Filament cut too short at end



  • Avatar
    Oscar Pruitt

    Chroma and Simplify 3D worked on the first print, although the transitions were progressively worse each purge.  The first was about 50/50, the second was maybe 60/40, the third was almost all the next color and barely any of the previous color (started printing the next color before the purge happend so the transition was too soon) and the last transition happened way before the purge, like 20mm before (rough guess).

    The second print I reversed the colors in S3D and sliced it then prepped it in Chroma.  The first two purges were ok but the second two were well after the transition printed on the key chain.

    Here are my pings from the second print.  Since the last two were ok I went ahead and saved it.  The first print did not ask me to save.

    Ping 1 70.55%
    Ping 2 100.04%
    Ping 3 116.26%
    Ping 4 102.88%
    Ping 5 102.88%

    I'm going to try that last one again and see if the pings are better now after I saved it.

    Comment actions Permalink
  • Avatar
    Oscar Pruitt

    Here are my pings this time and the third purge had some of the previous but mostly the next color.  The last purge is still after the transition.  But it's getting closer.

    Ping 1 69.27%
    Ping 2 98.73%
    Ping 3 115.33%
    Ping 4 102.09%
    Ping 5 102.09%

    What I'd really like to know is why did Canvas ignore my Bowden tube length.

    Comment actions Permalink
  • Avatar
    Jonny Yeu

    Hey Oscar, thanks for sharing all this information, it certainly helps to determine what may be the issue. There's a few things I'd like to check:

    • Which printer are you using? Are you using PLA?
    • Concerning the Bowden tube length, this adds a length of filament at the end of your print to ensure that your extruder is able to feed filament to your hot end. Because of this, it's better to overestimate this length than underestimate. We usually measure from the filament feed to the tip of your hot end, then add a few mm.
    • Taking a look at your pings, the first ping is always very low. This can sometimes be caused by the nozzle sitting too close to the bed, so re-leveling the bed should help.
    • Can you please share pictures of a recent keychain print along with the bottom of its transition tower?
    Comment actions Permalink
  • Avatar
    Oscar Pruitt

    Tried the earbud case per the troubleshooting guide and not only were the colors horrific but it ran out of filament before the last layer even started.  I think I still had 2 layers to go. 

    I expected the colors to be off, but they were barely on.  Pings were great before the transitions started.  I did get a buffer error because the short transitions and my 4mm/s purge.  Slowing down purge to see if it helps.  Chroma says 24 ping but I only got 18.  

    Ping 1 99.25%
    Ping 2 99.41%
    Ping 3 99.85%
    Ping 4 100.57%
    Ping 5 100.7%
    Ping 6 100.3%
    Ping 7 99.78%
    Ping 8 99.78%
    Ping 9 99.78%
    Ping 10 86.67%
    Ping 11 86.67%
    Ping 12 73.47%
    Ping 13 73.47%
    Ping 14 73.47%
    Ping 15 73.47%
    Ping 16 73.47%
    Ping 17 73.47%
    Ping 18 73.47%

    Comment actions Permalink
  • Avatar
    Jonny Yeu

    Hi Oscar, are you using Side Transitioning when running the calibration prints as well? If so, please factory reset and re-print the keychain with a transition tower. Once that's done, please provide us with pictures and the additional information requested above.

    Comment actions Permalink
  • Avatar
    Oscar Pruitt

    The first few prints the nozzle was too close.  I adjusted that and it helped.

    I overestimated by around 40mm on the bowden tube length and set it to 620mm.  Chroma worked with the keychain just fine.

    I am using side transition.  I slowed the purge to 2mm/s and I have only hit the buffer switch 1 time that I noticed.  All colors seem to be on track now.  The transition is at around 40% of the purge so that is good.  I'm hoping this one finished and the issue from the last print was purging too fast.

    The second earbuds holder did not finish either, but the colors were better.  I will print the keychain with a tower and post pics.

    Here were the 21 pings it got (was supposed to be 24?)

    Ping 1 109.88%
    Ping 2 110.2%
    Ping 3 110.65%
    Ping 4 111.16%
    Ping 5 111.62%
    Ping 6 111.41%
    Ping 7 111.2%
    Ping 8 112.2%
    Ping 9 112.2%
    Ping 10 99.8%
    Ping 11 100.2%
    Ping 12 99.54%
    Ping 13 99.54%
    Ping 14 99.54%
    Ping 15 102.97%
    Ping 16 102.97%
    Ping 17 114.65%
    Ping 18 114.03%
    Ping 19 114.03%
    Ping 20 114.03%
    Ping 21 114.03%

    Comment actions Permalink
  • Avatar
    Oscar Pruitt

    It worked great with a tower and I set the Bowden tube length to 640mm.  The last transition was a tiny bit early, but that's expected before calibrating.  Why did it only work once from Canvas when I used side transitions and why would it ignore my bowden tube length until I put it 60 or more mm longer than it actually was?

    I printed 2 in a row from Chroma and they both finished with enough left to pull it back out.  But the headphone holder did not.  It did however start getting better after the first one.

    Ping 1 98.48%
    Ping 2 98.04%
    Ping 3 97.69%
    Ping 4 97.01%
    Ping 5 97.98%
    Ping 6 98.61%
    Ping 7 98.01%

    Comment actions Permalink
  • Avatar
    Oscar Pruitt

    Where do I go from here?  I want side transitions, but do I need to calibrate first then do side transitions or should side transitions not matter?

    Comment actions Permalink
  • Avatar
    Jonny Yeu

    Great! For now, please complete another 1-2 prints with a tower and share your pings. Once we have consistent pinging, you should be ok to switch to side transitioning.

    Comment actions Permalink
  • Avatar
    Oscar Pruitt

    Thanks! I did the 4 color keychain and t did great! The Bowden length is 640 now and I have about 20-30mm left to remove what was left. Didn’t realize I had to use the tower during calibration. I was thinking the side purge would be the same. I forgot to download the pings but I got a good sized print going now and transitions and pings look great! Pings are between 99.93 and 101.74 right now. I’ll post an update in the morning.

    Thanks for your help and sorry for freaking out lol.

    Comment actions Permalink
  • Avatar
    Oscar Pruitt

    I was doing a 2 color tree frog and there were 40 pings when I got up this morning all between 98 and 100, but mostly 98.  All transitions were perfect and at around 30% of the purge block.  Unfortunately, shortly before I got to it, something happened and it ended up grinding the filament right at a transition and printed a layer or two in the air.  I suspect maybe a transition that was a bit wide or maybe a bend from inside the buffer (I've seen it happen) didn't get all the way into the nozzle or something.  I didn't want to risk saving bad data so I just stopped it, but up until that point it was all good.

    I'm trying something a bit smaller with a slightly higher temp that can handle that many retracts and possibly a squished or bent transition.  I think the key was to start with the tower to begin with then move over to side purges and see how it does.  I'll post the results shortly.

    All in all I am very happy with the P2SP :).  Compared to my Diamond five color nozzle, this thing is a piece of cake lol.

    Comment actions Permalink
  • Avatar
    Oscar Pruitt

    Well that print failed due to a splice breaking in the buffer.  It was doing so good too!

    Ping 1 100.44%
    Ping 2 100.82%
    Ping 3 100.85%
    Ping 4 99.96%
    Ping 5 99.99%
    Ping 6 100.43%
    Ping 7 100.49%
    Ping 8 100.32%
    Ping 9 99.9%
    Ping 10 100.1%
    Ping 11 100.29%
    Ping 12 100.12%
    Ping 13 99.95%
    Ping 14 99.73%
    Ping 15 100.01%
    Ping 16 99.61%
    Ping 17 99.64%
    Ping 18 99.95%
    Ping 19 99.86%
    Ping 20 99.8%
    Ping 21 99.69%
    Ping 22 99.96%
    Ping 23 99.67%
    Ping 24 99.34%
    Ping 25 99.89%
    Ping 26 100.96%
    Ping 27 100.9%
    Ping 28 100.44%
    Ping 29 100.45%
    Ping 30 100.55%
    Ping 31 100.17%
    Ping 32 100.1%
    Ping 33 100.08%
    Ping 34 99.82%
    Ping 35 99.91%
    Ping 36 100.13%
    Ping 37 100.46%
    Ping 38 100.35%
    Ping 39 100.27%
    Ping 40 100.51%
    Ping 41 100.65%
    Ping 42 100.65%
    Ping 43 100.65%
    Ping 44 100.65%
    Ping 45 0.0%

    Comment actions Permalink
  • Avatar
    Jonny Yeu

    Sorry to hear about the broken splice! If you're using older filament, a lot of users find that dehydrating the filament before use with Palette helps to avoid breakages. Otherwise, your calibration looks great!

    Comment actions Permalink
  • Avatar
    Oscar Pruitt

    Yea I tested the yellow and it is a bit brittle in some areas.

    Comment actions Permalink
  • Avatar
    Oscar Pruitt

    lol at this rate I wont need the end filament to be long enough.  I swapped yellow out for some newer green that has plenty of flex in it.  We'll see how it turns out in the morning.

    It was breaking at the splice, but not on the splice.  The yellow was snapping right at the splice when the buffer got too low.  All the pieces that I have pulled out from all the failed tests had strong splices, but yellow would either give or just snap right at the splice.

    On the bright side, All the transitions and colors were perfect and pingswere between 99 and 100.5 until the filament broke.  I have a good feeling about this next print.

    Comment actions Permalink
  • Avatar
    Oscar Pruitt

    I ordered a dehydrator as this seems to be happening with most of my filament.  Although I have seen the filament really close to the buffer switch several times.  This can put a strain on even dry filament.

    Any way to make sure the buffer stays more full and does not pull the filament so tight?  It's not that the buffer can't keep up, it's that the Palette does not feed the buffer for a while.  I've sat there watching it.  Now it could be that it is letting the buffer get small in prep for a splice since it feeds a batch into the buffer at one time before the splice.  But it doesn't feed too much as to overload the buffer.

    Or is it hitting the buffer switch for a pong?

    Comment actions Permalink
  • Avatar
    Oscar Pruitt

    Its not quite done printing, but the Palette is done and I saved the job.  Now to see if it left any at the end.  Here are the pings:

    Ping 1 101.8%
    Ping 2 102.19%
    Ping 3 101.79%
    Ping 4 100.81%
    Ping 5 100.61%
    Ping 6 101.01%
    Ping 7 101.01%
    Ping 8 101.03%
    Ping 9 100.99%
    Ping 10 100.97%
    Ping 11 101.02%
    Ping 12 100.92%
    Ping 13 101.37%
    Ping 14 100.93%
    Ping 15 100.6%
    Ping 16 101.21%
    Ping 17 101.27%
    Ping 18 101.11%
    Ping 19 101.07%
    Ping 20 101.26%
    Ping 21 101.36%
    Ping 22 101.49%
    Ping 23 101.57%
    Ping 24 101.46%
    Ping 25 101.45%
    Ping 26 101.52%
    Ping 27 101.49%
    Ping 28 101.5%
    Ping 29 101.65%
    Ping 30 101.56%
    Ping 31 101.65%
    Ping 32 101.28%
    Ping 33 101.01%
    Ping 34 100.81%
    Ping 35 100.8%
    Ping 36 101.1%
    Ping 37 100.69%
    Ping 38 100.56%
    Ping 39 100.61%
    Ping 40 101.02%
    Ping 41 100.74%
    Ping 42 100.52%
    Ping 43 100.91%
    Ping 44 101.03%
    Ping 45 101.12%
    Ping 46 100.59%
    Ping 47 100.59%
    Ping 48 100.84%
    Ping 49 101.07%
    Ping 50 101.02%
    Ping 51 100.74%
    Ping 52 101.28%
    Ping 53 100.94%
    Ping 54 100.62%
    Ping 55 100.89%
    Ping 56 100.98%
    Ping 57 101.06%
    Ping 58 100.75%
    Ping 59 100.64%
    Ping 60 101.02%
    Ping 61 100.9%
    Ping 62 100.19%

    Comment actions Permalink
  • Avatar
    Jonny Yeu

    Hey Oscar, thanks for sharing this information, the print looks great and you're very well calibrated. If you're using a Hub with your setup, there's a Slowdown During Splice option when you turn on Advanced Options in the P2 plugin which will slow down the print speed when a splice is occurring in order to keep the buffer from collapsing. 

    It doesn't pong every time the buffer switch is triggered, but it does tell Palette to create more filament. If we sense that the buffer is too small and a splice is coming up, Palette should also fill the buffer to help avoid any collapses.

    Comment actions Permalink
  • Avatar
    Oscar Pruitt

    I found that option this weekend.  I had nothing else to do lol.  It's such a hard life being stuck at home with nothing but a tv, laptop and 3D printers :D

    One last question before I move on with my testing and learning.  Why does the transition tower have a brim and how do I remove it?  Not a huge deal as I will be going to side purges soon.

    Comment actions Permalink
  • Avatar
    Jonny Yeu

    Great to hear! And we hear you, it's a new world that we're all adjusting to. Hope you're staying healthy and safe!

    We usually include a brim to help with bed adhesion, but I would check your Minimum Brim Count value under the Transitions tab:

    Comment actions Permalink
  • Avatar
    Oscar Pruitt

    It is set to zero.  But it's no big deal as I will be moving to a side transition.  

    Also, I was able to take all the failed attempt PLA and use the Palette to splice it all together and am printing a benchy now with all that waste.  So more calibration data and I'm only using leftovers from other prints :)

    Comment actions Permalink
  • Avatar
    Jonny Yeu

    Great to hear! Let us know if you ever have any other questions.

    Comment actions Permalink

Please sign in to leave a comment.