Dynomotion Forum

I have been building a small 3-axis cnc router with KFlop and KStep boards. The machine is functioning well except for a lack of stability where I will run a Gcode program and frequently have the "Read Failed - Auto Disconnect" error pop up throughout the machining.
When I click OK to clear the error the program will continue. This recurs at varying intervals, sometimes short, sometimes quite long but it can be half a dozen times during a cycle.
I am using a small water cooled spindle driven by a Sunfar VFD. Sometimes when I switch the pump on for water cooling it will trigger this error but not every time. However, the error occurs regardless when everything is running.

Any help with this would be much appreciated as I am using this project to decide whether I will upgrade a larger cnc to the KFlop board in the future. But I cannot do that until I can get these stability issues sorted out first.

Hi axeman,

That is usually an issue with grounding and noise. Are you using a good quality and short USB cable? The cable should have a good shield connection from shell to shell of less than 1 ohm. Ferrite bead/chokes can be helpful. Any noisy signals near the USB cable? Is the frame of the machine and Spindle Earth Grounded? Are there Ground loops?

Regards
TK

Thank you for the prompt reply.

There doesn't appear to be any noisy signals, unless the water pump is causing noise. As far as I am aware I have no earth loops. The machine frame and spindle are grounded. But I checked the USB cable and found it was over 10 Ohms! Yikes!
I found a USB cable (1.5m) with 0.5 Ohm resistance and put it into service. The machine was a lot better but still crashed a few times like before. I will obtain a shorter cable and ferrite choke and see if that helps.
I'll post back next week with the results.

Thanks again.

It has been a while but after thinking about noise I have done what you suggested Tom, and also investigated shielding of cables. To cut a long story short, I discovered that if I grounded the shielding of the spindle cable at the spindle, and not at the control box end, my issues were solved. Grounding it at the control box did not fix the problem at all.
This may be helpful for others if they encounter similar issues.
If anyone can explain why this should be the case I would be very interested to know.

Cheers

Hi axeman,

Thanks for taking the time to post back.

It isn't clear to me why that helped. Grounding is complex. It would be hard to speculate without a lot more details. Maybe it coupled all the Spindle noise into the Control box Grounds? Maybe somehow formed a Ground Loop? Dumps the Spindle noise to Earth Gnd more directly and away from contol signals?