[MadTooler]

I am not sure if this is a symptom, problem, or normal and I simply didn't notice it before. While running gcode in KmotionCNC, the DRO is occasionally changing color to yellow then back to the normal enabled blue.

Gantry router with Kflop, SnapAmp, Kanalog, DC brushed servos, PC router, USB isolater, WinXP, and Kmotion/KmotionCNC v. 4.33.

Any thoughts if bad or normal?

Thanks.

[TomKerekes]

Hi MadTooler,

If the Status/DRO update has difficulty getting access to KFLOP for an extended period it gives up and displays the DRO's yellow as if disconnected. This shouldn't directly effect machine operation but is an indication something isn't normal like excessive USB congestion.

You might try our later code V4.34 including updating the KMotion/KFLOP USB Driver in Windows Device Manager. There have been some improvements.

Other FTDI USB Devices on the PC might be causing a conflict. Has anything been added recently?

Trajectory Planner Setting of Facet Angle less than 2 degrees may cause excessive motion vector data to be sent to KFLOP.

GCode data with an excessive number of tiny vectors, without a large enough Collinear Tolerance to allow the Trajectory Planner to combine/filter vectors, can also cause excessive communication to KFLOP.

Regards

[MadTooler]

Thanks for the quick reply, Tom.

I should clarify the program isn't usually halted when it flashes to yellow. Normally, it is a very quick flash of yellow then back to blue like nothing has happened, but it may do it several times during the run. I do recall there have been a couple of recent programs run where the DRO was yellow and the control was in the disabled state after the program was completed, but I wasn't sure if I had accidentally hit the e-stop button on my joystick pendant when trying to jog out of the way.

No new hardware changes, but a while ago I had tried pushing the limits of the trajectory variables. I make a lot of parts with very tiny and important details where shaving 0.001" can make things disappear. I did push it so far the data became massive and overwhelmed the control. It also made the PC very sensitive where any other simple tasks done while running, including moving around in the gcode viewer, would result in program failure. As far as if the flashes were occurring before these changes, I am not sure.

These are the current values that were backed off from the failing extremes, and yes I know they are still pretty aggressive:

Break Angle : 45
Look ahead : 5
Collinear Tolerance : 0.001
Arcs to Segments : Yes
Corner Tolerance : 0.001
Facet Angle : 1

I am also displaying the encoder values to the DRO.

I am anxious to try v4.34 after I finish running my current jobs. Glad to hear it is released!

Thanks again.