New fully assembled i3 MK2 bed crash
So, I received my fully assembled i3 MK2 the other day. I opened the box and removed the top level of packing foam to see the 1kg of filament sitting directly on the bed, sliding around, with no packing around it. Now I didn't check the top foam that holds the top of the frame to see if it also held the 1kg filament box.
Anyway, cut a long story short, I followed the setting up guide, using the 'assembled' tree, which only required the Z axis to be calibrated (run Z into the top brackets until you get jumps from both sides).
So, it started on its merry way, position 1 through 3 fine. Then 4 I noticed it getting very very close, 5 and 6 the same. When it came to the position over by the 'printed hand' on the MK42 (top right), it crashed fully into the bed, flexing it down quite a lot, I would say 5-10mm. It did this a few times before i could scrabble for the off switch, hunting for god knows what. I turned it off rapidly. No error message no nothing obviously.
So after some chatting with support (Monza), tech didn't know why it happened but said to do an XYZ. Anyway, this failed, so I was then told to lower the sensor head. I lowered it by 1.5 turns, and did a recalibrate XYZ, this all worked fine.
Now I am running with a live z offset of 0.475 to get the correct profile for layer 1.
So, has anything likely to have been damaged during all this? and why would the fully assembled not just work out of the box? Perhaps that 1kg filament was smashing into everything, because it was just sat on the bed when I opened it up, and nothing around it at all. Why would the sensor need moving down? Why did it smash into the bed some 5-10mm when I would have thought -1mm beyond some reference point (ie point 1) and it should start to worry and abort.
I have printed successfully over the last couple of days, but when something costs this sort of money I left somewhat concerned by what happened.
Feedback would be most appreciated.
Richie.
Re: New fully assembled i3 MK2 bed crash
It's always possible the spool of filament knocked something out of alignment, or it may just be that the probe had worked lose in transit, these things do happen. The main damage that could be caused by a nozzle crash like that is to the heatbed and/or the PEI film on it, so if there are no obvious tears in the film and no cracks in the bed then you are most likely good to go.
What version of the Firmware did it come with?
It's always best to do a quick check even when you get a pre-assembled printer:
Re: New fully assembled i3 MK2 bed crash
It came with 3.0.8 firmware.
There didn't seem to be any damage to the bed as far as I could see, and all axis look correct.
As far as the checks go, having never owned one, and just following the instruction book, the bed crash happened. 6.3.1 calibration flow for the assembled unit doesn't mention checking axis or anything like that. I would assume it would come suitably packaged to survive the trip, otherwise there isn't much point in having it calibrated already. If it needed a full calibration it would have said so in the manual.
The main question, is how the thing managed to smash into the bed when it was ok on the front 3 measurements. This implies that the bed was slanted upwards at the back in relation to the z frame. Either that, or the firmware did something crazy, but not finding the reference point in that corner/and or the height being out, is not a reason to smash into the bed in the down detect movements. Thankfully it was not doing the X/Y just the up/down.
After the bed crash and the subsequent support conversation I did the full paper XYZ calib, and had to abort it as the sheet of paper starting moving, the head was hitting it. At this point the sensor was adjusted down. How does a threadlocked tight nutted sensor move upwards during transit?
Is there any way to see the values stored for each reference point?
Also, as an aside, the live Z adjust is a bit of a strange one. It seems to count the number of rotations of the dial and apply them no matter what. It is probably an idea to mention this in the manual or change the firmware so the moves are not 'queued', otherwise you can end up just spinning the dial, thinking it will stop when dial is stopped, only to find that you get 400 odd Z limit bump stops at the top of movement.
Richie.
Re: New fully assembled i3 MK2 bed crash
Richard,
we are sorry you had such a trouble with our pre-built printer. It really seems something shifted in the printer during the transit. Unfortunately the firmware has no way to find out.
> Also, as an aside, the live Z adjust is a bit of a strange one. It seems to count the number of rotations of the dial and apply them no matter what. It is probably an idea to mention this in the manual or change the firmware so the moves are not 'queued', otherwise you can end up just spinning the dial, thinking it will stop when dial is stopped, only to find that you get 400 odd Z limit bump stops at the top of movement.
The live adjust value is applied immediately, but quite slowly.
Vojtech