My printer has suddenly started acting up. At random points in a print it looses registration on the x-axis (as seen in the image).
I’ve never had this issue before - until it happened 5 hours into a 7.5 hour print of this part. Now it happens every time during the print of the first layer - although at a different place each time.
The first thing I checked was the timing belts, but they are tight and not worn. To see if I could reproduce the error I wrote a simple g-code file that moves the x-axis back and fort increasing the speed incrementally up to the max. Even running this repeatedly I could not get it to miss registration.
Normally I use a Raspberry Pi with OctoPrint as a print-server, but I’ve also tried using a Mac with the latest Pronterface and get the same result.
The printer is controlled by a RAMPS board with Marlin firmware and has the following settings (these are from after I tuned things down quite a bit speed wise - to see if the mass of the carriage was the reson it failed on rapid changes in speed/direction): #define BAUDRATE 115200 #define DEFAULT_AXIS_STEPS_PER_UNIT {106.76, 106.76, 800, 52.45} #define DEFAULT_MAX_FEEDRATE {215, 215, 5, 25} #define DEFAULT_MAX_ACCELERATION {500,500,100,10000} #define DEFAULT_ACCELERATION 2000 #define DEFAULT_RETRACT_ACCELERATION 2000 #define DEFAULT_XYJERK 5.0 #define DEFAULT_ZJERK 0.4 #define DEFAULT_EJERK 5.0
The part was sliced in the latest stable release of Slic3r, but I’ve tried a bunch of different settings to see if anything would change - to no avail.
I’m at a complete loss of what to try next so any suggestion will be welcome! Is it possible that the RAMPS board or the Polulu drivers can be damaged somehow?
Have you consider that a chip on the printer board may be getting to hat and changing electrical values. You can try pumping a lot more air in around the printer board and add a few finned heatsinks to larger chips. It could require a 7 hour or longer test to verify if this is the problem.
So there are lots of things that are possible, the trick is figuring out which one. There are two ways to debug this, one is from first principles, the other deduction.
Using first principles you would write some code that would talk to the RAMPS board directly and tell it to move the X axis by a fixed amount, then pause, and move Y, then move, then pause and move Y back and forth, etc. Then attach a sharpie pen to the hot end and tape a piece of paper to the build bed and get a reproducable X failure. Using that failure use an oscilloscope to watch the signals to the stepper and verify 1) that the correct # of steps are being sent, 2) that the stepper is seeing voltage for every step, 3) that the gear on the end of the stepper is turning the right number of degrees for each of those steps, and that the belt is moving things the right distance. One of those will not be true.
Using deduction, you could swap the Polou driver for the X and Y steppers and see if the problem ‘moves’ to the Y direction. If so replace the Polou board. If it stays, return the drivers and swap out the X stepper motor with a replacement motor. My guess is that one of those two things will fix it if you have confirmed the mechanicals between the stepper and the carraige (timing belt) are sound.
@Joseph_Prugh I’ve considere this, since it’s been a bit hotter here than usual, but since it’s manifesting in the first layer of a print, even after the printer’s been off over night I’ve written it off so far. I’ll try to attach small heat sinks just to be sure.
@Chuck_McManis My thinking exactly - I just ran out of ideas of what to try next.
I’ve been using tape marks instead of paper/sharpie, but that’s a great suggestion that I’ll try out for sure. I don’t have access to a oscilloscope, but hopefully some of the other suggestions will pan out.
Swapping the polulu drivers I will definitely try out if I can manage to faithfully reproduce the error (without having to do an actual print), something I haven’t had much luck doing so far.
@Evan_Gillespie Quite a while actually - thanks for the reminder. Both X and Y move without much force needed, but I’ll try that as well
I have a flashforge creator x it stopped in mid print randomly. Some times the stepper would vibrate trying to find zero or something. I updated the firmware and all my issues went away.
@wesman_Wesman I’m at a pretty recent version of Marlin and have been reluctant to upgrade since has been behaving well (don’t fix what’s not broken - and all that), but I’ll try that as soon as I have ruled out the mechanics and pololu temeratures.
After no less than 15 failed prints (thankfully which all failed during the first layer, but never at the same spot) I now have a working printer again.
There was nothing wrong with the belts, steppers, linear bearings or firmware - although I did upgrade my firmware, rechecked the e steps, tune my Pololu drivers and add heat sinks / extra fans and oiled the rods.
Slic3r defaults to 200% width on the extrusion on the first layer - which in my case made the edges pucker up. When the print head next came back to lay the next line it would hit the rigid edge and the whole print head would shift. I thought it was mounted tightly, but the forces here are quite high and since nothing else could shift the weakest spot had to go.
Dialing down this setting to 100% and lowering the extruder temperature to 180 degrees (down from 185) fixed it completely.
I think I’ll print a new holder for the E3Dv6 head next
