Multiple print failures  

Page 7 / 10
  RSS
bobstro
(@bobstro)
Illustrious Member
Posted by: @epzcaw

I do have a Raspberry Pi, but have done only a limited amount of work with it and not for some time now.  Interfacing it with the printer would be a whole new learning experience.  Maybe later, I will get back to you for advice.  Thanks

Setting up Octopi is literally a 10 minute process following the instructions. You mostly configure the wpa_supplocant.conf file. That's it. I was stunned at how simple it was. 

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...
Posted : 07/12/2019 3:21 pm
Tim
(@tim-m30)
Illustrious Member

But you forget the minor issues of whether or not settings like network wifi peer visibility are enabled or not; plus some of the other things like firewalls not letting traffic get through a device or are being redirected.   I for one would have to do a lot of work to point outside traffic to my Rpi for Octoprint to work as expected; and, I have yet to see the UI show up on my setup even though I've tried to make it work that way (local rather than remote).

I finally found a way to install a GUI shell, but still no Octoprint UI is showing up.  It's all command line background tasks so far.  Easy?  Not really.

This post was modified 8 months ago 2 times by Tim
It is always wise to get more than one opinion......
Posted : 07/12/2019 6:53 pm
bobstro
(@bobstro)
Illustrious Member
Posted by: @tim-m30

But you forget the minor issues[...]

None of which have a thing to do with Octoprint. If you've made your network difficult to use it will be difficult to use. I have forgotten nothing.

I finally found a way to install a GUI shell, but still no Octoprint UI is showing up.  It's all command line background tasks so far.  Easy?  Not really.

That doesn't sound like OctoPi. It is a pre-rolled image for the RPi with all features enabled on the first boot.

OctoPrint can be more difficult depending on your Linux skills and distribution.

The individual I intended my response for indicated existing skills with the Raspberry Pi, so the process will be familiar. Furthermore, it will help with diagnosing a persistent problem with a far more complicated and expensive device. Comments of "it's too hard" are unhelpful.

 

This post was modified 8 months ago by bobstro
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...
Posted : 07/12/2019 9:20 pm
epzcaw
(@epzcaw)
Trusted Member

Today I printed three of model Mk3_Buddy_150um successfully. There was one failure but this was due to filament jamming in the spool.  I printed them at 225 via “Tune” as suggested by Prusa adviser, and also because there were fewer frog failures among the 225 test prints (though this is probably not statistically significant)

I can’t do time-lapse video either with camera or with phone.  The phone does indeed do time lapse, but only provides 10 frames however long you record for.

After a few false starts, I managed to print this which took about 5 hours 20 minutes (not fully fettled!) At one point during the print, I noticed that the temperature had dropped to 215. I paused it and used “Settings” to get it to go up again. After that, the temperature remained with 2 degrees of 225 for the whole print.

If there is an intermittent temperature drop, then running at 225 has the advantage that even a 10 degree drop keeps it above the melting point temperature.

So where so I go from here?

I will carry on trying to print.  If there are no more failures, I will be very happy!

Re Raspberry Pi - if I could set this up easily to record the nozzle temperature, it would enable me to identify whether temperature drops are indeed the problem.  But I don't think this is a trivial job.  I will try and do a long print run in the daytime where I can use two cameras to view the display sequentially, taking one away to charge while the other takes a video.  Could be a long job and Christmas is coming with lots of other stuff to do....

But thanks again for suggestions/advice

Intermittent faults are the devil’s work.... 

A soft answer turns away wrath....
Posted : 07/12/2019 9:38 pm
bobstro
(@bobstro)
Illustrious Member
Posted by: @epzcaw

[...] I can’t do time-lapse video either with camera or with phone.  The phone does indeed do time lapse, but only provides 10 frames however long you record for.

Octoprint/Octopi sound ideal for troubleshooting. More below...

[...] After a few false starts, I managed to print this which took about 5 hours 20 minutes (not fully fettled!) At one point during the print, I noticed that the temperature had dropped to 215. I paused it and used “Settings” to get it to go up again. After that, the temperature remained with 2 degrees of 225 for the whole print.

The drop is indeed worrisome. Any possibility of a draft or other environmental issue? Using a silicone sock can help maintain temps, but usually more related to cooling fans that are misdirected onto the heater block kicking on.

[...] If there is an intermittent temperature drop, then running at 225 has the advantage that even a 10 degree drop keeps it above the melting point temperature.

True, but you are likely to encounter a lot of stringing and sagging at elevated temps. In general, with PLA you want to keep temps as low as possible for finish quality while still maintaining part strength.

[...] Re Raspberry Pi - if I could set this up easily to record the nozzle temperature, it would enable me to identify whether temperature drops are indeed the problem.  But I don't think this is a trivial job.  I will try and do a long print run in the daytime where I can use two cameras to view the display sequentially, taking one away to charge while the other takes a video.  Could be a long job and Christmas is coming with lots of other stuff to do....

Ignoring the increasingly annoying and unhelpful curmudgeonly comments by others, Octoprint provides exactly this feature. Mine's configured to maintain 30 minutes of temp history ... and I'm the middle of a 23 hour print, so can't reconfigure it ... but here's the graph it provides by default.

Not much to see and indeed my temps have been stable for the last 12 hours, but if you increase the graph span, you can see fluctuations graphed over time. This is all done via a straightforward GUI.

[...] But I don't think this is a trivial job.  I will try and do a long print run in the daytime where I can use two cameras to view the display sequentially, taking one away to charge while the other takes a video.  Could be a long job and Christmas is coming with lots of other stuff to do....

It sounds like you've got a lot going on, but if you're spending more time on tossing prints than succeeding, I still recommend giving the RPi with Octoprint a try. You can do it while a long print starts. If you've worked with the RPi before, it is a familiar process (download image, burn image to microsd, edit wifi config file, boot). It supports a camera out of the box, so if you've got a USB camera laying about, you'll be in business quickly. Trust me, I'm not making money pushing a product on you. Octopi has been a huge productivity gain for me.

On a related note, my 2 grown sons visited over Thanksgiving and we had an unexpected afternoon of fun playing old tabletop wargames that we enjoyed when they were much younger. I'm now printing up "toys" for Christmas myself in anticipation of their next visit, so I'm all about getting Santa's Factory in full production. I've launched and completed some of my more ambitious 24+ hour prints. Here's what easily would have been $250+ of miniatures back when we were pushing metal around the table:

Hardly top quality, but plenty good for the tabletop. I've been able to salvage a couple of large-scale prints when one part fails. The Cancel plugin is great when one of ten bits fails mid-print. OctoPi is extremely useful for busy elves!

This post was modified 8 months ago 2 times by bobstro
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...
Posted : 07/12/2019 10:23 pm
bobstro
(@bobstro)
Illustrious Member

Edit: Just realized I failed to mention the built-in timelapse feature provided by OctoPrint. If you've got a native RPi or USB camera, it works well enough for monitoring and timelapse purposes.

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...
Posted : 07/12/2019 11:21 pm
Tim
(@tim-m30)
Illustrious Member

Ignoring the increasingly annoying and unhelpful curmudgeonly comments by others [snip]

@bobstro - my comment was simple: while easy for some, like yourself, others may find Octoprint more challenging to set up than you did.  Yes, I finally got my setup working.  Yes, it does many things well.  Yes, you can control the printer remotely.  Yes, the video integration is good.  Yes, it does time-lapse.  That said, my installation took a bit of an effort - and like others have reported has crashed on me for no particular reason ... or maybe it didn't crash but appeared locked while it was waiting for something it wasn't telling me about, PINDA preheat delays maybe? 

But, easy?   Not the initial 'make it work' portion - it took some tinkering to get it to work well enough.  Though, my use case isn't plug it in, load some 3rd party IP-DNS clients downloaded from unknown users and sites to point at my ISP supplied DHCP issued IP address and internal web server the install creates, then allowing my router to redirect web traffic to a DMZ install I haven't really tested, then loading a few phone apps for accessing from public wifi hot spots (though I believe most phone browsers can do an adequate job if you can deal with tiny screens).  

I am pretty proficient in the networking world... and reasonably computer savvy...  I would not call the setup easy, especially when the computer skills of the recipient is unknown. Was it hard?  To me?  No, but I knew where to find the problems I encountered along the way.  

 

It is always wise to get more than one opinion......
Posted : 08/12/2019 3:34 am
bobstro
(@bobstro)
Illustrious Member
Posted by: @tim-m30
@bobstro - my comment was simple: while easy for some, like yourself, others may find Octoprint more challenging to set up than you did.  
And... So what? You had a hard time with it. Therefore it is useless for diagnosing the problem at hand? You know @epzcaw will fail equally hard? You're owed an apology by everybody recommending Octopi?
 
Your  experience in this and many other things are not universal. If someone has an RPi sitting around that might help with diagnosing a problem with an expensive and complicated process, why do you feel that insertion of your own diatribe against Octoprint is in any way helpful? 
 
I literally wrote "I still recommend giving the RPi with Octoprint a try" and you're now into a 2nd post railing against it. 
This post was modified 8 months ago by bobstro
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...
Posted : 08/12/2019 4:30 am
bobstro
(@bobstro)
Illustrious Member

As luck would have it... I got a little over-ambitious in factory mode and kicked off a 30+ hour print without thoroughly cleaning the bed. I noticed one part had failed about 2 hours in. Using Octoprint with the Part Cancel plugin, I was able to identify and cancel the offending part and salvage the rest of the print without worrying about loose bits being dragged about:

Well worth the effort in my book!

This post was modified 8 months ago by bobstro
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...
Posted : 08/12/2019 7:52 am
zoltan
(@zoltan)
Member Moderator

@bobstro

The drop is indeed worrisome. Any possibility of a draft or other environmental issue? Using a silicone sock can help maintain temps, but usually more related to cooling fans that are misdirected onto the heater block kicking on.

We are puting on @epzcaw too much. That drop could be the part of gcode which could  change after first layer. 😊 If he did the temperature increase manualy at the print start, one of the later gcode could change it. What I understood the set temperature droppoed not the actual one.

Let us keep the original subject here Octopi is very interesting, but then we fall in the situation "what was the issue?" again. 😀 

even an old man can learn new things 🙂
Standard I3 mk3s, FW 3.9.0, no closed box, sketchup , fusion 360, PrusaSlicer, Windows 10
PRUSA MINI FW 4.1.0, technical background...
Posted : 08/12/2019 8:05 am
bobstro
(@bobstro)
Illustrious Member

@zoltan

I do think Octopi and graphing the temp swings and other data during long unwatched prints is germane. It is a useful tool in these instances. @epzcaw has expressed interest in both timelapse and temp monitoring, both of which it provides using resources he has at hand. 

This post was modified 8 months ago by bobstro
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...
Posted : 08/12/2019 8:27 am
epzcaw
(@epzcaw)
Trusted Member

I greatly appreciate all the support and advice offered here. These sorts of forums show the best of human nature where people want to help complete strangers for no return. So thank you all again.

Now, with all your help, I think I have made a breakthrough.  After completing 3 “Buddy”s successfully, I decided to try and print my own stuff again.  I printed two very small pieces, and started on the animal on the plinth for third grandchild.  This was a four and a half hour print.  I set up my camera to video the control display. , I set the printing temperature to 225 About ten minutes into the print.  I noticed that it was no longer printing - and the actual temperature was now 210

 I paused the print, set the temperature to to 280, filament started pouring out, and I resumed printing. Lo and behold, it was printing again. 

I left it overnight, and this morning it had finished the job - see below - though one ear came away when I was removing support so maybe a printing hiccup there.  I examined the video, and found that the temperature had briefly dropped to 190, before going back up to 210 but only returned to 225 after I heated it up to 280, and made a second via “settings” for 225 nozzle temperature.

So something is causing the temperature to drop occasionally.  I don’t believe it is an environmental thing.  The printer is in a centrally heated room with the temperature of about 22C, there are no significant draughts. The temperature drops somewhat after midnight and many prints have completed successfully during this time.

So my questions now are  what is causing this and how can I remedy it.  I assume it an internal hardware problem.  The demand temperature remains correct so I don’t think it can be software.

Since early intervention appears to be able to rescue the print, the raspberry pi could be a solution, where the temperature is monitored and an alarm sounded so I can intervene.  If all else fails, I might try this, but this will have to wait until the New Year.

However, this does not seem to be a practical long-term solution, so would like advice about what else there is to be done.

Tim, here is an image of the block and nozzle as you requested. 

Thanks again  all.

PS I am a "she".

A soft answer turns away wrath....
Posted : 08/12/2019 11:44 am
zoltan
(@zoltan)
Member Moderator

@epzcaw

Let me express my great compliment. To have lady as counterpart with such technical knowledge is rising my respect. 

From the previous I understand that the temperature dropped under the set (desired) one.

But be aware the Prusaslicer has option and using it to set different temperature fornthe first layer (usually higher) and for the rest of the print.

Nevertheless when you had the actual temperature 210°C and the set 225 there is something wrong with the heating apparat so we are probably back to my point 4.

Which can be thermal resistance increasing during print or hw issue of the main board heating drivers.

This post was modified 8 months ago by zoltan
even an old man can learn new things 🙂
Standard I3 mk3s, FW 3.9.0, no closed box, sketchup , fusion 360, PrusaSlicer, Windows 10
PRUSA MINI FW 4.1.0, technical background...
Posted : 08/12/2019 2:52 pm
bobstro
(@bobstro)
Illustrious Member

Congrats on your persistence with this. At this point, I'd suggest making up some notes and contacting support via online chat when logged into the online store. Something is definitely amiss. The more complete your summary, the more quickly support will be able to move on past the basics. Pics and video will help. Not to be annoying, but the ability to repeatedly capture the failure in progress will be useful! 

When the temp drops, do you see it in the set or actual temp on the display? The distinction will be important (settings versus printer swings). Also find as small a part as you can to trigger the failure. Are they all at the same height? 

This post was modified 8 months ago by bobstro
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...
Posted : 08/12/2019 3:13 pm
bobstro
(@bobstro)
Illustrious Member

More thoughts...

Look for consistency if there is any. Do they fail at similar heights? Similar durations? Similar features (e.g. point ears).

Will a simple tall block trigger a failure?

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...
Posted : 08/12/2019 3:22 pm
zoltan
(@zoltan)
Member Moderator

Few words about the heating chain:

Main board:

  • receiving from user (via wheel and button) temperature change requests -this I think works
  • receiving from G-code temperature change commands - this should work as well
  • receiving from thermistor the actual temperature value - as it shows even the drift beween the set and actual temp. I guess works too
  • connectors - could be potentional bottleneck - needs to be checked they are well connected  if they could  not loose during print
  • controlling the heating block current via power drivers (mosfet amplifiers) - questionmark - after longer operation it can get warmed up and start to work  as not required - this could be rather expensive as it requires the whole mainboard replacemet - having the printer under warranty worth to discusse with support guys, 

I left it overnight, and this morning it had finished the job - see below - though one ear came away when I was removing support so maybe a printing hiccup there.  I examined the video, and found that the temperature had briefly dropped to 190, before going back up to 210 (Edit question: just to be sure at the temperature drop the requested temp. was still 225°C?) but only returned to 225 after I heated it up to 280, and made a second via “settings” for 225 nozzle temperature. this should be observed a bit more and discussed with support sending them the video, could prove my theory, the heating drivers work on their edge and at the moment the fans start cooling on the full throttle it cannot heat powerfully

Cabling :

  • broken cable between the main board and hotend heater (cheapest damage only £ 6.23 + delivery costs hm, probably higher than the spare part)

Physical heat transfer 

  • between the hotend heater and heating block
  • betwen the heating block and the nozzle
  • this is related to assembly and way they are attached to each other - could be observed when you request the pre heat from ambient temperature (let´s say your 22°C) to see how much time it takes to get the required temperature ( let us say 225°C) and ask support if this is within normal threshold - if you keep us informed we can check against our printers

Gentlemen, any comments or other opinions?

 

 

This post was modified 8 months ago 3 times by zoltan
even an old man can learn new things 🙂
Standard I3 mk3s, FW 3.9.0, no closed box, sketchup , fusion 360, PrusaSlicer, Windows 10
PRUSA MINI FW 4.1.0, technical background...
Posted : 08/12/2019 4:11 pm
epzcaw
(@epzcaw)
Trusted Member
Posted by: @bobstro

Congrats on your persistence with this. At this point, I'd suggest making up some notes and contacting support via online chat when logged into the online store. Something is definitely amiss. The more complete your summary, the more quickly support will be able to move on past the basics. Pics and video will help. Not to be annoying, but the ability to repeatedly capture the failure in progress will be useful! 

I have had many online chats.  The last one was on Thursday.  The second last instruction was to try printing at 225, and the last was to email them if I had any further problems.  I emailed them on Friday with the results of my frog tests but have not had a reply yet.

When the temp drops, do you see it in the set or actual temp on the display? The distinction will be important (settings versus printer swings). Also find as small a part as you can to trigger the failure. Are they all at the same height? 

It is the left hand temperature reading which drops - this is either the actual temperature or the temperature reported by the temperature sensor.

The failures normally happen well into longer prints and usually well into the print but not at any specific time or height as evidenced by the frog series.  The near-failure that I intercepted in the last print occurred at about 10 minutes, but failures have also minutes away from the end of 5 hour prints.

 

 

A soft answer turns away wrath....
Posted : 08/12/2019 5:44 pm
epzcaw
(@epzcaw)
Trusted Member
Posted by: @bobstro

More thoughts...

Look for consistency if there is any. Do they fail at similar heights? Similar durations? Similar features (e.g. point ears).

Will a simple tall block trigger a failure?

There is no consistency that I can see except that failure usually occurs after an hour or more, but can be much less than that - see previous raply.  I ahven;t printed nay tall blocks. It looks to me like a random error. A very coarse estimate of the mean time is 100 minutes.  Would need a lot more tests to find a better value, and I a bit reluctant to waste more plastic and power, given current concerns about the environment!!

 

 

A soft answer turns away wrath....
Posted : 08/12/2019 5:48 pm
bobstro
(@bobstro)
Illustrious Member
Posted by: @epzcaw
The failures normally happen well into longer prints and usually well into the print but not at any specific time or height as evidenced by the frog series.  The near-failure that I intercepted in the last print occurred at about 10 minutes, but failures have also minutes away from the end of 5 hour prints.

Try slicing the part in PrusaSlicer, select the preview mode and see if there's any commonality at the layers in which failures occur. We had instances with people encountering failures at fine details like fingers where lots of retractions occur. You can turn on the feature to see retractions which may help.

This doesn't explain your temperature swings though. Do these failure occur printing the pre-generated gcode file samples provided on the SD card? That would eliminate any of your settings from consideration. The left-hand temp value shows the settings value, so if that changes while printing the sample gcode, something is way off.

Apologies if you've already done some of these steps. Just trying to narrow down the range of problems.

 

This post was modified 8 months ago by bobstro
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...
Posted : 08/12/2019 5:51 pm
epzcaw
(@epzcaw)
Trusted Member
Posted by: @bobstro
Posted by: @epzcaw
The failures normally happen well into longer prints and usually well into the print but not at any specific time or height as evidenced by the frog series.  The near-failure that I intercepted in the last print occurred at about 10 minutes, but failures have also minutes away from the end of 5 hour prints.

Try slicing the part in PrusaSlicer, select the preview mode and see if there's any commonality at the layers in which failures occur. We had instances with people encountering failures at fine details like fingers where lots of retractions occur. You can turn on the feature to see retractions which may help.

This doesn't explain your temperature swings though. Do these failure occur printing the pre-generated gcode file samples provided on the SD card? That would eliminate any of your settings from consideration. The left-hand temp value shows the settings value, so if that changes while printing the sample gcode, something is way off.

Apologies if you've already done some of these steps. Just trying to narrow down the range of problems.

The frog test was done using a Prusa g-code.  I did 5 with Prusa settings and had 4 failures.  I did  5 with the temperature set at 225, as recommended by my  Prusa adviser, and had 2 out of 5 failures.  People have commented that because the frog has 50um layers, it is more likely to fal, but given that it si a Prusa sample code, I would expect it not to be a problem.

I did print 3 "Buddy" from teh sample st successfully but set the temperature at 225.  Perhaps I might try another run at Prusa settings.

 

A soft answer turns away wrath....
Posted : 08/12/2019 6:01 pm
Page 7 / 10
Share:

Please Login or Register