1
General Mach Discussion / Re: MPG problem, mach freeze and mach crashes with art code 904
« on: January 06, 2012, 03:32:58 PM »
I now worked on both lathe and mill for the whole day with lots of manual jogging and did not encounter a single freeze - very strange...
i turned backlash off because it really caused some strange behaviour with the mpg. i mechanically adjusted my machines to have about 0,01mm backlash, that is OK for my application without compensation.
the screaming noise is just very short, maybe 50 milliseconds or so and the steppers do 2 full steps (20microsteps) for 0,01mm travel, so I hope that there is not one lost step included. it might be just the geckodrive g540 current reduction that causes that noise and i should not give too much attention to it. maybe the g540 is switching to reduced current and back to full current in the next moment as one step is needed because of the mpg being turned one step.
i do have lost steps when moving the 3d view of my gcode with screaming steppers - then I'm loosing millimeters up to centimeters! I got no solution for that so far but that's not my biggest problem at the moment, I just don't touch my computer as long as the machines are working
Still, maybe someone else has seen the problems with mpg and found a solution. If not, I will try to record the mpg signals with an oscilloscope and try to catch some glitch or something when the freeze occurs
Tom
i turned backlash off because it really caused some strange behaviour with the mpg. i mechanically adjusted my machines to have about 0,01mm backlash, that is OK for my application without compensation.
the screaming noise is just very short, maybe 50 milliseconds or so and the steppers do 2 full steps (20microsteps) for 0,01mm travel, so I hope that there is not one lost step included. it might be just the geckodrive g540 current reduction that causes that noise and i should not give too much attention to it. maybe the g540 is switching to reduced current and back to full current in the next moment as one step is needed because of the mpg being turned one step.
i do have lost steps when moving the 3d view of my gcode with screaming steppers - then I'm loosing millimeters up to centimeters! I got no solution for that so far but that's not my biggest problem at the moment, I just don't touch my computer as long as the machines are working
Still, maybe someone else has seen the problems with mpg and found a solution. If not, I will try to record the mpg signals with an oscilloscope and try to catch some glitch or something when the freeze occurs
Tom