Printer self-restarts after calibration  

RE: Printer self-restarts after calibration

Sounds like a firmware problem (scrolling aliens) ... check that you have MK3s firmware, and appropriate MMU firmware installed.  It's common for folks to grab the wrong files when downloading.

Also - the timing might be hinting at a extruder motor driver problem.  Seems like when it goes to print the purge line it crowbars the PS (forcing a full reboot like a power fail would).  If the printer load and unload filament works normally, not sure; but if that also has issues, inspect the extruder drivers and motor for problems.

RE: Printer self-restarts after calibration

@tim-m30 absolutely there are firmware problems. I installed the correct file, but I'm still getting all kinds of issues. I can't even begin to list all the screw ups this thing has. At the moment I'd like to ship it back to Joseph packed in a box of dog crap.

I've only seen the scrolling aliens once. I've taken it all apart and reassembled the extruder head and there's nothing wrong there. The MMU seems to be operating correctly as far as the mechanics, but behavior for the whole thing is bizzare. I can't even get the damn thing to load filament to the nozzle. I've tried dozens of times. I'll reload the firmware again. It's the same firmware it has now, but I'll start with a fresh download.

Post replacement of firmware (yes, I did it right, yes really I did) it won't come out of boot screen.

"Original Prusa I3 Prusa Research"

I've tried reflashing a few times and it's all the same. Looking for solutions, I keep running across trying a factory reset. I tried both the light and full factory resets. I get a beep, but it doesn't do anything else. 

I'm a very capable person with just about every machine. Every time I touch this thing, it's worse off than before. Have I bricked this piece of garbage?

RE: Printer self-restarts after calibration

Check your FINDA/IR sensor. There's a known issue where the printer will hard-reset if the filament sensor in the extruder reads 1 but the FINDA reads 0, which sounds exactly like your original issue.

RE: Printer self-restarts after calibration

So, I've reflashed the printer several times. Each time, it wouldn't go past "verifying 100%" . If I unplugged, it would stick at the home screen. I finally figured out that I didn't need to use a file with a second language. Why are those files even there if they don't work? I'm running again with the correct firmware. First things first, I wanted to run a first layer cal after some work on the extruder. It had been clogging and I wanted to clear everything out of the nozzle. Just like clockwork I have an MMU failure for filament 1. Open the MMU lid and check the filament - it's good. I cut the end at an angle and try again. The printer goes to Live Z adjustment with no filament in the nozzle, or even in the bowden tube from the MMU. I press the menu button and it goes to the main screen. The nozzle temperature drops although it's flashing "heating" at the bottom of the screen. It's forgotten about the FLC entirely. I've done this loop dozens of times. I could load the filament first, then do the FLC, except it refuses to load anything. Ever.

I take the advice of others and turn off the filament sensor. I then do a "load to nozzle" on filament 1. It advances to the extruder, I see the door move and I know it's reaching the gears. It pulls the filament all the way back out, and then asks me if it's extruding correctly. No. No it's not extruding correctly and it never will, because you haven't advanced the filament to the extruder you piece of garbage! What the F can I do with this thing if it's doing this crap?!?! I can't do the minimum steps to see plastic drip from the nozzle, how the hell could it possibly print something?

I reset with the X. I try it again. It preheats, makes some noise, Then the MMU flashes all 5 green. I leave it alone until it says "MMU needs attention". I open the MMU and look. I fix nothing, it's absolutely fine. I press the button and it does it's dance and tries again. Now it says "MMU ok, resuming position" I see filament extruding at the tip! But no, now the MMU is upset again. It has ground a divot in the filament. Before anyone asks, it's adjusted correctly. The screws are ever so slightly below the surface of the MMU body. Hoping I can salvage this attempt, I open the MMU, remove the Bowden tube from the extruder, pull the filament through and cut off the bad part. Then I push the filament through the gears on the extruder and put it all back together. I actually did this twice. Now it did it again, even after I tightened the MMU a bit more to try to keep the gears from spinning.

Fine then. I'll remove the bad portion a third time. I'll reset and ask it to unload - just so it knows where the filament is. It pulls the filament back about 50mm. Maybe I'm using the wrong command? I use the Eject Filament command. MMU is unhappy. There is no problem. Nothing. The MMU will not respond to a button press on any of the three buttons to continue. I wait. I press and hold the right button. The selector moves and shears the filament off. I asks me to remove the filament, but I can't. The selector design is really bad and has pinched the filament between the bodies. I pull hard with pliers and finally remove it. When My selector moves it now has a fun new mouse squeak. That's probably good, right?

Back at a stable condition, I X reset. I then ask it to do a FLC from brand new PLA on filament 3. It heats up, advances well to the extruder, retracts, advances again, retracts, advances a third time. The head moves to the back of the platform and retracts the filament. MMU load failed. It isn't asking me for anything. I press the menu button and it begins heating again. There is no filament in the tube, but we're back on the "Adjusting Z" screen. How can it possibly think that this is happening? After 2 minutes, it says that the MMU needs attention. It's dug a hole in the filament again. There's nothing wrong with it, the pathways are clear. The filament moves without too much effort in the tubes. I see it move the gears each and every time it advances, but no luck. 

Clearly there are some firmware issues. I'm going back to the previous version and I'll keep you updated.

RE: Printer self-restarts after calibration

Update:

Reverted firmware to V3.7.0 (1.0.5) for both rambo and MMU. 

Tried FLC. It loaded filament once, retracted, then ran the FLC empty. 

Tried to Load to nozzle. Ran filament to the extruder once, retracted, then asked if it was extruding the correct color.

Turned the filament sensor back on. Tried to load to nozzle again. It moved the selector without cutting the filament and jammed it all up. Surely the MMU is made for this and is absolutely indestructable! It's really excellent that I have to take the whole thing apart to get to the blade that is so poorly designed! 

What gives?

RE: Printer self-restarts after calibration

I don't own an MMU; I've read way too many folk reporting issues getting it to work well so I dodged that bullet.  I wish I could help, but since I have no experience with the MMU, perhaps Vintage or someone else can help.   Or, contact Prusa directly through CHAT and have them help.  You paid for a warranty.

RE: Printer self-restarts after calibration

Yes, I'm pleased to say I had minimal issues getting mine working after some initial teething and very quickly ditching the buffer setup and factory PTFEs. I'm not sure how relatable my experience is as I have a Skelestruder setup. The stock extruder design has some less-than-optimal design areas of the filament path for the MMU functionality. In my setup there is PTFE to just above the gears, and PTFE immediately below, the filament in the extruder has absolutely minor contact with printed parts and the minimal number of transitions out of/into PTFE.

A couple more comments/observations: 

- Remove the blade. It's useless/unused right now so all it does is cause problems when there's a bind.

- I don't think you *can* run the MMU with filament sensor disabled. It's part of the core system logic, so if it can't sense where the filament is then all bets are off.

- I get you're frustrated/angry/annoyed, but the inline rants/anthropomorphism obscure the core content of your post. You're more likely to get helpful replies if people don't have to sift through all of that to get the important info.

- As suggested previously, have you verified the filament sensors (FINDA/IR) are calibrated correctly? Both of them need to work properly for the system to function. And that means that neither of them must flicker/change state when filament is moving through them, or you're in for a world of hurt and weird issues, and many of your symptoms are consistent with what I'd expect if the sensor readings are incorrect.

- I found my screws to be too tight when slightly below the body; I ended up having them flush or just a hair proud for best results.

- If you have *ever* replaced the hot-end PTFE with something not from Prusa, check its diameter/tolerance and that it is installed correctly. In fact, check the hot-end PTFE installation anyway. Unsecured PTFE (print/buy a collet clip if you don't already have one) or something larger than the official 1.85mm ID stock is going to cause you a world of pain in getting filament tips correctly formed.

- If you're willing to put in the time, consider making a Skelestruder. Even just the serviceability is a major QoL improvement for jams when debugging. The IR sensor setup is still a little finicky, but much less so than the stock design to get right.

RE: Printer self-restarts after calibration

Vintage, 

Thank you for your response. I have considered ditching the buffer as well. Using a straw, I can usually get things loaded pretty quickly, but it's still pretty far from optimal. I also considered widening the area between the plates by printing new brackets as well as buying PTFE with a wider internal diameter to go between the buffer and MMU. I've seen all the other buffers on thingiverse and I'm not thrilled with them. Either way, it doesn't seem to be the biggest issue - that the printer retracts when it should load. 

I only have about 70 hours on the machine. Pretty much all of that was before I installed the MMU. So my machine is still on the original PTFE tube inside the hot end that the machine came with. I did remove it to inspect it, and reinstalled it but that hasn't changed the behavior of the machine or the success rate at all. And there have been a couple of times where it loaded to the point it was oozing from the nozzle, but then retracted to the MMU before continuing. So I'm sure that the path is clear.

When I have the time, I'll check or remove the blade on the MMU. I wish it were easier to remove from the machine. A connector of some kind would be nice. Maybe that's a mod I can do in the future.

I rechecked the sensors this morning and both Finda and IR are working very reliably. No issues that I can see. 

As for the rants, you're right. They detract from the argument. I'm incredibly frustrated. If it worked some of the time, I could narrow down the issue with tests. As it is now, I'm helpless. No loading = no testing. Emotions are difficult to express adequately in technical writing and I want to be sure that the Prusa people know how upset I am. Zero improvements in the machine have wasted huge amounts of my time and ruined many days.

I ran another test this morning. Left the IR sensor on, changed to 7x7 leveling. Preheated and then ran a first layer calibration.

RE: Printer self-restarts after calibration

Are you using Octoprint??
If yes, check out the Printer Profile in Octoprint, choose your printer, and click on the "Hotend & extruder" tab.

MAKE SURE the "Number of Extruders" is set to "5", and then the "Shared nozzle" box is ticked.
I think one of the updates of Octoprint messed up the printer setting....I saw many many people are having the same issue. 

I ran into this issue last month, tried upgrading/downgrading the firmware, disabling / enabling filament sensor, checking almost all of the things I could check...but no luck. Hope this post helps you.