I have my k8200 now since a month up and running very well and I was surprised about the great printing quality.
During a print, the temperature sensor wire has lost connection and was connected to 15V. This destroyed the analog multiplexer on this pin. Because there is no other adc I had to replace the board. So I ordered a Mega 2560 and a RAMPS 1.4 and configured Marlin correctly, set the mainboard type to 33.
But the layers were extremely shifted.
So I searched this forum and did the following corrections/modifications:
-checked if the pulleys were fixed and aligned well with the belt.
-Exchainged the belt and pulleys with GT2 types
-I tried different tensions for the belts
-I varified that the schematic of the k8200 board and RAMPS are identical
-the endstops don’t produce random signals.
-exchanged the Pololus to make sure that they are working correctly
-resoldered some bad connections on the RAMPS
-i increased the reference voltage of the Pololus to 0.55V
-i tried different versions of Marlin and am now using v2
-i even checked if there is a difference between arduino 1.0.5 and 1.6.7
-In the Configure.h file i tested the following modifications:
–I lowered the acceleration to 300 mm/s²
–adjusted the steps per unit values because of the 16 tooth gt2 belt
–lowered the value for XYjerk to 10mm/s
Additionally i tried different print speeds.
None of these modifications did improve the print quality.
None of the prints was accetable.
Sometimes the error occures while printing the first layer and there is a shift of 1-2mm.
Often there is a continous shift of a few .1s mm every layer
And nearly every time there is a shift of 1-2mm in layer 6 (about 1.2mm at .2 mm layer height) and the pattern is always the same
I think I have tried everything I could imagine to eliminate the problem.
I don’t have any ideas anymore.
I would be glad if someone could give me some advice
I will post some pictures of the failed prints later on
I also tried different models (that i have already printed successfully), created with different 3d programs and sliced with cura and slic3r, without any difference.
I double checked, that none of the cables is causing problems (broken wiring/get stuck and cause lost of steps)
im by no means an expert, but it looks like you may need to re-tune the drive voltages. i had similar print issues in my first few prints (layer shift). if you remember on the original k8200 board, the small ceramic screw to adjust voltage to each motor. see section 2 - “connecting the printer” in the “using the printer” half of the manual, about 1/3 of the way down.
Velleman’s recommended voltages did not work for at least one of my motors, i had to do a little trial and error to get the motor to move properly without skipping.
i just started having a layer shift issue, but it was not continuous. i’ve had several successful builds up until this point.
For me, everything was going fine until about 1 hour into the print, (coincidentally as soon as i left the room for 2 minutes).
when i came back, my Y axis had shifted about 2mm, as if it had jumped a tooth on the belt. it was loose, so i shortened the belt by a tooth.
I started over, and again after an hour or so the same thing happened (and again in a 2 min span of me walking out). i had to pause the print and physically move the bed to get it back in position (close enough to finish).
I finally caught the malfunction in action, it seems that the motor stalled/ jumped a step.
I probably need to increase the voltage to the motor, but i don’t know why the issue happens after half way through the build. if the problem is due to mechanical stiffness, it would seem that the issue would appear throughout the build, and not just after a set amount of time.
Thanks for your advice O918,
but I figured out what the problem was:
I use a E3d Hotend that was mounted using a printed plastic mount found on thingiverse.com.
After a few prints the plastic mount was indented due to the force excerced by the two mounting screws.
This caused the Hotend to be loose enough to move a little bit.
I observed that this was happening everytime when there was a retract. So the motor retracted the filament and caused the offset.
The fact that the offset occured on many different models can be explained because these models all have to use retracts.
The slicer started to implement retracts at about the same height for all of the models because they all formed two independent parts from about 2mm height.
To solve the problem I used a washer, cutted and bent so that it would fit nicely around the upper end of the hotend.
Then I heated it up in a flame and melted it into the indented region of the mount.
Since this mod i did not experience any offset during print anymore.
Hy,
with E3D you can use lower retract : 1,5cm is ok with 3mm filament.
For the shifting, did you check that the pulley for GT2 belts are stongly locked (stop and go with motor movement) ? It occurs to me once, shifting was due to a pulley screw not well locked.
Hope it can help
