylinder on both sides of the cable. What can I try now? I'm pulling my hair out! But if that spindle is running, I have only a few seconds before I get the error or lose communication. I've even tried grounding the aluminum frame to the casing and therefore the Arduino. Without the spindle running it makes all the way through the 20min gcode file. I have even tried running the spindle without a bit and this error still happens within a few seconds. But I have tried this with and without those two connected while the spindle is running with no change. I generally leave it always connected, because otherwise I forget and crash bits. I tried hooking my tiny kit Oscilloscope to the shared ground at the spindle casing, but didn't see any jump in EMI on the 1V/div scale.Īs a note I usually run with my ground connected to the spindle casing and A5 connected to a copper clad for probing height maps. I have the wires going from the power supply plug to the spindle twisted, (thinking it could be EMI, but DC? hmmm).

The spindle is running from a completely separate 24V power supply too. I'm running it now without the spindle (or a bit of course) and it is running just fine. I can't run my spindle for more than a few seconds before I get this error or lose communication. I'm running gcode generated using Fusion 360's grbl post process options. So my computer shouldn't be powering it, just sending messages. I have a switching regulator converting the 24V power supply down to 5V, and feeding into the Arduino.

(I recently fried the DeveloPower V2.0 it came with, my own fault). I'm running Grbl 1.1h, on an Arduino Uno, with the CNC shield, on my 1610 I've had for several years. The machine will continue to run until it gets to the end of what was last sent, but the communication is broken. I either get this error, or my computer simulation in Candle just freezes.

How to sync port cnc 3018 gbrl candle