Calibration Problem
Hello All, Prusa fan here!
I am experiencing X|Y|Z calibration (mainly X-axis)/ Auto Homing / Z-Height measure problems following an X-Axis crash report. Below is a description of events, problems and attempted solutions to return the machine back to working beautifully normal. Any help is most appreciated.
Date: 2-13-2019
Machine: PRUSA i3 MK3
Firmware: 3.5.1-1778
History: Machine has been an extraordinary performer, a masterpiece of totally impeccable brilliance for a total combined run time of 12 days according to the statistics.
Event Marking Change In Performance: Machine recorded an X-Axis “crash” (2/11/13) during two print efforts of the exact same part. No evidence of part interference from build plate up.
Current State:
Self Test X-Axis length error.
During X|Y|Z Calibration Test reports Z failure.
Machine will not return to exact X=0, Y=0 Factory Home Position.
Machine does not maintain X|Y|Z calibration settings between prints.
Self Test X will pass so long as the print head is positioned in the center of guide rod travel.
During Auto- Home sequence, the machine is off in the X position as it relates to the PINDA probe and circle location on the board. Approximate distance shifted right from intended location is ½ an inch (12.7mm).
During the PINDA TEST:
When following the spatula procedure, the PINDA TEST is a pass in accordance to:
(Z0=LED on) without spatula/metal object present
(Z1=LED off) with spatula/metal object present
Attempted Solutions:
Checked all wire routes, connections, pulleys, wire tie interferences, manually traveled all axes by hand and by user interface with no apparent travel limit problems.
Reset PINDA height in accordance to Zip Tie height offset procedure. Re-Ran Live Z Tune. Saved new calibration settings.
Turned machine On/Off, Unplugged/Replugged.
Ran the Set Up Wizard 7 times with different results.
Sometimes it would pass X|Y|Z checks and then make it all the way to Live Tune Z then fail 9 point bed inspection prior to print with a collision midway thru bed level check. (*In order to pass, print carriage has to be centered on X-Axis guide rails after Auto-Homing.)
Other times it would fail X-Axis check by reporting X-Axis Length error.
Ran Factory Reset. No return to normal.
Results: Unable to repeat a run without recalibrating X|Y|Z which usually fail X-Axis Length or Z-height offset measure.
In search of help. I am lost in finding a working solution/repair. Thank you for any and all suggestions, feedback, help and most of all, your time.
Respectfully,
Big Fan/Admirer Wes
Re: Calibration Problem
I'll start with "I am NOT an expert" ... and with the first error in the list.
Current State:
Self Test X-Axis length error.
This hints at there being an extruder side to side movement issue. Something is limiting how far the extruder can move. The extruder plastic should be able to contact the Z-Axis mounts when moved to either side. Did you add belt tension springs?
A common problem is the hotend fan is not installed right. The fan wire should exit the fan on TOP, not on the side, towards the back of the extruder. If you assembled it with wires out the side, the wires interfere with X-Axis range.
This manual image shows the wire "looking" like it comes off the side. It doesn't. It comes off the top into the wire channel (image 2).
Re: Calibration Problem
Thank you for posting Tim.
I will double check these suggestions. What baffles me though is that from day one, the self test, live z and auto home worked perfectly and as such, each of the prints were beautifully executed piece parts. So far, the machine totals 12 days worth of run time with no real run mysteries during that time. It is a remarkable marvel which I fully admire. I am just baffled in my search for "cause" in an effort to remedy with a solution.
I will report back whether or not I am able to return to normal.
Thanks for your time,
Wes
Re: Calibration Problem
During Auto- Home sequence, the machine is off in the X position as it relates to the PINDA probe and circle location on the board. Approximate distance shifted right from intended location is ½ an inch (12.7mm).
Auto-home doesn't use the embedded calibration points. Instead, the firmware slams the extruder into the X and Y stops, and uses the PINDA to find the Z limit.
The bed cal points are used only once, during calibration, the X and Y distances measured (in motor steps) are then constants stored in NVRAM; they get recalled from memory at each printer startup.
Re: Calibration Problem
I am just baffled in my search for "cause" in an effort to remedy with a solution.
Have you checked the set screws on the X-stepper gear?
Re: Calibration Problem
Hello Tim and Marco,
Thank you for you replies. I appreciate your help. I removed the X-axis stepper motor. I then cycled the stepper. I assembled the stepper back to its mounting position. I re-ran setup and cycled thru machine test as if checking operations first time following build completion. The machine cycled through all tests with passing results. At the moment, I appears to be operating as intended.
My best guess for the X-Axis overrun failure was due to running the largest build envelope part yet. It fit the bed in SLIC3R at a 45 degree angle and the G-code posted successfully. I did notice at the beginning of the print that the part was skewed/offset slightly to the right (X-Positive) though it was "arranged" perfectly centered in SLIC3R. When running the "brim" portion of the print however, it appeared to just narrowly exceed the bed width. While there did not appear to be an obvious interference during the beginning portion of the print, my best guess is that this is what eventually lead to the "Crash Detected" Print Stopped culprit that when trying to recalibrate revealed that the machines' coordinates were clearly off.
Upon further review, the way that the zip tie "lugs" for lack of better description were oriented on the printer head in a way that they could have ever so slightly interfered with the power supply. I repositioned the "lugs" to point in an up-and-down position over a side-to-side position. So far I have run two successful prints since the crash occurrence.
Thanks for you help Prusa family.
Wes