Help me troubleshoot if you would?
I have some strange behavior. As in the other thread, autobedlevel ended up at max Y trying to press Z down into the mud.
I recompiled the source (thanks Ant0ny) for v1.0.4, and tried it as 200x200x200. This kept the Z-axis probe over the glass, but when I went to print the "test_print_hollow.gcode" file, I got a series of adjacent scallops of filament prior to running off the back edge of the bed... all of this behavior is with the code that shipped and the code I downloaded/built for the 1.0.4 build.
It's as if it forgets to retract/move -Y, and instead moves +Y.
I've changed the bed dimensions back to 200x300x200.
So, I'd like to ask some basics because I am newb, and hope you can help. I start this test at (manually moving the axes) X: HIT, Y HIT, Z:100 (or so).
After running Home Axes from the LCD, where should the reported/actual position of the bed be?
(MIne are X:100; Y:105; Z:11)
After running G28 (I made a single command file and loaded it via SD-card. I need to get a laptop set up next to this machine), where should the reported/actual position of the bed be?
(After G28, then G29 mine are X: 100; Y: 105; Z:11 If I run it again, I'm at the back-third of the bed. This is not 'homing axes' behavior, I think)
Again, manually setting the X, Y to HIT, HIT, and starting bedlevel from LCD, where should the reported/actual position of the bed be?
(Mine are X:15, Y:205, Z:11.04. The probe seems to be X 14, Y 5; X 185 Y 5; X 15 Y 205; X 185, Y205. It seems like the bed is actually in the position reported at finish. If I rerun from here, the bed goes full +Y and I shut it off before I damage anything...).
And, G29 issued from the SD card file from the manually-set-at X HIT and Y HIT seems really weird - it runs off the back of the glass and leaves the bed fully extended. It's reporting that the printer is X 15; Y205; Z 4.04. Yet Y-axis is clearly full-extension, and the nozzle is below bedheight (springs compressed).
Now, when I went to Prepare->Move Axes to recover from this, I got jerky Y-axis retraction coupled with Z-axis (+Z motion)!
I don't know enough about how to dump printf/debug out. I'm willing to set up a laptop (Mint Linux) next to the printer. What should I do to look at debug output?
Since the behavior is so different time-to-time, I wonder if we have a memory corruption/access to y-position register issue?
Edit: is there anyway hardware or crosstalk - in stepper encoder, maybe - could be causing this? I'm skeptical as I've separated out the wire bundles; also the behavior seems like it's just forgetting to move the bed backwards in some cases.
I'm gratefully accepting advice. I can use the development tools, but have no familiarity with Marlin.