Print head getting dropped in middle of print  

RE: Print head getting dropped in middle of print

We just had this happen too. Any ideas on how to make this not happen?

RE:

It looks to me that the machine went out of register in the X axis. How it could do that without sensing a crash from the load cell sensor or a motor stall is a real mystery. I note that there’s quite a bit of excess material toward the front of the print. No crash?

Note that sensing and reporting all forms of crashing and stalls are not even half baked in yet in the FW.

RE: Print head getting dropped in middle of print

I had this happen once. The print was well adhered to the build plate, so much so that when the Nextruder bottomed it actually broke the print in two (and the bottom remained firmly adhered to the plate). 

What I did next was a full factory reset, FW reflashing and re-calibration both before and after firmware flashes. The problem has not occurred again after 50 or 60 large print jobs.

I also noticed that the latest software/firmware seems to have implemented Z movement between layers so there is little chance of the Nextruder hitting the model during non-print movements.

RE: Print head getting dropped in middle of print

FYI, the Z axis movement is a setting in the Prusaslicer preset not the firmware. Z Lift is typically referred to as Z hopping on retractions in non Prusa contexts. All multi materials printing typically needs z hopping or the nozzle can easily crash into another tools printed toolpath. Why Prusa didn’t have that set high enough is just another example of what we have seen a lot of with PS since the release of the XL.

RE: Print head getting dropped in middle of print

I haven't had the drop happen again, but I have had the X-Y position shift now on multiple prints. I thought this might have been related to running from Octoprint, but I've had 3 failures now running off the USB thumb drive. I have attempted to email Prusa support about it as I haven't had time to sit and wait in the queue for the support chat, but 4 days later I still haven't heard anything back from them.

RE: Print head getting dropped in middle of print

At one point I was having some oddness with sending gcode/bgcode to the XL. I bought a couple of these as they don't stick out very far (that might impinge on opening the EnXLosure doors). I haven't had any issues since then...but unable to prove a cause/effect. I needed to reformat as FAT (and lost partition space, but can reuse for other purposes later and they were cheap). https://www.amazon.com/dp/B07D7PDLXC

RE:

This collision / layer shift issues seems to only happen on prints that occupy space over 200mm x 200mm on the print bed and are not vase mode prints. I've done a couple of 260mm x 120mm, 276mmx108mm, and 220mmx220m (last one in vase mode) without issues, but prints occupying 200x200 or more on the bed, I'm sitting at about 8 failures for 10 prints, all the failures have been this layer shift issue. I'm running the latest firmware which my understanding was had collision detection turned off because of problems they were having with it, but I'd already disabled it on the printer because of other posts I'd read about it causing issues with print quality. On this print because of the corners in the print it was very easy to measure the shift in X and Y which was 22mm in both directions. Interesting on a Core X-Y machine since an equal shift in X and Y could be caused by one of the 2 steppers not moving when trying to make a move in either X or Y. That extra material you're seeing on the front is because of the layer shift, it wouldn't have caused it, it's a result of it.

RE: Print head getting dropped in middle of print

Weird. Have you tried:

1. Use original Prusa XL thumb drive that is erased to FAT32
2. Download latest firmware
3. Reset XL to factory settings
4. Re-apply latest firmware
5. Reset XL to factory settings
6. Recalibrate all modules and settings

When I was having problems this seemed to have fixed them (head bottoming out, weird display issues)

RE: Print head getting dropped in middle of print

Hi, just happened to me last night on a big print for the first time. It was close to the right corner. 

I could make use of the cancel part feature and continue.

I was not sure, but from what I can read here it seems it happens on big prints only? 

RE: Print head getting dropped in middle of print

I'm only seeing it on large prints, I have 2 different large prints it's failing on pretty consistently. Prusa Support thinks it's a belt tension issue (I don't agree) but I did miss a step when adjusting belt tension where you check the alignment (I thought you just had to make sure the tension was equal on both belts), and so I need to go through the process again. Because of the large prints though I've run through a bunch of filament and had to get some more ordered so I can do more test prints.

RE:

After reading this thread, I looked at the tool change system and why it might drop a tool. Since the tool is held in place by two sliding unpowered catches held in place by nothing but friction, it's very likely that the tool holder itself is out of spec. It could be that when the tool is held, there's not enough friction to hold the catch in place. When the machine whirrs around and around, the catch starts to slip and then plop. The tool falls off.

If the tool holder has enough friction to overcome any of the motions the machine will generate, then the tool will never fall off. 

Somewhere there should be a specification as to how much force it takes to pull the catch back. The machine itself should be able to capture that number and either use it to hold the maximum acceleration to a value less than it will take to overcome that friction, or throw an error to the operator to indicate the tool holder is below specification.

One possible mitigation would be to change the tool every so often. I'm sure there's a way for the gcode to send the tool back to the dock and then pick it up again. That would reset the catches and prevent the tool from being dropped.

RE: Print head getting dropped in middle of print

So I've been having a problem with these large prints where the X-Y system loose steps (it a layer shift) I'm still not sure why, but I've seen shift as large as 21x21mm. When the X-Y location is off that much, when a move take the head close to the right side of the machine, if it's shifted enough it will push the locks against the right endstop and disengage them. It's out of frame on this video, but it's definitely the cause of the issue. You can verify this by taking a docked tool head and then just pushing the head all the way to the right, you can undock it when it hits the end of travel. It would have been nice if Prusa had designed this so it wasn't possible to undock the tool head by not allowing the latches to make contact with that side of the printer, there might be a way to do it without losing print volume but so far the only ideas I've come up with basically take 9mm off the right side of the build volume.

Posted by: @eastmemphis

After reading this thread, I looked at the tool change system and why it might drop a tool. Since the tool is held in place by two sliding catches, it's very likely that the muti-tool holder is out of spec. It could be that when the tool is held, there's not enough friction to hold the catch in place. When the machine whirrs around and around, the catch starts to slip and then plop. The tool falls off.

If the tool holder has enough friction to overcome any of the motions the machine will generate, then the tool will never fall off. 

Somewhere there should be a specification as to how much force it takes to pull the catch back. The machine itself should be able to capture that number and either use it to hold the maximum acceleration to a value less than it will take to overcome that friction, or throw an error to the operator to indicate the tool holder is below specification.

RE: Print head getting dropped in middle of print

Hello 

I have been printing long multi material parts that are placed along on the right side boundary. I too have had the printhead disengage from the dock and fall into the print causing a crash. I agree with your hypothesis that the tool head is getting unlocked when it transitions between heads and it moves to the far righthand side of the printer causing the dock to become unlock. I will be studying how the printhead moves. Also I am wondering about the Travel Lift Section of PrusaSlicer and how effects this scenario. 

To continue my print jobs, I will be offsetting my parts 10mm from the right side of the print bed.

RE:

I was having this issue a lot while printing parts for my enclosure, these parts are very long so have a tendency to pull-up off the bed at the edges.  I worked with support to resolve it. Most likely issue was the parts were pulling up just enough to snag the tool head and cause crashing and tool head falling off.  Look very close at you parts for pull-up at the corners.  I never would have thought that was the issue because at first look they looked prefect to me but the parts had a lot of warping.  I’m printing in PETG. Per the suggestion by support I also updated my firmware to 6.0. At one one point I covered the printer to keep some heat it so the printer was not exposed to the cold. Problem solved!  Now I have my enclosure built and I’m printing parts for a filament dry box. These parts are not as long but I have not had the same issue since these changes.  A stable environment to keep parts attached to the bed is key! Another big one is making sure your steel sheet is perfectly clean. Hot soapy water and scrub if you start seeing 1st layer issues, this is another one I never would have believed BUT it works! IPA before every print, and don’t touch your build plate with hands or fingers.  My 2 cents. 

RE: Print head getting dropped in middle of print

Hello Moto Skwid

As per Prusa's recommendation, I updated the Firmware to 6.0.0 alpha. I did run into the Dock position calibration issue. So downgraded to the previous version, completed all the calibration, then reflashed to Firmware 6.0.0 alpha and verified all calibration remained (it did).

I also offset my print 10mm from the right side and about 8mm from the back side. 

Knock on wood.... so far I have had no issues. Printed over night making TPU seals with PLA supports. 

The Prusa Help Forums are golden. I am very appreciative of everyone's knowledge and determination. 

-Spoon

RE:

I also had the same printing long enclosure parts. Far right hand corner and the print head just fell off (twice now). I was able to pick it up and stick it in its dock and resume. I will try the 6.0 alpha to see if helps.

RE: Print head getting dropped in middle of print

Well I can tell you what pulled my tool head off the carrier... a filament tangle on the filament roll.  I did 3 prints in one day, well 5 if you count the failed prints; but anyway the day proceeded as follows.  First print printed fine.  During the second print the tool head disconnected from the carrier but the XL did stop the print and started beeping.  I canceled the print and reattached the tool head to the dock, pulled the failed print off the bed and hit reprint and the "third" print worked fine.  Started a fourth print which failed just like the second.  Followed the same procedure described above and the last print finished fine.  Of course I was mucking about with the printer each time the print failed but ultimately failed to determine a problem with the printer so after I failed to find a problem I just tried reprinting and for the most part it worked.  Went to bed with the problem still unresolved and when I woke up the next morning the first thought in my head was "IDIOT, check the filament spool".  Sure enough I found that the filament had a tangle, one crossed filament on the roll.  Probably not the problem you're contending with but just wanted to remind everyone to check the simple things first, something I failed to do here.  Also it might just be my imagination but while watching your video I believe the tool head on your machine seems to wobble a lot more than mine does, almost looks like the magnets don't have a firm attachment between the tool head and the carrier.

RE:

Sure enough I found that the filament had a tangle, one crossed filament on the roll.

I'm not seeing how a tangle could cause the tool to fall off the holder mid-stream. The tool changer mechanism appears to use a friction fit for the locking pins. If that friction is too low, then the head can fall off. I think that's the basic issue with that failure.

You can probably see an impending issue by looking closely at the locking pins when printing. Are they all the way home or are they drifting out of position? They shouldn't move at all and should homed.

In my experience, tangles, where the strand feeds from under another layer, are caused by losing track of the filament end. If the filament is allowed to get loose, and unwraps around the spool, then it's very likely that the end of the filament ducked under another wrap when it came loose. Keep track of the end and you'll never have a tangle like that.

RE: Print head getting dropped in middle of print

I went looking through the XL manual and I can see I've been using some incorrect terminology so I'll try to use the Prusa designations; at least when I remember what they're named.  Only I couldn't find what those "locking bars" that connect the Nextruder to the tool-changer are named so I'll just call them locking bars. 

EastMemphis, I see what you mean. I couldn't pull the Nextruder off the tool-changer when I pulled on the filament (I was trying to simulate a filament tangle).  However when I slid one of those "lock bar" to the left, it just fell right off.  There is absolutely no way the Nextruder is coming off the tool-changer when those bars are engaged.  And it takes a good amount of finger pressure to slide those locking bars to the left.  I even tried pulling on the filament (again trying to simulate a filament snag) while trying to push one of the lock bars but it still felt tight and didn't want to move.

It actually feels like there's a ball bearing locking thing going on in there.  To get the locking bar to move initially takes some force but once you get it past that engaged position it moves really easily.  I can feel the "lock" engaging and disengaging as I move the "lock bars" back and forth.   Note that the Nextruder has to be attached to the tool-changer otherwise the lock bars move very easily.  And if you're going to test this yourselves make sure you hold onto the Nextruder otherwise it will drop onto your build plate.

I Picked and Parked all my Nextruders and hand tested them, they all seemed to "lock" correctly onto the tool-changer and I couldn't move the lock bars without a good amount of pressure.  I also moved each Nextruder all the way over to the right side of the print bed using the "Move Axis" control and the LCD X-axis display showed the furthest it would go was 361mm.  That's as far as the machine will let me move the tool-changer and the lock bars were still about 2mm from hitting the stops on the right side of the printer.  If I could get the XL to move another 3-5mm to the right I believe that metal-stop that would disengage the lock bars and drop the Nextruder.  Unless there's a way to trick the XL to print beyond the 361mm x-axis position then the only thing I can see that can apply enough pressure to move the lock bars are the dock locking pins at the back of the printer.  

So yeah the lock bars have to be moving somehow... but how... maybe that heavy vibration I hear when certain movements are taking place is enough to vibrate the "ball-bearing lock" out of place?  It's probably not really a ball bearing lock but I don't know what else to call it.  If that's the case then FW 6.0.0 should fix it.  I'll be doing that update tomorrow morning.