Notifications
Clear all

[Fermé] I challenge Prusa to use MMU2S  

Page 6 / 10
  RSS
Nailig
(@nailig)
Active Member
RE: I challenge Prusa to use MMU2S

Casey, He did ask a direct question, which could be answered with a short direct aswer. You could give a short descroption of your problem. But instead.....  lost of complaining and i still don’t really know what you need. 

call Prusa for help, it is obvious you will not get what you want here.  

Publié : 31/07/2019 2:37 pm
CybrSage a aimé
Casey
(@casey-2)
Estimable Member
RE: I challenge Prusa to use MMU2S
Posted by: CybrSage

Casey, reported.

Stop now.

Same to you.

Publié : 31/07/2019 2:39 pm
CybrSage a aimé
Scott
(@scottmg)
Active Member
RE: I challenge Prusa to use MMU2S
Posted by: Casey/NZ2O

Agree totally with OP. This seems to have made my printer less reliable/not work at all. 🙁 I have spent hours upon hours. My issue? It doesn't seem to make it to the bondtechs. It's a snag right around the chimney. Take the bowden tube off top, and feed it yourself? Works fine. Let it feed as designed, NOPE.

I had similar issues and thought I'd share my experience so far here.  I've had a frustrating couple of days.  Most of what I know is probably already here somewhere, but it would have been nice to know.

1.  When in doubt, calibrate your sensors.  Then do it again.  Almost all of the crazy behavior is related to sensor readings.

The first sensor issue was with FINDA.  I don't think I was rough at all with it, but it was broken.  The worst part was it worked fine for a while, then eventually went wacky.  I eventually realized if I fiddle with the cable, the value changes in the Support > Sensors area.

The next issue is with the new filament IR sensor.  I really wish that'd be redesigned.  I believe the root of the issue here is that it's too easy to get "on the edge".  Even something like changing the tension in the idler screw can make it stop working.  I spent forever one night pulling my hair out, wondering why the damn MMU wouldn't send filament to the extruder.  I finally realized my IR filament sensor was right on the bubble.  Even jostling the extruder around would sometimes trip it, and I think in reality it was constantly flipping.  My suspicion is that this confused the MMU/I3 so the MMU just stopped sending filament. 

Prusa - how about some debouncing on your sensors and/or some more specific error messages.

2.  The filament tips matter - sort of.

They make a big deal in the document about getting nice, pointed tips after extracting filament.  In my (admittedly short) experience, stringing is fine as long as it's not too thin.  If I get wispy strings that break of, those will clog the filament path, even get caught in the FINDA ball. 

3.  We need a better solution for purging. 

My experience with the purge tower has been less than great, with some filament combinations not sticking well and creating a mess.  Although in retrospect that might have been related to a partially clogged nozzle I had. 

Lately I've been editing the gcode and entering my own purging solution, but that's going to be a no-go on prints with more than a few dozen filament changes. 

4.  Somehow it all needs to be faster.

My latest issue has been around some thin features next to supports.  It requires several filament changes, which equal lots of time.  The thin structures cool, and inter-layer adhesion is abysmal (PETG).  When I skip the supports and print faster, it's incredibly tough, all with the same settings.

Anyway, I like the unit better now than I did a few days ago (when I wanted to throw it out the window).  If could ask Prusa one thing it would be to work on the sensors - make them more resilient and harder to setup incorrectly.

 

Publié : 31/07/2019 11:48 pm
Casey a aimé
bobstro
(@bobstro)
Illustrious Member
RE: I challenge Prusa to use MMU2S

Holy smokes, guys. I've lost interest in my MMU2 so haven't been following along. The party is happening in this thread!

I'd just like to add...

LAST.

 

My notes and disclaimers on 3D printing

and miscellaneous other tech projects
He is intelligent, but not experienced. His pattern indicates two dimensional thinking. -- Spock in Star Trek: The Wrath of Khan

Publié : 31/07/2019 11:54 pm
Nikolai a aimé
CybrSage
(@cybrsage)
Honorable Member
RE: I challenge Prusa to use MMU2S

I wish we could get to the sensors menu while it was doing loading and unloading. That would help a lot with troubleshooting.

Publié : 31/07/2019 11:59 pm
bobstro
(@bobstro)
Illustrious Member
RE: I challenge Prusa to use MMU2S
Posted by: CybrSage

I wish we could get to the sensors menu while it was doing loading and unloading. That would help a lot with troubleshooting.

Might the terminal output from Octoprint not provide that info? Not the most readable, but a script might be doable. What specific sensor info do you need?

My notes and disclaimers on 3D printing

and miscellaneous other tech projects
He is intelligent, but not experienced. His pattern indicates two dimensional thinking. -- Spock in Star Trek: The Wrath of Khan

Publié : 01/08/2019 12:03 am
Nikolai
(@nikolai)
Noble Member
RE: I challenge Prusa to use MMU2S
Posted by: bobstro
Posted by: CybrSage

I wish we could get to the sensors menu while it was doing loading and unloading. That would help a lot with troubleshooting.

Might the terminal output from Octoprint not provide that info? Not the most readable, but a script might be doable. What specific sensor info do you need?

Best solution what I've seen so far was the led on top of the sensor. Easy mod and always good visible status of this sensor.

Often linked posts:
Going small with MMU2
Real Multi Material
My prints on Instagram

Publié : 01/08/2019 12:05 am
CybrSage a aimé
CybrSage
(@cybrsage)
Honorable Member
RE: I challenge Prusa to use MMU2S
Posted by: bobstro
Posted by: CybrSage

I wish we could get to the sensors menu while it was doing loading and unloading. That would help a lot with troubleshooting.

Might the terminal output from Octoprint not provide that info? Not the most readable, but a script might be doable. What specific sensor info do you need?

It might, but I do not use octoprint.  I want the pindar, finda, and ir ones.  That way I can hit Load to Extruder and watch the sensors change as it loads.

It would help with removing one potential item for teoubleahoiting if it could be monitored in real time from the LCD panel.  But they prevent its use while loading and many other things.

Publié : 01/08/2019 12:08 am
CybrSage
(@cybrsage)
Honorable Member
RE: I challenge Prusa to use MMU2S
Posted by: nikolai.r
Posted by: bobstro
Posted by: CybrSage

I wish we could get to the sensors menu while it was doing loading and unloading. That would help a lot with troubleshooting.

Might the terminal output from Octoprint not provide that info? Not the most readable, but a script might be doable. What specific sensor info do you need?

Best solution what I've seen so far was the led on top of the sensor. Easy mod and always good visible status of this sensor.

I will probably go that route unless Prusa surprises me in the next firmware release (hint, hint, Prusa).

Publié : 01/08/2019 12:09 am
Scott
(@scottmg)
Active Member
RE: I challenge Prusa to use MMU2S
Posted by: CybrSage

I wish we could get to the sensors menu while it was doing loading and unloading. That would help a lot with troubleshooting.

That would be nice.  Even if it was just 3 characters on the screen (1/0 in the corner somewhere for each finda pinda ir).

Publié : 01/08/2019 12:13 am
CybrSage
(@cybrsage)
Honorable Member
RE: I challenge Prusa to use MMU2S
Posted by: Scott
Posted by: CybrSage

I wish we could get to the sensors menu while it was doing loading and unloading. That would help a lot with troubleshooting.

That would be nice.  Even if it was just 3 characters on the screen (1/0 in the corner somewhere for each finda pinda ir).

I like it.  P0,I0,F0  For Pinda IR Finda and 0 or 1 for off and on.  🙂

Publié : 01/08/2019 12:24 am
bobstro
(@bobstro)
Illustrious Member
RE: I challenge Prusa to use MMU2S
Posted by: CybrSage
I like it.  P0,I0,F0  For Pinda IR Finda and 0 or 1 for off and on.  🙂

And a really loud beep that can't be turned off! 

My notes and disclaimers on 3D printing

and miscellaneous other tech projects
He is intelligent, but not experienced. His pattern indicates two dimensional thinking. -- Spock in Star Trek: The Wrath of Khan

Publié : 01/08/2019 12:26 am
CybrSage a aimé
Casey
(@casey-2)
Estimable Member
RE: I challenge Prusa to use MMU2S
Posted by: CybrSage
Posted by: Scott
Posted by: CybrSage

I wish we could get to the sensors menu while it was doing loading and unloading. That would help a lot with troubleshooting.

That would be nice.  Even if it was just 3 characters on the screen (1/0 in the corner somewhere for each finda pinda ir).

I like it.  P0,I0,F0  For Pinda IR Finda and 0 or 1 for off and on.  🙂

I like that idea as well. Or an icon that inverts. Either way, more info (condensed ok), the better.
:: adds to Christmas wishlist. ::

Publié : 01/08/2019 12:42 am
CybrSage a aimé
gnat
 gnat
(@gnat)
Noble Member
RE: I challenge Prusa to use MMU2S
Posted by: bobstro
Posted by: CybrSage

I wish we could get to the sensors menu while it was doing loading and unloading. That would help a lot with troubleshooting.

Might the terminal output from Octoprint not provide that info? Not the most readable, but a script might be doable. What specific sensor info do you need?

I looked into that around another thread. Seeing the sensor values appears to be a LCD function only. No gcode appears to be available over the serial console.

MMU tips and troubleshooting
Publié : 01/08/2019 1:30 pm
Casey
(@casey-2)
Estimable Member
RE: I challenge Prusa to use MMU2S

Well... since he reported me... lol...

Hey Jakub:
This may be quite relevant to why I have had issues with the MMU2S having to re-live-Z on every print (even after it was in Prusa's shop for repair).... Please see attached.

In doing the tests with Taso, in the transcripts that I forwarded 3-4 times to Jennes now, the hotend heater was off, and the heatbed was 60c with an ambient temp of 76F (however that converts...).... That's totally insufficient to melt the PINDA holder during the test, so that is important.

Taso in chat determined the PINDA was not quite working properly. However, Jennes is not reading the transcript attached. I'm sure of it. A PINDA holder can't melt at that 60C temp, or if it could, then that's a massive flaw.

Taso did say, three times in the chat, that the PINDA should be replaced. Jennes, Sr tech, is basically trying to say that the reason why it's off is the PINDA is melting during the test...... but at 60C????

Thoughts Jakub?

Attachment removed
Publié : 01/08/2019 2:32 pm
gnat
 gnat
(@gnat)
Noble Member
RE: I challenge Prusa to use MMU2S
Posted by: Scott

3.  We need a better solution for purging. 

My experience with the purge tower has been less than great, with some filament combinations not sticking well and creating a mess.  Although in retrospect that might have been related to a partially clogged nozzle I had. 

Lately I've been editing the gcode and entering my own purging solution, but that's going to be a no-go on prints with more than a few dozen filament changes. 

4.  Somehow it all needs to be faster.

My latest issue has been around some thin features next to supports.  It requires several filament changes, which equal lots of time.  The thin structures cool, and inter-layer adhesion is abysmal (PETG).  When I skip the supports and print faster, it's incredibly tough, all with the same settings.

I really should save a link to the thing since I will need it sooner or later, but there is an excellent post-processing script documented in a thread around here ("Real multi material printing" or something like that) which properly adjusts temp changes for switching between material types which produces much better prints. I believe someone is working on incorporating this functionality directly into PrusaSlicer itself. If so, hopefully it will get merged into the base code.

The KS is closed, but as part of their upcoming SimplyColor3D application, BigBrain3D is also including a bucket design to improve the purge process. Other than videos, however, they haven't supplied much detail on the bucket (as far as a comparison of waste and time): https://www.kickstarter.com/projects/bigbrain3d/simplycolor3d-every-3d-model-multicolor/description

Finally, @vintagepc and I have been working on our own post processor that will remove the purge tower from the gcode and replace it with using a bucket for tool changes. It still needs work, but we've gotten as far as being able to print a 2 color Benchy though while the reduced purge volume was encouraging (<20g vs ~40g) there are speed improvements to make as it added an extra 3 hours to the print 😣 Unfortunately my printer is down while I'm waiting on parts to rebuild my extruder so I haven't been able to debug/tweak the code. In a perfect world I'd like to see the functionality exist in PrusaSlicer, but while I can read C++ it is not one I'm that comfortable coding in. https://gitlab.com/gnat.org/originalprusa/tree/master/devices/mmu2/bucket/tool/3DBenchy

It would indeed be nice if Prusa supplied such functionality themselves, but the great thing about open source hardware and software is that when it is paired with a community full of talented and driven people we can fill in the gaps for things that Prusa hasn't thought of or doesn't have time/resources to get to.

MMU tips and troubleshooting
Publié : 01/08/2019 2:32 pm
gnat
 gnat
(@gnat)
Noble Member
RE: I challenge Prusa to use MMU2S
Posted by: Casey/NZ2O

Taso in chat determined the PINDA was not quite working properly. However, Jennes is not reading the transcript attached. I'm sure of it. A PINDA holder can't melt at that 60C temp, or if it could, then that's a massive flaw.

Taso did say, three times in the chat, that the PINDA should be replaced. Jennes, Sr tech, is basically trying to say that the reason why it's off is the PINDA is melting during the test...... but at 60C????

I've seen a thread somewhere (maybe in the MK3 section?) about doing the bed leveling and temps. I didn't pay too much attention as I'm not having issues, but I seem to recall this was chalked up to firmware bug. Maybe dig into that.

As far as the part melting, you are already in the relevant thread on that subject. Early on people were blaming it on users printing high temps and in enclosures, but that's been found not to be the case. Honestly I wouldn't have thought it possible until after about 40 hours of printing PLA I had to tear down my extruder to fix a jam and found my fins melted and deformed. Another 100 hours later (cause I didn't have new parts on hand so I had to reinstall what I had) and I had to snip off the fins because they were drooping enough to start being a concern. As far as I know my PINDA mount is OK, but I'll find out this weekend when I rebuild it (I do have replacement parts on hand this time). So while I agree that if you are printing PLA temps that you shouldn't have an issue, don't dismiss it until you've been able to confirm.

MMU tips and troubleshooting
Publié : 01/08/2019 2:42 pm
Casey
(@casey-2)
Estimable Member
RE: I challenge Prusa to use MMU2S
Posted by: gnat
Posted by: Casey/NZ2O

Taso in chat determined the PINDA was not quite working properly. However, Jennes is not reading the transcript attached. I'm sure of it. A PINDA holder can't melt at that 60C temp, or if it could, then that's a massive flaw.

Taso did say, three times in the chat, that the PINDA should be replaced. Jennes, Sr tech, is basically trying to say that the reason why it's off is the PINDA is melting during the test...... but at 60C????

I've seen a thread somewhere (maybe in the MK3 section?) about doing the bed leveling and temps. I didn't pay too much attention as I'm not having issues, but I seem to recall this was chalked up to firmware bug. Maybe dig into that.

As far as the part melting, you are already in the relevant thread on that subject. Early on people were blaming it on users printing high temps and in enclosures, but that's been found not to be the case. Honestly I wouldn't have thought it possible until after about 40 hours of printing PLA I had to tear down my extruder to fix a jam and found my fins melted and deformed. Another 100 hours later (cause I didn't have new parts on hand so I had to reinstall what I had) and I had to snip off the fins because they were drooping enough to start being a concern. As far as I know my PINDA mount is OK, but I'll find out this weekend when I rebuild it (I do have replacement parts on hand this time). So while I agree that if you are printing PLA temps that you shouldn't have an issue, don't dismiss it until you've been able to confirm.

Appreciate that, but they DID commit to replacement of the old version PINDA in there a few times. Would like to rule it out, but unfortunately, data is telling me otherwise.
More or less, trying to get their attention with that chat screenshot, because it's a bit crazy. 

The repeatability tests in Pronterface, if you review the screenshot/chatlog, were definitely performed below any known melt temp for PETG/PINDA holder, so that rules out the fin-melting being related to this one... 60C on the bed and 0C on the hotend is not a high enough temperature to cause it, at all.

Ce message a été modifié il y a 5 years 2 fois par Casey
Publié : 01/08/2019 2:47 pm
gnat
 gnat
(@gnat)
Noble Member
RE: I challenge Prusa to use MMU2S
Posted by: Casey/NZ2O

The repeatability tests in Pronterface, if you review the screenshot/chatlog, were definitely performed below any known melt temp for PETG/PINDA holder, so that rules out the fin-melting being related to this one... 60C on the bed and 0C on the hotend is not a high enough temperature to cause it, at all.

Ah. I wasn't getting that they were trying to relate it to active melting while doing the leveling. I haven't seen a documented case of active melting at any temp. It's more of a gradual/cumulative thing.

Have you leveled your bed at all? I bought my printer pre-built and once I started understanding more about all this I found that there was a 2mm swing from one corner to the other. When I installed the MMU I applied the nyloc mod for bed leveling and built a little tool to help run the leveling process. In doing it (same as you where only the bed was warm) I was seeing swings between tests even when I didn't make adjustments. What I ultimately decided on was that it was just very minor temp changes in the bed that were causing the issue. When I hit the bed with an infrared thermometer I found that the temp could be almost a full degree off what the display read and it took it getting a full degree under before it seemed to work to raise the temp back up.

MMU tips and troubleshooting
Publié : 01/08/2019 3:03 pm
Casey
(@casey-2)
Estimable Member
RE: I challenge Prusa to use MMU2S
Posted by: gnat
Posted by: Casey/NZ2O

The repeatability tests in Pronterface, if you review the screenshot/chatlog, were definitely performed below any known melt temp for PETG/PINDA holder, so that rules out the fin-melting being related to this one... 60C on the bed and 0C on the hotend is not a high enough temperature to cause it, at all.

Ah. I wasn't getting that they were trying to relate it to active melting while doing the leveling. I haven't seen a documented case of active melting at any temp. It's more of a gradual/cumulative thing.

Have you leveled your bed at all? I bought my printer pre-built and once I started understanding more about all this I found that there was a 2mm swing from one corner to the other. When I installed the MMU I applied the nyloc mod for bed leveling and built a little tool to help run the leveling process. In doing it (same as you where only the bed was warm) I was seeing swings between tests even when I didn't make adjustments. What I ultimately decided on was that it was just very minor temp changes in the bed that were causing the issue. When I hit the bed with an infrared thermometer I found that the temp could be almost a full degree off what the display read and it took it getting a full degree under before it seemed to work to raise the temp back up.

Once I get the PINDA properly replaced/support woken up/reading the chat that they claimed to not be able to find, I might do the nyloc mod and hyperfine tuning and such, depending on what the real level looks like.

For purposes of dealing with support, I'm avoiding any and all mods for now. After all, the Sr Tech seems to have freaked out from me printing parts for the motor mounts with a different color PETG earlier... Tried to call me out for mods that weren't there, even when it was in their shop. lol... Motor mounts are clear PETG. Maybe they mistook it for polycarbonate? 😉

Publié : 01/08/2019 3:18 pm
Page 6 / 10
Partager :