I have a standard model CRP thats 24x24 running the DIY nema 23 setup, so gecko 540 and a ESS. Over the last week, or about 4 hours of actual use I have had one of the slave axes malfunction significantly 3 separate times after hundreds of hours cutting/drilling these same parts with no errors at all.
The first one was while I was drilling the mounting holes in my parts, the machine was a little over halfway through its cycle and all of a sudden starting making a grinding type noise like it was missing a ton of steps, like it would make if you ran into the stops and just kept trying to drive through them. But it was still moving. So i stopped the cycle and quickly determined that it was the x axis, the one with two motors, that was having the issue. I figured one motor had lost some steps and the gantry had gotten enough out of square to cause binding issues. So i re zeroed it against my homing switches this had no effect. I unplugged the master motor, this made it a little better but it still sounded liek it was binding up pretty badly. i replugged the master and unplugged the slave, and the binding went away and the master could drag the gantry around pretty well. It was definitely not staying square and would chug occasionally but nothing like before. I plugged the slave back in and it worked fine. re zeroed and finished my work for the day, about an hour and a half of drilling.
The next time was maybe 3 hours of machine time later that week, during a cutting operation when I pulled the material off my vacuum table some of the parts were out of square and you could tell that at some point either the master motor had gained steps or the slave lost them, a significant amount too the gantry was over a 1/4 inch out of square, but no binding. And no need to unplug the motors to get it to reset, just re zeroed and finished my work, about an hour.
Third time it happened today, a few days and a couple of hour of work later, similar to the second time but this time the gantry was out of square in the opposite direction, so either the master had lost steps or the slave gained some.
I'm not getting any error lights on the gecko or the ESS, and nothing in the status window in Mach 3. At this point I'm not sure where to go to determine if it's a motor going bad, or a cable, or one of the drivers in my gecko, or something even further up the line.
I thought about swapping the cables on the x axis motors, since that would be easy, but given the behavior I don't think that would give me a definitive answer even if it was a cable. Can't swap other cables since the x axis cable arent long enough to reach the Y or the z motors.
What was the thought process that led you to the electronics? Not to say that you are wrong, I just noticed that you didn't mention a potential mechanical issue. Although I haven't seen the standard version in action, it could a small error or glitch in the linear motion mechanism (loose set screw or belt misaligned, etc) that is adding up. If during your 4 hours of cutting the gantry is moving back and forth a lot, then the error could be cumulative and significant. That is to say, the very small error, say .0001, is allowed to creep in hundreds or thousands of times, that could equal the .25 of error on only one side. The issue may always be there, and if you home/zero your machine every time (like we are supposed to) it could disguise it self during the shorter runs and only present itself as noticeable after a couple of hours of running.
Just a thought, I could be completely wrong. Something to check before you start replacing the expensive stuff though. If I remember correctly, there was an issue early on with the CRP gear train and the set screws backing out on the main drive gears on the stepper motors.
I'd be curious to hear how this turns out for you.
-Mike
I mostly ruled out mechanical because with the slave motor disconnected it moved smoothly, and only happened randomly. Also it wasn't something that was adding up over time, it was a sudden error. The parts I cut only take about a minute 45 to run each cycle (9 copies of the same part per cycle). I just run a crap ton of cycles in a sitting and the error occur during one copy on a cycle, throwing the rest of the parts on that cycle off if I didn't catch it.
99% sure I'm losing a driver in my g540, it started doing the two motors fighting each other again today. This time it didn't just go away when I unplugged the slave so I was able to experiment some. I swapped the x and a motors in software and swapped the cables around and the problem was now happening on the X instead of the A(slave). When I was swapping them back around I noticed that with the A axis plugged into the G540 and screwed down tight if I bumped the connection the motor would jiggle. Kinda like it does when you first turn the system on and the motors start holding.
I went ahead and upgraded to the leadshine mx4660, I liked the idea of less motor heating, and smother running, plus no need to transfer my heatsink and fan combo over. If the gecko is still under warranty I'll set it on the shelf so I won't have to scramble for a replacement part next time.
Help with troubleshooting motor/driver issue
Reply with Quote
