search

prusa3d / Prusa-Firmware-Buddy

Firmware for the Original Prusa MINI, Original Prusa MK4 and the Original Prusa XL 3D printers by Prusa Research.
Other
1.12k stars 219 forks source link

[BFW-5545] [BUG] MK4 keeps requesting filament path to be cleared upon empty spool change #3129

Open bolsoncerrado opened 1 year ago

bolsoncerrado commented 1 year ago

Please, before you create a new bug report, please make sure you searched in open and closed issues and couldn't find anything that matches.

Printer type -

MK4

Printer firmware version - [e.g. 4.0.5, ...]

Any

Original or Custom firmware - [Original/Custom]

Original

Optional upgrades - [e.g. Filament Runout Sensor]

Stock Factory Assembled MK4

USB drive or USB/Octoprint

USB

Describe the bug

  1. Start a print
  2. Spool runs out of filament
  3. Filament sensor kicks in
  4. Filament is removed BUT
  5. even if you reply YES to the "was filament removed ok?" the printer KEEPS asking you to remove filament. Nothing helps, the dial, the clicky, nothing. Only a RESET, ruining the print.
  6. Upon rebooting filament sensor works as intended, so there's no debris no filament messing the sensor, no nothing.

How to reproduce

Read above.

Expected behavior

Since it has happened to me already TWICE and non correlative, I asume this is a firmware glitch. I already opened the idler door and removed any filament (first time none was there, second time a section was there, I chose NO to the question, but apparently it didnt detect I removed it), blowed air, moved filament on the port hole, nothing helps when the bug is present. ONLY Reset button fix the issue and once the printer reboots, everything works as intended.

G-code

No particular gcode.

Video

Next time it happens I'll tape it.

eklynx commented 10 months ago

Already tried that. Didn't help

unclej84 commented 10 months ago

@Prusa-Support Did I get the last Paragraph under section "Current problems of MMU3 on the MK4" in the recent blog post right, that you are also referring to this issue?

Prusa-Support commented 9 months ago

The post on the blog kind of confirms these anomalies are most likely related to debris and irregularly shaped filament that may cause a mechanical blockage of the filament sensor moving parts.

As you probably already know now, they are working on improving the filament tip shaping through a sequence of movements. That is a work in progress but theoretically will drastically reduce the chances of filament sensing problems.

That mostly concerns Multi Material prints where several hundreds of filament changes may occur in a single print. As of now, while we work on the best filament shaping, our best recommendation is to clean the filament sensor. https://help.prusa3d.com/article/filament-sensor-mk4-mk3-9_465702 Nonetheless, sometimes reinserting the filament a few times can do the trick.

Michele Moramarco Prusa Research

bolsoncerrado commented 9 months ago

With all due respect these issues have been happening since day 0 way before mmu3 was even released :/

But ok. We'll take your word.

IftiBashir commented 9 months ago

The issues have been happening since the beginning, but the seriousness has only been highlighted now since it causes huge issues when using the MMU3. Otherwise I wonder how long these issues would have been apparent before Prusa worked on a fix. In my opinion, this 'fix' with the change in filament tip should be implemented asap - MMU3 or no MMU3.

bolsoncerrado commented 9 months ago

And TBH this filament tip thing is just a "patch" to a meh sensor configuration...

jakkaj commented 9 months ago

Happening to me too now. Can't even get the calibration steps now after printer reset :/ 5.0.1 firmware.

signal-2023-11-26-100201_002 signal-2023-11-26-100201_003

jakkaj commented 9 months ago

The cleaning instructions here did help: https://help.prusa3d.com/article/filament-sensor-mk4-mk3-9_465702

jhenley01 commented 9 months ago

I've had this issue multiple times as well. Sometimes wiggling filament will fix it, sometimes not. Between this and the thermistor issues the MK4 has been disappointing. For the price of these machines compared to the competition you'd think they could get basic things like a filament runout sensor to work correctly.

PhLacoude commented 8 months ago

I just experienced this issue for the first time after running out of PLA filament 1h20 minutes into a 1h51m print. This is VERY annoying when a simple GUI change could override the sensor.

bartbergman commented 7 months ago

I had the same issue yesterday. It's definitely a hardware issue (as already acknowledged by Prusa). I had to manually clean the sensor because I didn't have compressed air available, but a tip from Prusa Support was to try to wiggle the metal ball inside the sensor with a magnet to unblock it. Haven't tried that yet, but will definitely try that if this occurs again.

n3wtron commented 7 months ago

It happened also to me and I solved (that time) brutally disconnecting the printer from electricity, inserting the new filament in the extruder opening the "blocker", reconnecting the printer and finally clicking "change filament".

It worked, I don't know if it will always work but...

Prusa-Support commented 6 months ago

The main cause is most likely debris around the filament sensor and this may affect pretty much all types of printers and filament sensors we tested so far, in various ways, and to different extents. Of course, there may be specific cases where we can't exclude other causes of failure - check with Customer Support - but again, it is usually a matter of debris.

The first suggestion should probably be keeping the filament dry as moisture makes the filament more stringy and brittle. Also, inspecting and cleaning the extruder gear and its surroundings is a simple practice that is often overlooked. https://help.prusa3d.com/article/regular-printer-maintenance-mk4_419000

Disconnecting components while the printer is running is something to be avoided. Unplugging the printer should not be necessary and may lead to complications. It may make the printer simply ignore the filament sensor status and encourage the customer to continue without solving the problem itself. I don't feel like recommending this method either.

Removing and inserting the filament a few times could knock off debris from the filament sensor, otherwise, you can clean the filament sensor with canned air, or take the sensor's ball holder off to proceed with cleaning. https://help.prusa3d.com/article/filament-sensor-mk4-mk3-9_465702 Consider trying cleaning with compressed air from below as well (through the idler door) but removing the ball holder for cleaning would be more effective.

Michele Moramarco Prusa Research

pigskins65 commented 6 months ago

I have to respectfully disagree, at least with the firmware level I am at. People have proven that they can get this issue to occur on demand. In my case it happened every time I ran out of filament mid-print. I had to shut off the sensor and pay close attention to the amount of filament remaining.

On Wed, Feb 14, 2024 at 1:51 PM Prusa-Support @.***> wrote:

The main cause is most likely debris around the filament sensor and this may affect pretty much all types of printers and filament sensors we tested so far, in various ways, and to different extents. Of course, there may be specific cases where we can't exclude other causes of failure - check with Customer Support - but again, it is usually a matter of debris.

The first suggestion should probably be keeping the filament dry as moisture makes the filament more stringy and brittle. Also, inspecting and cleaning the extruder gear and its surroundings is a simple practice that is often overlooked. https://help.prusa3d.com/article/regular-printer-maintenance-mk4_419000

Disconnecting components while the printer is running is something to be avoided. Unplugging the printer should not be necessary and may lead to complications. It may make the printer simply ignore the filament sensor status and encourage the customer to continue without solving the problem itself. I don't feel like recommending this method either.

Removing and inserting the filament a few times could knock off debris from the filament sensor, otherwise, you can clean the filament sensor with canned air, or take the sensor's ball holder off to proceed with cleaning. https://help.prusa3d.com/article/filament-sensor-mk4-mk3-9_465702 Consider trying cleaning with compressed air from below as well (through the idler door) but removing the ball holder for cleaning would be more effective.

Michele Moramarco Prusa Research

— Reply to this email directly, view it on GitHub https://github.com/prusa3d/Prusa-Firmware-Buddy/issues/3129#issuecomment-1944407797, or unsubscribe https://github.com/notifications/unsubscribe-auth/AHZD3ZC7BPM22YZIJJPMRQTYTUBTTAVCNFSM6AAAAAA2A4ZWDGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSNBUGQYDONZZG4 . You are receiving this because you are subscribed to this thread.Message ID: @.***>

PhLacoude commented 6 months ago

I agree with @pigskins65. Could you simply add a button in the GUI that would allow us to override the faulty detection of the filament removal?

Just to recap, the filament is removed BUT even if you reply YES to the "Was filament removed?" the printer KEEPS asking you to remove filament. At this point, it would be very simple to add a button say would say "Are you REALLY sure the filament has been removed?". It would let us go on with our lives.

In my case, I tinkered with the wire of the filament sensor, making sure it would stop indicating there was filament left when there weren't. But this is a hardware solution to what should be a software issue.

jhenley01 commented 6 months ago

I’m not really buying the “this may affect pretty much all types of printers and filament sensors” answer.

I’ve owned several printers from different manufacturers and the MK4 is the only one I’ve had repeatable failures with the filament runout sensor.

Simply admitting you guys made a mistake with the design and working on a fix will go a lot farther with your customers than trying to claim it’s a normal thing that happens with all printers.

fkirill commented 6 months ago

UPD: Turned out the filament sensor ball got stuck for whatever reason, I pulled the filament in and out a few times and it appeared to clear up the issue.

+1, same problem here. THe filament ran out, the printer is stuck with "Please remove the filament" message and there is nothing I can do. I will have to start the print from scratch. It goes without saying that the filament channel is empty and it looks a lot like a firmware bug.

I must say this I have already run out of filament multiple times before and everything worked fine. So maybe this is a more complex issue that only occurs sporadically, maybe some kind of a race condition etc. The problem is definitely there.

Also, an important thing: this is the third time I ran out of filament on a single print (I'm just finishing a few almost empty spools), maybe a repeated out of filament events in the same print causes an issue. Or maybe I have have a high enough concentration of "out of filament" events that I was able to hit the bug.

You can try using small pieces of filament (20cm ones should do) to reproduce the problem.

miyconst commented 6 months ago

Today my MK4 got stuck at the same annoying message, in my case it turned out that the filament end was bent and it got stuck in the path, from the outside it looked empty but with a torch light, I saw the issue. Used a needle to straighten the string and pull it out, then I was able to resume the print.

Still, the current behavior of the firmware/printer is simply not acceptable.

pjvenda commented 5 months ago

Had this issue triggered mid-print when a filament ran out. Current firmware is 5.1.2.

Was asked to check the filament had been removed (no filament was visible - it had been pulled into the print), confirmed, then got stuck in the "Changing filament' - 0% screen. Had to reset the printer and loose the print.

After resetting and inserting new filament, a large amount of fairly thick filament came out of the nozzle, as if too much had been pulled in... nonetheless, new filament was loaded successfully.

jascha-sacc commented 5 months ago

+1 to all this! I just lost a print twice because of this problem (the printer's in a shared space, the spool had some nasty tangles in it that didn't show up until partway through, and that's why I was changing the filament).

Specs:

Absolutely having a GUI override would make sense here.

pjvenda commented 5 months ago

Had this issue triggered mid-print when a filament ran out. Current firmware is 5.1.2.

Was asked to check the filament had been removed (no filament was visible - it had been pulled into the print), confirmed, then got stuck in the "Changing filament' - 0% screen. Had to reset the printer and loose the print.

After resetting and inserting new filament, a large amount of fairly thick filament came out of the nozzle, as if too much had been pulled in... nonetheless, new filament was loaded successfully.

I should add that the exact same situation happened again the same day. This time a small amount of filament was left on the printing head. I tried the technique described around using a piece of filament and poking it in the hole repeatedly, which after some insistence actually worked.

The sensor does not appear dirty or malfunctioning because it correctly detects the presence of filament when swapping filaments while a print is not running. My suspicion is that the path of the filament is not clearing enough to return the sensor detector because when poking the hole with the filament I could not feel the springy resistance of the ball as I did when replacing the filament outside the printing operation.

I hope this helps. And I hope somebody is working to fix this.

bolsoncerrado commented 5 months ago

The very few interactions "Prusa" has done in this thread suggest otherwise.

bolsoncerrado commented 5 months ago

And oh well! Here goes karma!

After replying i launched a 2 color print (2 color by M600 in the middle of the print).

Upon reaching M600 it unloaded fine but it complained of filament still being present. No matter what I did, no matter how much air or filament i moved on the tiny hole, no matter how much I cleaned with the idle door opened (and no filament could be seen in there), NOTHING could be done, so I tried to UNPLUG the printer in hopes to load the filament when POWER PANIC kicked in....

Upon booting up, the printer tried to resume printing with NO filament and printing WAY HIGHER than expected!

So definetly needs a firmware revamp on all these issues starting with MORE options on screen and defiently powerpanic NOT resuming correctly when youre changing filament...

SUCKS BIG TIME to say the least.

jascha-sacc commented 5 months ago

Well, i just pasted the link into a Live Chat with a Prusa help person, so hopefully that might put it back on their radar?

jascha-sacc commented 5 months ago

Had a long chat, actually - maybe it'll make a difference!

PhLacoude commented 5 months ago

So definetly needs a firmware revamp on all these issues starting with MORE options on screen and defiently powerpanic NOT resuming correctly when youre changing filament...

I strongly agree with "MORE options on screen". This whole thing could be circumvented with additional screen options. I understand the hardware aspect might be difficult to sort out. But there is no excuse for not adding a couple more buttons on the screen.

jhenley01 commented 5 months ago

Another day and another failed print at 98% complete because of the runout sensor issue.

I'm also running 6.0 RC1 firmware and can confirm the issue is still there and still no menu option to override. Seems to me that Prusa really has no intention of fixing this issue.

bolsoncerrado commented 5 months ago

Perhaps as easy as "Ignore filament sensor for the rest of the print" thing or option to be pressed available during a print and filament change or filament runout or something like this besides the retry/cancel requests...

It's sad Prusa guys are no longer sharing further insights here....

bolsoncerrado commented 5 months ago

GEez, v6.00 ahead and no word for sensor improvement. Awesome :( https://github.com/prusa3d/Prusa-Firmware-Buddy/releases/tag/v6.0.0-RC1

jascha-sacc commented 5 months ago

So I had a great convo with a Prusa help person via chat where I dropped the URL to this thread - let's have more people also do that and keep the issue fresh in their minds.

unclej84 commented 4 months ago

After having modified the sensor and idler for the mmu3 I had no further issues with this. Unfortunately the seem not to sell a kit or spare parts for this modification without the mmu3 and the assembly manual still shows the old ball/magnet/spring and idler.

bolsoncerrado commented 4 months ago

That would clearly prove their really BAD hardware decision on this and why they've been so silent about it.

jhenley01 commented 4 months ago

I’m betting they’ll do one of two things.

They’ll either quietly do a hardware fix for new units going out and try to avoid having to replace parts for existing customers.

Or they’ll do what they did when the Mini had all heat issues with the MINDA probes. They came out with the “SuperPINDA” kit and expected people to buy an “upgrade” to fix their design flaw.

Prusa-Support commented 3 months ago

Unfortunately, our reply hasn't changed - https://github.com/prusa3d/Prusa-Firmware-Buddy/issues/3129#issuecomment-1944407797 - but rest assured that your comments are not being ignored.

As explained in the similar/same issue before this one - https://github.com/prusa3d/Prusa-Firmware-Buddy/issues/2970#issuecomment-1924886097 in fact opened before the release of Prusa MK4 but also listing MK4 cases - similar problems may be observed to some extent with pretty much all kind of filament sensors we tested and/or released so far, and, as many highlighted in this thread, the problem seems to be only reproducible in the case of mechanical blockage.

There are chances that the stamping process introduced in Prusa Slicer 2.7.3 will reduce some scenarios where the filament conditions and settings are the potential problems.

However, as per https://github.com/prusa3d/Prusa-Firmware-Buddy/issues/2970#issuecomment-2016865010, our best recommendations at the moment remains the following.

Basic guidelines can be found in the relevant Help article linked below. https://help.prusa3d.com/article/filament-sensor-mk4-mk3-9-xl_465702

Michele Moramarco Prusa Research

bolsoncerrado commented 3 months ago

How sad to read...

PhLacoude commented 3 months ago

For the record, I strongly disagree with Michele's reply.

When the spool runs out of filament, and the filament sensor correctly senses it, there is no reason to keep the user in an infinite loop asking him/her to remove the filament after it has already been removed.

At a minimum, instead of perpetually asking to remove the filament that is no longer present, the menu on the screen should let the user CONFIRM that the filament has been removed, and let the user proceed at his/her own risk. From experience, we know if we have filament in our machines.

FOR ALL PEOPLE WHO WOULD COME TO THIS PAGE LOOKING FOR A WAY OUT IN THE MIDDLE OF A 10+ HOUR PRINT, you can fix the issue by unplugging the filament sensor on the LoveBoard. The MK4 will detect that the filament is gone and let you out of this endless menu. You then plug the filament sensor back. Cumbersome but it works. https://help.prusa3d.com/guide/5-nextruder-assembly_434014#435286

bolsoncerrado commented 3 months ago

Not only I agree with the software solution of a confirm question but I also have the feeling that upon unloading midprint, somehow the sensor calibration value may go nuts, that's the only explanation on why moving in and out in quickly succesion the filament triggers the response BECAUSE WHEN THERE'S NO PRINT GOING ON THERES NO SUCH ISSUE WHEN UNLOADING FILAMENT.

So it must be something firmware glitch side. A variable not correctly called or emptied or whatever.

Anyhow the confirmation should be more than enough and not leaving you option less with just a "please remove" when you have already done so mid print!

Prusa response to this is just lame.

For the record, I have the feeling unplugging and plugging back live components MAY FRY something so beware!

Prusa-Support commented 3 months ago

Disconnecting components while the printer is powered on is something you should really avoid. Not to mention that this workaround would most likely only postpone the problem. https://github.com/prusa3d/Prusa-Firmware-Buddy/issues/3129#issuecomment-1944407797

Cutting power to trigger an intentional Power Panic and resume later, in this case, is not an intended feature - results are not guaranteed. I guess it may work in some cases, but again it is not a solution and will most likely only postpone the problem. https://github.com/prusa3d/Prusa-Firmware-Buddy/issues/2970#issuecomment-2016865010

If the ultimate request is to have an option on the screen to postpone the problem troubleshooting, I can't assure a positive outcome but it will be my concern to collect all feedback about it and share it with the developers. For now, our recommendation remains to clean and inspect the filament sensor. https://github.com/prusa3d/Prusa-Firmware-Buddy/issues/3129#issuecomment-2105938590

Michele Moramarco Prusa Research

mobilla commented 3 months ago

I agree disconnecting components is a bad idea mid print, and relying on power panic is also a bad Idea. However they are the ONLY options right now because a "skip check" button isn't available. If I am 20 hours into a 21 hour print and need to change filament I am NOT going to cancel the print and start again with a larger spool when a simple button will do. I would rather delay that problem, and try to fix it in an hour, after the print is done.

Until the real issue can be fixed, please implement the simple solution of a skip check button so we can stop needing to do riskier things which could result in real damage or even fire if you accidentally unplug something like the heater element. I would rather risk a failed print than waste 20 hours or break my printer completely.

miyconst commented 3 months ago

@Prusa-Support

I love my Prusa MK4, I plan to buy an XL, and I want to love Prusa as a company, but this issue makes me doubt the competence of people answering here.

What I am going to write will most likely be considered by some as rude and unprofessional, so I say upfront that I don't want to hurt or target anyone, I want Prusa to improve.

bolsoncerrado commented 3 months ago

A-Men!

PhLacoude commented 3 months ago

"Disconnecting components while the printer is powered on is something you should really avoid. Not to mention that this workaround would most likely only postpone the problem."

Yes, @Prusa-Support, I strongly agree that this is something we should really avoid.

But if you gave us an "ignore and continue" button, as @miyconst suggest, we could finish the print after the spool ran out on a long print. I believe that none of us have proposed this change to do 20+ filament changes. Most of us would have liked to have this option to use once to complete a long print after a spool of filament ran out.

It is not even always true that it postpones a problem. As many of us said, there was nothing in the way of the filament and, after resetting the machine, the filament sensor would no longer "see" a filament (or debris).

otaviojr commented 3 months ago

+1

I'm facing the same problem with the new MMU3 and the newer idler filament sensor.

A 2-hour, single filament print stopped five times.

I need to click continue after it ejects the filament, and it will reload it and continue printing.

Annoying. I need to be there to click a button.

Now, I will need to wake up at night hourly to see if I need to click the button. LOL

rwillett1 commented 3 months ago

My Prusa Mk4 arrived a few hours ago, set it up, got it working, decided to simulate an out of filament issue and I have the same sodding problem as everybody else. If I push an old piece of filament in 5-10x, it appears to work.

So I can safely say that Prusa has not quietly fixed the hardware problem or used software to fix it. This thread started 11 months ago and we still have it on brand new printers. Mine is three hours old.

I've just brought a brand new printer that is not as good as my old Mk3 that it replaces and was going to be given to the local school.

I am furious that this major issue has not been fixed. I'm contacting Prusa to send it back, its not fit for purpose.

Rob

bolsoncerrado commented 3 months ago

I'd really like to see Prusa staff to come here one last time and repeat, after this last evidence, that it's filament debris the ONLY issue this filament sensor has...

As the saying goes... "There's none so blind as he who will not see."

Geez.

rwillett1 commented 3 months ago

I've just cut the filament to see what happens.

Filament unloaded OK, this time it did load but I had to reload it to get it to work. it put up the message "Ramming" on the screen. I pushed a little harder and it worked.

So lets do another one and see what happens. This is on a printer that has less than two hours usage from coming out the box.

The last two have worked OK, lets keep doing it and see.

Rob

otaviojr commented 3 months ago

I've just cut the filament to see what happens.

Filament unloaded OK, this time it did load but I had to reload it to get it to work. it put up the message "Ramming" on the screen. I pushed a little harder and it worked.

I never cut the filament. In my case I just click continue and it will reload and continue to print. No interaction with the printer.

I will try to calibrate the finda better.

This never happened with mk4 alone. It started after the MMU3.

rwillett1 commented 2 months ago

The filament sensor just failed to recognise it was out of filament. Six hours of printing down the drain. If this was a Creality Ender 3 Pro V1 from five years ago, fair enough, this is a £1,100 machine and it doesn't even recognise it's out of filament. Its three days old.

Just tried it again and its happily printing with no filament. This is garbage.

Also Prusa do not have a Live 24/7 chat, isn't working at the moment.

otaviojr commented 2 months ago

I did calibrate my Finda sensor and was able to print a 5h part with no issue.

I believe my problem is not related to this.

I think it was the Finda all the time.

MikeDabrowski commented 2 months ago

Happened to me second time. First time it was on last 5v and this time is on latest 6.0.2. Old filament unloaded - asked to remove the end, tried to load to finda twice and just sits there waiting for idk what.

Last time I talked with support they couldnt help it and we threw it off as a random bug. This time I am waiting 20min for the chat. Already restarted the mmu3 twice, on this latest version there is an option now to disable MM.