Palette 2 Jamming

Comments

61 comments

  • Avatar
    Redemptioner1

    Yeah, I have started running into this issue again but I am unsure if it is related to my issue with the scroll wheel not staying within calibration

    0
    Comment actions Permalink
  • Avatar
    dkoloms

    My setup is 

    Slic3r PE ver 1.41.3

    Chroma 3.1.3

    Palette 2 Pro in connected mode.

    Only printing in PLA using stock splicing settings.

     

    This  happens at the end of the print, when the P2 is done with splicing, There is more filament left over than normal and there are splices in it even though there is very little left to print of the object.  Last time I had this problem it occurred when there was only a little left of a single color to print, however there was extra filament with multiple splices remaining.  On one file I tried it multiple times with different brands of filament, and got the same  jam at the same place in the print both times.

    While the print is in process all the pings are close varying by a % or so at the most, until it starts reporting missed pings when  everything jams up.  Pongs are always ~99.5 or so.

    I am not sure if this is a P2 issue or more of a slicer/chroma issue since there is extra filament with splices produced and it is reproducible with a specific file.  I have not tired it  with Canvas or another slicer.

     

    Debra 

    0
    Comment actions Permalink
  • Avatar
    Manmeet S

    Thanks for the additional context. Missed pings would explain why you see spliced filament leftover in the buffer/tube as Palette 2 as the filament consumption is not being tracked accurately. Is it safe to assume you are running a Prusa printer? If you could share a photo of the print along with the ping data that would be helpful. A few immediate things I would recommend

    1) Double check that your teflon tube entering your extruder is secure at different points on your bed and heights
    2) Try out the "slowdown while splicing" feature through CANVAS Hub as there is a chance your print speeds up during that point in the print outrunning the Palette 2 momentarily. Enabling this would guard against that situation.

     

     

    0
    Comment actions Permalink
  • Avatar
    Redemptioner1

    @ Manmeet - I think you are interpreting the details wrong. The missed pings are starting when the palette starts to jam, there is left over filament because the palette jammed internally for no go reason, it is not a case of the missing pings start then it jams.

    To your point about slowing down the print, this is most definitely not the answer (don't use Johnny's default answer for everything), clearly if the filament is getting jammed in there it has to be in there for some time getting melted against something to make it stick. If it was out running the buffer then the filament would continue to feed out once the splice finished and the buffer was restored, only there would be a skipped layer/section in the model as a result (which has not been mentioned by anyone). 

    I would suggest you replace the Teflon tube in the splice core, they only seem to last about 3000-5000 splices before they wear enough to allow the splice to stay stuck to the inside of the Teflon tube (seems to start happening within 1000 splices), or the tube diameter wears enough to allow the diameter of the filament at the splice to become larger than the gap on the back side of output extruder causing the splice to get jammed and then the output extruder continues to run (as there is no feed back between scroll wheel and extruder to ensure filament is still being fed out) which results in the filament being friction melted into the extruder and filament pathway. The only way then to get the filament out is to remove a lot of parts and use a bunch of force.

    0
    Comment actions Permalink
  • Avatar
    Redemptioner1

    Also, still waiting on what I should be doing next (short of a paypal claim).....?

    0
    Comment actions Permalink
  • Avatar
    dkoloms

    I will try,  and set slowing down.  However I have to agree with Redemtioner1, the missed pings do not start until the Palette jams.

    I do not have the ping data from the failed print to share and I tossed the failed print.  If this occurs again I will post the data and a picture

     

    Debra

    0
    Comment actions Permalink
  • Avatar
    Manmeet S

    Sorry did not mean to imply that the missed pings caused the jam in some way if that’s the thought, I stated that to give an idea of why leftover splices might still be in the buffer/outgoing tube. It's certainly possible that a communication error may have occurred where the buffer switch triggering was not registered.

     

    When it comes to not being related to speed while what you state is true for most cases you still have the situation of outrunning a buffer temporarily which could lead to the filament stripping in a way that the extruder gear cannot grip it and move it forward causing you to print in the air. Unless it was in the process of a pong or a splice the outgoing of the Palette 2 will not start forcing more filament out and new filament would not be made available for the extruder to grip onto. Is that the problem for sure? Maybe not but it rules out one variable so we can get to the bottom of this and get back to printing. This is not a default answer specific to Jonny, we all share a common knowledge of what works for our users and what does not if we find certain things have helped in previous cases we will recommend them if we believe it will either solve the issue or help us isolate the issue at hand. If he is recommending something its because he has reasoning behind it.

     

    Before changing the Teflon tube can we actually say this was a jam? Debra, were you able to use the Clear Palette command after cancelling the print or did you have to remove the splice core? If you do run into this issue again Debra could you please take a shot of the print/pings and send your Octoprint log file our way (support@mosaicmfg.com) so we can take a deeper look.

     

    If you can give me a bit more context behind your situation Redemptioner we can work towards fixing it.

    0
    Comment actions Permalink
  • Avatar
    dkoloms

    I was easily able to remove the filament from the Palette.  It was not stuck in the splice tube.  It just seemed to be pulled hard against the buffer switch.  I only had it occur on a couple of prints, and am currently printing without issues.  I haven't tried it again with the file I had an issue with (twice).  I am currently half way through am estimated two day print with 2233 splices and so far have had no issues with it. I do have slowing enabled in the Canvas hub.  I will see when I get to the end since this is where the issues seems to occur

    I appreciate your help

    Thanks

    Debra

    0
    Comment actions Permalink
  • Avatar
    Redemptioner1

    I can confirm in my case it is jamming in the palette and the buffer switch is coming on. I can also confirm that part of the issue is the teflon tube wearing out quickly (using plain PLA) which creates a significant difference in splice diameter making it exceed 2mm in diameter causing it to catch/jam in the filament path inside the palette. This is made worse by the small gap between splice core and output extruder, if the filament has a little bulge from the splice it gets caught/wears against the lip of the filament path just before the extruder. As you can't see the alignment here (nor can you adjust alignment top and bottom) so you can't get the filament path from the splice core perfectly aligned with the filament paths on either side of the splice core.

    Now the lead in filament path to the splice core could be argued is less of an issue until you take some time to investigate what happens here. In my case I noticed it can cause a tiny bend/kink at the very end of the filament about to be fed into the splice core. This results in the splice not being perfectly aligned and this results in an edge to the sidewall of the filament which is just perfect to catch the edge of the filament path (in the gap) as it exist the splice core. The single screw attachment for the splice core that is off centred prevents the alignment of the splice core and palette filament paths to be perfectly aligned. This is better on the pro version due to the aluminium being used for the core being much stiffer and more likely to be square and is probably why I have not had many "jamming" issue on the pro's.

     

    @ Manmeet - Generic responses shows a lack of care, attention and respect for your customers.......

    If it was just a case of the buffer running out, clearly the palette would no continue to run the output extruder so there would be no melting/jamming of the filament in the palette extruder nor would it wear the filament to a point where the palette extruder combined with the printer extruder would not be able to move the filament along. Now there is a possibility the extruder on the printer will grind enough of the filament away to then be unable to feed the hotend more filament (should not happen with any of the geared down extruders and would require speeds above about 200mm/s to have enough feed rate to cause enough wear on the filament to stop it being gripped), in this case the buffer would fill again without jamming but the print would still be a failure (even though the printer extruder can't push more plastic through, the palette would still replenish the buffer).

    I had another standard palette 2 jam last night, I will pull it apart (again) this avo and take some pictures and put them up.

     

     

    Red arrow shows where the extruder has melted the filament (hard to see because black filament) from grinding on the last splice stuck in the teflon tube/filament path, this has cause a melted section of filament to migrate back down the filament path out into the section above the extruder. This is a really bad design area of the palette as there is a gap between extruder and filament path which is made worse by the lower edge of the filament path starting at a point. You can see the black area of melted plastic which should not happen if the palette has some very basic safety features like making sure the scroll wheel is turning before running the output extruder on the palette. Obviously the issue causing the jam in the first place has taken place before the extruder (some where between extruder and splice core), but this additional crap shoot from the extruder additionally melting the filament after the extruder makes it very difficult to remove the jam. You have to remove extruder and splice core at same time while using something like alcohol to convince the melted plastic to let go of the plastic in the palette filament path. Enough to make you want to throw the palette 2 across the room.....

     

    0
    Comment actions Permalink
  • Avatar
    Manmeet S

    Hey Redemptioner,

     From reviewing your photos the problem your facing is related to one or a combination of these factors
    -Filament being used is out of spec (too thick to enter the splice core without resistance)

    -Poor quality filament (if you take a look at the red circle by the outgoing motor you can see you have a strange irregularity in the filament at a point which is not a splice

    -Splices that are 2.0mm thick will cause issues in the Palette 2 system and lead to jams so splice tuning these to have these properly in spec is necessary. The splice I circled in the image looks quite large and if something like that was to reach a drive gear I would expect it to grind away and melt as you have described. 

    Now, this isn't to say your filament manufacturer is at fault but I would ask that you look into this and confirm that is not the issue on top of splice tuning to get the resulting splices within spec.

    I understand that you may feel we provide generic responses but these are the questions that help us solve the problems our customers encounter as I have mentioned before. Sometimes a problem encountered by a user can be an oversight of a setting and other times it might be a more complex problem. With these questions, we are able to build a better picture of the problem and while they might be pointless questions to you they could make the difference in identifying a piece of information the customer did not think was important. So please work with us so we can work with you and get you back to printing.

    0
    Comment actions Permalink
  • Avatar
    Redemptioner1

    @ Manmeet

    Filament out of spec - Not the case, the experimental optical filament gauge says otherwise ans shows the 9 filaments I have tried all to be with +/-0.02mm at worse so well within tolerances for good filament. This is backed by the fact that rolls of filament that have worked no problem previous show the exact same problem with increasing pongs which is further supported by the fact multiple brands of filament all show the exact same issue. 

    Poor quality filament -  I think if you look closer you will find the point circled by you on the right is the curve plastic of the cover distorting the image so no issue there. As for the one circled on the left, this is normal as the palette forces a decent angle change straight after the output extruder which can cause the filament to get a tiny bend in it, it is only really noticeable when printing fast as the filament is still a little warm when it comes out of splice core, warm enough to put a slight bend in the splice and I think you will find if you do some testing this is normal as all my palette's do the same thing 

    Splices that are 2.0mm  - The diameter of a splice is set by the internal diameter of the Teflon tube in the splice core, no amount of "splice tuning" will change the diameter of a splice, only the internal diameter of the Teflon tube. If a splice is coming out over 2mm it is because the Teflon tube is allowing this. Maybe I am unlucky and got a crappy Teflon tube insert for the splice core (something I have not changed out again since the last once changed about 7000 splices ago).

     

    The problem is not the filament, the problem is one of 3 things;

    - the Teflon splice core is not "non-stick" enough 

    - the internal diameter of the Teflon splice core tube is too large allowing for the splice to grow too much in diameter (remember the output tube is over 3mm in diameter to make up for splices being a little oversized to begin with) 

    - the Teflon splice core tube is wearing too easily causing it not only to grown in internal diameter but also to reduce the "non-stick" nature of the Teflon

    To be honest it could be a combination of all 3, but the main thing to note is the issue is happening in the splice core (splice getting stuck in there), I also forgot to mention that there is an audible sound (like a dully click) when the splice is finished and it goes to refill the buffer. The sound is very clearly the filament releasing from being stuck to the tube inside the splice core. All my Palette 2's make this sound, some worse than others (pro's make less sound and have not jammed....yet). 

    0
    Comment actions Permalink
  • Avatar
    Jim Neill

    So I just had this happen to me again and in this situation it had nothing to do with short transitions.  I printed the same exact print that was completely successful two days ago and this time the jam was so bad that I couldn't get the spliced filament out of the splice core, and actually during the process of trying the splicing tube yanked out with the filament.  It seems that the splice had swollen to a diameter that the tube couldn't handle or it spliced to the tube itself (unlikely). I ended up having to slice open the tube with a razor to get the filament out and inspect.  I'd provide measured diameter but my caliper's battery is dead right now (hate those small batteries, they die so quick). In my situation, the splicing tube only had 1611 splices.

    My biggest problem is that the firmware of the palette 2 doesn't detect this jam.  It should be fairly clear that the switch in the buffer is triggered for an extended amount of time.  In that situation, the firmware should notify the octoprint plugin/canvas hub to cancel or at least pause so that it doesn't continue to try and print in the jammed state.

    0
    Comment actions Permalink
  • Avatar
    Manmeet S

    @Redemptioner

     

    The optical filament measurement device sounds interesting is this something that is commercially available? I've seen quite a few of these available through the years but they all fall short in one way or another due to the oval nature of filament/the twisting of filament as it moves to your extruder. The plastic by the right splice doesn't quite cover the filament area I've circled so I don't think its distortion but let's assume it is and the filament is perfect.

     

    The diameter of a splice is set by the internal diameter of the Teflon tube in the splice core, no amount of "splice tuning" will change the diameter of a splice, only the internal diameter of the Teflon tube. 
    I get your point of view but that is incorrect and it's not going to help you to keep this logic. Yes if a splice tube has become too large splice tuning won't help it, we've never said otherwise and this is a stated reason to swap out your Teflon splicer tube in our replacement guide (http://mm3d.co/p2splicer-tube).

     

    However, splice tuning gives you control over three variables which will affect the diameter of your splice assuming your splice tube is in working condition. Think about heating a splice (having a high heat factor) and having a low cooling factor and a low compression factor, this means you would move the splice out of the splice core before it has had adequate cooling cause it to neck and become too thin. Conversely, imagine setting the compression factor value to be too high, in this case, you're pushing together both materials beyond what they need to be filling the splice tube beyond what it should so that when you pull out you now have a splice that is too large. These factors directly affect the diameter of the resulting filament. While the output tube is spec'd to be a bit larger you should not be exceeding 1.9mm in splice diameter or you will have issues. The dull sound your hearing is very likely the overly large splice struggling to exit the splice core and make it past the outgoing drive momentarily. So while we are in agreement that a splice tube that has become too large will not be helped by splice tuning to say splice tuning has no effect on filament diameter is not accurate. The two work hand in hand, splice settings + Teflon tube diameter it is not just reliant on the diameter of your splice tube. 

     

     

    0
    Comment actions Permalink
  • Avatar
    Manmeet S

    Hey Jim, was this the same filament you used with the successful print a few days prior? That does sound like a really low number to experience a jam what were your splice settings? That is a feature we are looking to implement but more based on the encoder value rather than the state of the buffer switch. Basically detecting that the encoder has stopped rotating and sending a pause command to stop the printer. Unfortunately, this functionality wouldn't be available in  Accessory mode.

    0
    Comment actions Permalink
  • Avatar
    Jim Neill

    It was the same exact filament and gcode. I've never tuned the splicing since receiving the palette because it typically always seemed to work. To be honest, I don't even know where to go to see the current settings.

    0
    Comment actions Permalink
  • Avatar
    Redemptioner1

    @ Manmeet 

    Yes they are commercially available, start from around $17K for a half decent dual measurement, or about $25K for the quads, not something I think we will see in general use in the 3D printing market for most people due to cost. The one I have been playing with is generally used for measuring fibre optic strands as they are made and is significantly more than the ones mentioned above. Oval shapes are a non-issue if you use multi laser measurements, in fact the one I am using calculates the volume of the material and not just the diameter, and no I have not tried to input the sensor output to a printer as it is not mine to keep :( 

    As for splice tuning, I stand corrected that tuning will have an effect on diameter, but only to make the diameter smaller, not larger. If you have a closer look at how the splice takes place you can see why the diameter of the filament will be the diameter of the inside of the tube if the splice is anywhere near ok. As you pointed out, not enough cooling and pull the filament out and it will neck-down the filament making the diameter smaller, but we are not talking about the filament diameter getting smaller so lets not muddy the water with talking about something that is not happening and has not been raise in this thread.

    Now for generally ok splice's the splice core heats the end of the filament that has already passed through cutter and sitting in splice core, it then takes the new filament from below and jams then end of the new filament into the melted filament in the spice core (the compression). This causes the new filament to be pushed up inside the end of the melted filament causing the melted filament in it's path to be pushed outwards till it hits the teflon tube and takes up its new larger diameter from the inside of the teflon tube. There is no other variance  in diameter that can happen here as the tube sets the new filament diameter, the length of increased diameter in the splice might be a little longer or shorter based on heat and compression settings but the diameter will not vary as it is set by the tube.

     

    So we are in agreement the issue is the teflon tube, so what is the deal with them, are the extra tubes provided made from a lower grade material or something.....? The new tubes are starting to have issue around 3000 splices which is not good (a couple of prints for me), with the tube degrading faster in the non-pro's, there is obviously an issue with the tubes residual heat after a splice not cooling fast enough on the non-pro's causing a few possibilities;

    - the additional heat causes tube to degrade prematurely making the internal diameter larger

    - the heat causes the tube to degrade prematurely causing the surface to lose its non-stick properties.

    - the heat causes the tube to soften too much which allows the tube to expand when the compression cycle takes place, the internal diameter of the teflon tube increases from the pressure pushing out on the teflon tube during the splice, the tube cools and is now wants to shrink down smaller than the spliced filament diameter causing it to jam

    Assuming that a bunch of us didn't somehow get a bad batch to teflon tubes (the Pro's did come later after all) and the issue is just overheating the teflon tube, is there a way to tune the splice core actual heat it reaches, or is that just a factor of time which means there would be now way to stop the telfon tube getting so hot?...admittedly this may reduce the types of filaments that splice "well".

    So the only real relationship between filament splice diameter and tuning in this situation is if the teflon tube is overheating and expanding during compression cycle, dropping the temp might reduce the impact but is unlikely to help as we will probably end up with under cooked splices between the filaments, to which case there is probably no fix other than using a different teflon mix for the tube if there is a need to maintain the amount of heat to melt the filament appropriately to begin a splice..

    0
    Comment actions Permalink
  • Avatar
    Manmeet S

    Yep a bit too pricey for 3D printers unfortunately but that sounds like a good opportunity to have access to one even if only temporarily. 

    My intent is not to muddy the water but to add to your knowledge as you seem like an individual who likes to learn about the devices you use. While it may seem logical for the inner diameter to be the maximum splice diameter you would ever experience the flexible nature of Teflon makes this invalid. You could have a splice tube with an inner diameter of 1.80 and with too much compression have a splice exit at 1.85 as the Teflon is able to flex. Overheating and overcompressing can lead to problems and we are looking at putting together a more thorough guide for how these should be used together in a more consistent manner. The splicer tubes that ship installed and the ones that are shipped as spares come from the exact same manufacturer and are specified in the same way. We don't have any desire to limit customers to certain brands of filament but at the same time we can't control the formulations certain brands use, some may require higher heat than others for proper splice production. I would ask that you try lowering your heat/compression to get a better resulting diameter exiting from your spice core. The temperature that the splice core achieves is not a variable that is tunable but instead how long it remains on/off which is what the heat factor controls. What are your current splice settings and brand of filament? It sounds like you are happy with your filament manufacturer and I can understand that you may want to stick with them but the reason we provide control over heat, cooling and compression is that we recognize not all filament formulations are the same. One brand of PLA could be very different from another. If you're up for it I can get you some extra tubes so you can work through finding ideal settings for that specific brand.
     
     

    0
    Comment actions Permalink
  • Avatar
    Redemptioner1

    So just to clear everything up, there is yet another design floor in this product;

    You are saying the issue is the teflon splice core tube (as expected), and this tube is not real stiff once it gets a little heat on it which allows it to be stretched creating a larger internal diameter to the teflon tube. Then dependant on the compression settings you can cause the tube to expand further increasing the filament diameter to a size greater than the rest of the Palette 2 can tolerate. Sounds like you need to correct the tolerance in the tube the heater wire goes round to house the teflon tube to prevent it expanding further than the rest of the palette can tolerate. It also explains why the metal splice core works better as it is pulling the temp down much faster and no doubt reducing how much total heat is put into the tube as well. 

    Yes I am very happy with the filament suppliers I use as I go through a lot of filament (and a lot of wasted filament thanks to the Palette 2) and I have never had a bad batch or any inconsistencies between runs. Realistically 90% (by volume) of major filament manufactures have very tight tolerances and produce amazing filament. The only real problems tend to come from the smaller third party companies altering the filament themselves and re-spooling it, they simply do not do enough volume of plastic, have the resources to buy the right production gear and don't have facilities anywhere near clean enough (i.e. printed solid) to produce filament at the same consistency as the big boys.

    0
    Comment actions Permalink
  • Avatar
    M Montminy

    At the risk of poking a beehive, I found this thread searching for a solution to my seemingly nonstop jamming. I've come to the conclusion that there seems to be something up with the teflon tube.

    I spent a few hours today running over 26 splice tunes. The largest splice was created was 1.88 using a compression of 3, just to see if I could a) cause a jam, and b) determine outliers. Neither happened, the majority of my splices were in the 1.8 range. That seems to be well within the 1.6 to 1.9 range referenced.

    Yet I still get pretty frequent splice core jams. I've replaced the splice tube once already, and between the two tubes combined, I don't think I have more than 3000 splices. I don't use the Palette as much as I'd like, because I can't trust it to print anything without jamming. Every time I think I've tuned it and I get a good print, it starts jamming again.

    I've had numerous exchanges with Manmeet and others in the Facebook group, but I'm still failing to get anything longer than a 2 hour print to not jam.

    It was mentioned earlier in this thread; "there is obviously an issue with the tubes residual heat after a splice not cooling fast enough on the non-pro's causing a few possibilities". The seems to resemble my experience. I say this because the likelihood I'll experience a jam increases with the splice frequency. If the splices are spaced well enough apart, the print has a higher chance of succeeding. So heat seems to be a factor. If I turn the heat in the splice tune down towards -4, my filament leaves the core with a barely audible "pop", if any. If I use default heat or higher, there's a very audible pop as the filament is clearly sticking to the teflon, even when it's a 1.8mm splice.

    Judging by the similar posts on Facebook, there appear to be others having similar issues, struggling to find the magic splice tune. But when all your tunes look alike, and within spec, you have to wonder if there's something else amiss. It almost feels like "splice tune" is the default response, when something else in involved.

    The last successful print I did took over a dozen failures and weeks of trial and error. I couldn't use splice tuning, as even the best measured and looking settings jammed. I just shotgunned it until something worked. I thought I had it, but I've now realized I don't, since if the print does more than 100 or so splices, it's still jamming, so I'm back to the drawing board.

    I went through the few jammed splices I still had around, and they were all different brands of filament, and all around 1.85 to 1.86mm. When you tune to a 1.8mm splice pretty reliably, and then a 1.86 shows up and jams, how do you deal with that?

    I'm suspecting that while the teflon tubes might be physically to spec, maybe there are thermal or "slippery" variations from batch to batch. 

    0
    Comment actions Permalink
  • Avatar
    Ed Williams

    M,

    Speaking purely from my experience, the only time I have ever gotten jams in the splice core was beacuse of subpar filament. It was kicking my tail for a few days, I thought I had bad splice settings or a failing core. To figure it out I simply took a new teflon tube and ran the suspect filament through it. Sure enough every couple of feet the filament would get tight and take alot of force to get through. I went and tried the same test on "good" filament (Protopasta, Matterhackers,Hatchbox) with no problems. The suspect filament in question was some old 3Dsolutech and Makergeeks. Good luck and let us know how it goes.

    0
    Comment actions Permalink
  • Avatar
    M Montminy

    The problem is "good" filament is subjective. I also don't feel it applies here, because every single time I've had a jam I've measured the filament, and it's never been out of spec. The filament didn't jam, the splice did. For what it's worth, I've had jams with Protopasta, Amazon, Prusament, Filamentum and others.

    I'm not here to bash the Palette, or complain. I bought it knowing full well that multi material printing can be a nightmare. I'm simply trying to suggest there might be something else affecting a number of users besides "splice tuning" or "bad filament". It's difficult to engage and diagnose when that's all people want to claim as the cause.

     

    0
    Comment actions Permalink
  • Avatar
    Ed Williams

    I completely understand. You very well might have gotten a bad set of tubes. Have you gotten any new ones from Mosaic yet? I am almost curious if your issue is not with your tubes but with the core itself. If you have put known good filament suppliers through itand changed out multiple tubes, there is still one more variable to test.

    0
    Comment actions Permalink
  • Avatar
    M Montminy

    I've not yet contacted support regarding new tubes or a possible core issue, I felt I needed more data first. It is entirely possible that some core are more prone to either running hotter or heat-syncing more than others, so it could be a core. I think it's also possible that batches of PTFE are more prone to softening than others. There are probably other variables as well.

    That was sort of my point behind chiming in here... to try to have a discussion beyond filament/tuning as to what might cause some of these outliers.

    I was finally able to get a 186 splice print to finish last night without jamming, oddly by using my crappiest filament, some Amazon basics. I'll try using those settings for a few other prints with different filaments and see how it goes. The main difference between the tune I had success with and the one I was using (that was suggested by Mosaic awhile ago), is the heat is 1 vs the prior 4.

    0
    Comment actions Permalink
  • Avatar
    M Montminy

    I'm not convinced I've solved it yet, but I've done 2 186 splice prints without a jam using different brands of filament, and I'm 14 hours and 300 splices into a larger print and still going. The audible "click" is also reduced when the splice pops free from the PTFE on the first move.

    My (admittedly limited) data set suggests that there's no linearity between a splice setting and the splice size. For example heat 0 and 2 resulted in bigger splices than 1 and 4, all other setting being equal. While I knew I couldn't approach tuning one setting at a time, I was expecting a little more linearity along the lines of 4 being too much, 2 being too low, so it must be 3, ruling out <2 or >4, and that doesn't appear to be true.

    I do think that there's some heat soaking happening that's affecting the PTFE over time, at least for some of us. In other words, if you're tuning purely for splice size, I think it's possible to end up at a heat setting that gives you a smaller splice, but ultimately increases the odds of your PTFE swelling during the print, allowing for larger splices, and more stiction. 

    There's probably some additional information that Mosaic could add to the splice tuning guide to help, such as how to measure the splices, and what to measure for. I think this thread or elsewhere it's mentioned 1.6 - 1.9mm being the range, yet that's nowhere in the tuning guide. Some guidance around what's expected (1.8?) vs too thin/too thick would be nice.

    Are you looking for the average thickness of the splice over it's length, or just the biggest? I found a huge amount of variability here. Some tunes I'd get a 1.82 throughout the splice, where others I'd get 1.75-8 on average, but a 1.86 at one point (using the tip of the calipers, not the flat). So is a very small larger bulge better or worse than a longer smaller bulge?

    I appreciate the goal of wanting to keep things simple for customers, but there are times where simple doesn't cut it, and you need more details and experienced guidance to troubleshoot.

     

    0
    Comment actions Permalink
  • Avatar
    Jonny Yeu

    Hi M, thanks for taking the time to share your experience here. IN our splice tuning guide, we do mentioned that a diameter of 1.75 mm +/- 0.1 mm is ideal as a splice that is too thick/thin can cause either jams and/or calibration issues. We also recommend using filament that is within +/- 0.03 mm spec of 1.75 mm as we've found the best results with these types of filament. We've also found that splicing the same brand of filament helps to create stronger bonds.

    For users who have had jam issues, we've found that settings of (2, -3, 2) has helped. Have you tried this combination in your Splice Tuning testing? If possible, could you please share some pictures of recent splices that you've created and the settings that are associated with them? Once we can examine the splices, we may be able to provide some more recommendations. Also, if you do require some additional splice tubes, I can definitely send you some additional tubes since you have run into these splicing issues.

    0
    Comment actions Permalink
  • Avatar
    M Montminy

    Did you just add the tolerance? ;) I swear I don't recall seeing that. So that suggests a 1.85mm splice is the upper end of "ideal".

    I did try printing with 2,-3,2 and got jams. In fact, here's the last 2,-3,2 splice that jammed.

    That 1.86mm was a short section, the majority of the splice was <= 1.8mm, I have to hunt for the 1.86 spot.

    I then tried over 26 tunes (though I didn't print with them all, just tuned). This is what's been driving me mad... Nearly all my splice tunes come out between 1.8 and 1.85, with only a few being above, yet I would still get pretty reliable jams, after ~100 splices. If I did prints with fewer splices, I generally had no issues. These are some of the tuned splices, and what most of my splices look like even without tuning.

    I have to go to heat around 4+, or positive compression, or other unusual tunes to get ugly splices. Around 4 I start to see necking, but strong enough splices to work. Cooling seems to have little impact at all.

    That's what led me to Google, and ultimately this post. It occurred to me there was something else going on besides crappy filament and splice tuning. I mean, what are you to do when the tunes look good, and measure 1.8mm, yet they still jam. Heck, even my jams once removed aren't far off, my last one was 1.86mm, and I've had others jam at less. I previously ended up with a 4,-3,2 tune. It created butt-ugly splices with some necking, but it was reasonably reliable, or I thought it was, but I wasn't doing very long prints with lots of splices.

    At least until I see data to suggest otherwise (like more jams), I can't help but posit that I suffer with some heat related stiction in the splice tube. It's not in an unusually high ambient condition, it's not starved for air-flow, and I run with the cover off. So either I have some tubes with a lower softening point or are more prone to stiction, or I have a splice core that runs a few degrees hotter or heats faster, but I would think simply playing with heat values would address that more easily.

    I've only got 1 tube left, so I do need some spares. I was damned if I was just going to replace it in hopes the problem would go away, I tend to need to know the cause of a problem before I can write it off, and just throwing tubes at it was giving up :) 

    0
    Comment actions Permalink
  • Avatar
    Redemptioner1

    I got a couple of questions for you M to try an add to my understanding if you will indulge me, unfortunately you probably don't have the tool to measure the internal tube diameter accurately. 

    Firstly I ended up changing over the last palette 2 to a pro as they standard palette 2 just seems not to work, jamming being a massive issues as well as poor scroll wheel readings. 

    questions

    1.  Does the teflon tube feel "bumpy" inside? Use a pit of filament and cut off the end square with a knife (not scissors), then run it up and down inside the splice core, can you feel and ridges/bumps on the cut end as it goes through the tube?

    I had a few with poor moulding internally, although hard to tell if this was from the heater wire.

    2. Does the teflon tube have any "singeing" on the outside from the heater wire (don't get this on Pro's but do on standards). I think this is the main issue, the wire heats the teflon too hot and warps and shrinks it. Then the internal walls get all distorted and bumpy causing jams. 

    This issue would not happen at all if the drive motor on the outgoing side of the splice core was geared down (how you did Mosaic not learn this from the thousands of printers that fail regularly due to this cheap/poor design which causes the filament feed to skip). The tiny little nema motor with a direct drive hobb gear is not powerful enough to break the bond between teflon and filament, which is not the only issue this design flaw has causes.

     

    3. What does the heat wire wrap look like? It appears the Pro has more wire wraps for the heat wire allowing the temp to come up much quicker and be on for less time. I think it would be reasonable to assume this (not as long to heat up and not get as hot as a result) would have an effect on both the stiffness of the teflon tube (stopping the tube diameter from bulging out during compression) and the porosity (stickiness) of the internal wall surface. 

    I think this may have the greatest effect, combine with the aluminium splice core shedding heat quicker, causing the filament to jam. The tube takes too long to heat up and gets too hot allowing the tube to expand during compression and shrink too much once cooled, it also allows for pores to open in the wall surface and have filament squished into them, then it cools down and goes to move the splice and now it is stuck. The problem here is the splice diameter measures fine, only the internal diameter in the teflon tube has shrunk slightly which results in the internal teflon tube diameter being too small to allow the cooling splice to shrink away from teflon tube inner wall and become free moving. 

     

     

     

     

    0
    Comment actions Permalink
  • Avatar
    M Montminy

    @Redemptioner1 to answer your questions:

    1) It feels "rough" inside. It's pretty smooth at the entry/exit, but it's rougher where the heater is. I don't detect any edges or lips, just a bit of texture where the heat is focused.

    2) I'm not removing the tube to find out :) as I only have one spare ATM. My prior tube had no singing though.

    3) This is what my wire wrap looks like:

     

    While the aluminum core might shed heat quicker, it's the body that's aluminum, and I wouldn't expect the body to experience much heat. I would expect the ceramic to be what soaks up the heat moreso than the body, and the ceramic liner is all that come in contact with the splice tube. But, I'm not a thermal engineer.

    A co-worker has a Flir, I should borrow it and play with it while doing tuning, but I'm sure Mosaic's already done better thermal analysis.

    I'm inclined to think this might be an edge case of QC, either with the PTFE tubes, or the heat output of the cores. Maybe some units are on the edge of being too hot and make it past QC (I have to assume they test them). Maybe their suppliers are passing along edge-cases to avoid scrap (not like any mfg ever does that...) and if Mosaic doesn't re-test other than spot-checks, which is reasonable to assume, they'll miss these.

    I guess an interesting test would be to swap out the splice core with another, and see if the recommended 2,-3,2 tunes are reliable.

    If you're still following this thread Johnny, I'll be at ERRF this year and plan to swing by the booth. Let me know if you'd like me to bring the core to look at.

    0
    Comment actions Permalink
  • Avatar
    Ed Williams

    M,

    Are you able to use the FLIR with both the old and new splice settings. I would love to see the those shots and any differences.

    0
    Comment actions Permalink
  • Avatar
    Redemptioner1

    Interesting your heater wire is different colours along the wind, I will have a closer look when I get home tomorrow but I don't believe I have that on any of mine, might be just a case of some uneven heating which is one of the reasons you get the ridges inside the Teflon tube. 

     

    Hopefully Mosaic can ship you out a new splice core and some new teflon tubes to see if the issue will be resolved, certainly worth changing out the teflon tube for now as they will definitely send you some more of them for free. 

    0
    Comment actions Permalink

Please sign in to leave a comment.