Printer initiates color change without M600 in G-Code.  

RE: Printer initiates color change without M600 in G-Code.

@groeniediedraak

Good day all.

I would appreciate help on solving a problem that occurred now on 4 occasions within 5 prints where the printer stops for a color change (which I did not request in Slicer). I unload the filament and then reload the same filament, but the new layer then does not stick properly and I have to abort the print.
I searched through the G-Code for a M600 command, but there is none.
Printer: MK3S with MMU2S (the MMU was disabled and the files were sliced for Original Prusa i3 MK3s & MK3S+).
Slicer version(s): This happened with 2.3.0 and then also with 2.3.1 (updated on 18 May).
The first 2 times this happened was with the previous version of the firmware (3.9.3). I then upgraded on 18 May to the new firmware (3.10.0 and it happened again.

I have never experienced this problem before with hundreds of prints sliced and printed with the same settings and methods as described above.

Any help to solve the problem will be highly appreciated.

Thank you.
Carl

Save the project file from Slicer (.3mf) ; zip the 3mf, and post it here. Doing this will let us see the profiles and settings you are using. You'll get meaningful help a lot quicker.

It is also possible your filament sensor is acting up. That will require monitoring the filament out flag to see if it is toggling at random times. That will be tedious and boring for you to watch. The other option - if not a slicer issue - is to tear down the extruder and verify the build of the sensor parts and assembly.

RE: Printer initiates color change without M600 in G-Code.

Hi Tim,

I hope I added it correctly. This was my first ever post on the forum.

The color change on this file happened at about 3.8mm (measured with caliper).

The others happened on different heights (even with the same print that was re-sliced)

Thank you

Swiwel_BasisNew

RE: Printer initiates color change without M600 in G-Code.

@groeniediedraak

Thanks. I don't see any reason there would be a filament change in your project; and my slice agrees with yours: no M600 present. And adding a layer change does in fact cause it to show up in gcode.

So it is probably hardware. Scan through the menus for the filament sensor (down in the Sensor menu near the bottom of the list).  I'd unload the filament, then open the sensor menu and watch it for a while looking for any flickering ON/off.  Then, I'd start loading and unloading filament after a few load unload cycles just let it sit there loaded and see if it is stable with filament in.  Probably wouldn't hurt to try wiggling the filament to see if bumping the ball has any effect. There are issues with the flag on the sensor arm being off and causing issues. More on the S than the S+ ...

https://help.prusa3d.com/en/article/ir-filament-sensor-troubleshooting-mk2-5s-mk3s_112226/

RE: Printer initiates color change without M600 in G-Code.

@tim-2

Thank you for the reply.

I will give it a shot on watching the sensor. I replaced the sensor arm many moons ago (at least 18 months ago) with a modified version I found in the Prusa Prints and it has been working flawlessly ever since.

I will report back any findings.

RE: Printer initiates color change without M600 in G-Code.

Most likely the filament sensor as @tim-2 suggested. I've seen exactly that behavior on one of my machines after some work to the extruder, and a quick readjustment solved it right away.

RE: Printer initiates color change without M600 in G-Code.

@fuchsr and @tim-2

I tried the suggestion of checking the filament sensor. No matter what I do (wiggling, pulling, pushing, etc.), the value stays 1 when filament is inserted.

I do think that I might have identified the reason for the problem, and it is not related to a color change. I suspect it to be a filament runout trigger. While scrolling through the menus, I checked the Print Fail Stats and the Last Print Fail shows Fil. Runout = 1.

The roll of filament is about 2/3 full, so no runout. I then dived into the dustbin and dug out the piece I cut off and there are 2 gouges on opposite sides of the filament where the Bondtech gear must have chewed it and triggered a runout (which I mistook as a color change request).

The filament (PLA) also feels much stiffer and harder than my other PLA filaments. I have not adjusted the tension spring in a long time, so maybe the harder, more brittle filament caused it to be chewed on a retraction. I will only be able to test the same print with another roll of filament tomorrow as I have a prior engagement today that will not allow the print to finish if I start it now.

Thank you so far for the advise and help. I will revert back when I have completed the test tomorrow.

RE: Printer initiates color change without M600 in G-Code.

Thinking about it now, it could not have been a Runout as the filament was still in and the sensor could not trigger...

RE: Printer initiates color change without M600 in G-Code.

It's not a runout but the IR sensor thinks it is. 

RE: Printer initiates color change without M600 in G-Code.

@groeniediedraak

Well - you have identified it IS the filament sensor failing.  The sensor is triggering even though you have filament.  The problem could be a glitch (random unexplainable events), a loose wire or bad crimp in one of the connectors, a plug not fully inserted, or just a bad connection (contacts not wiping cleanly).  Unplugging both ends of the filament sensor, plugging them back in and wait and see.

Or, the sensor has deteriorated (would be odd), or the mechanicals of the flag mechanism have changed (likely).  It wouldn't hurt to do A) above, then take the extruder apart enough to verify the mechanism per the Prusa troubleshooting guide.  

As a last resort - disable the filament sensor.  Prusa is known for unreliable filament detection as far back as I've read these forums.

RE: Printer initiates color change without M600 in G-Code.

I would disable the filament sensor before dissembling anything and see if the problem goes way, depending on your usage you may not even need it mine works but I have had it turned off for 4 years now. never had a single incident.

RE: Printer initiates color change without M600 in G-Code.

I agree with disabling the sensor.  You will need to take other steps to avoid run outs of course but as long as you're aware of the need you should be OK.

The unusually stiff PLA is probably showing symptoms of dampness, try drying it, it should return to normal pliability.

Cheerio,

RE: Printer initiates color change without M600 in G-Code.

@swiss_cheese

That's exactly how I confirmed the issue when I had the same problem as the OP. Disabled sensor in menu, problem went away. But I did like the runout detection, so I ended up partly disassembling the extruder and recalibrating the sensor as per Prusa Knowledgebase. Problem never reappeared on that machine.

RE: Printer initiates color change without M600 in G-Code.

I have this on one printer.  It was baffling so I have kept it disabled.  I may replace the sensor on the rebuild in a few weeks. 

RE: Printer initiates color change without M600 in G-Code.

First of all I would like to thank you all for the replies and advise.

I think the problem is with the specific roll of filament I was using. As mentioned before, this filament is much stiffer and "feels harder" than my other PLA filaments. I printed the parts again with another roll of filament that feels more flexible and softer. All the prints completed without fail. I also sat here and watched the filament sensor, as suggested, for as long as humanly possible (+- the first 2 hours of the first print & +- 1.5 hours on a second print), following each layer on the Slicer. The value did not change from 1 to 0 at anytime my eyes were on the LCD.

For now, I will assume that this was the problem. If I do get another failure like this, I will do a test with the filament sensor disabled and then try and fix the sensor problem. I will also try drying the roll of filament, as suggested, and see if it completes a print without fail.

Should anything change (i.e. if I suspect the sensor in the near future), I will return to this thread and update the information.

One question that I have is, when using the MMU2, must he filament sensor be on to communicate a filament load to the MMU2 in order to tell it to stop pushing the filament?

Thank you all for the contributions.

RE: Printer initiates color change without M600 in G-Code.

@groeniediedraak

I think the answer is yes, else the 'system' doesn't know when to stop advancing via the MMU motors.

Regarding the filament spool being the issue.  There's a chance the harder springier filament spool bounced as it turned as spools will do and somehow sprung the BB from rest to flick the IR flag arm. But the notion pushes credulity. But if the flag is already marginal, and a marginal springy filament snaps back ... lol. Strange things happen.