IN DEVELOPMENT Dual Volcanoes

Discussion in 'Guides, Mods, and Upgrades' started by Spoon Unit, Mar 25, 2016.

  1. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,546
    Likes Received:
    481
    Dead in the water on the traffic cone volcano dual test. The filament won't even feed into the 0.6mm head. The filament just keeps still, despite the attempts of the hobbled gear. It's as if the heatbreak has filament in it too high up, possibly the result of earlier dual printing and then re-assembly.

    Worse that that, in attempts to fix it, a piece of filament slammed into one of the heat sink fans and tore off a fan blade. Now it rattles and is the noisiest item on the machine. For anyone else who's removed the metal extruder parts a few times, and as a result, have taken the fan off the heatsinks more than once, have you also found that the blue snap-on plastic has become incredibly loose?
     
  2. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,546
    Likes Received:
    481
    Pulled apart the whole right extruder assembly again to find something very odd. The PTFE tube at the top seemed to have been pushed very slightly too far into the heatsink. One result from this seems to have been that the hole at the bottom of the PTFE tube was too small, hence the inability of the filament to get pushed through. Re-assembled, heated, pushed, and that flow issue is now sorted. So, now to try the dual volcano traffic cone.
     
  3. JvdP

    JvdP Well-Known Member

    Joined:
    Aug 3, 2014
    Messages:
    60
    Likes Received:
    24
    I wonder whether Simplify3D offsets for different nozzle diameters. I suppose it doesn't and it just puts the 0,4mm wide stripe on top of the 1,2mm wide one. This means that you can't make seamless walls. however, you can of course use the 1,2mm nozzle for infill and the 0,4mm for outlines. What was your idea?

    Edit to answer some of your questions.

    As I mentioned earlier I've experienced jams with small nozzle diameters on volcano. However I don't have hundreds of hours experience on the volcano so it might as well have been due to my incompetence.

    Regarding the fan. On my other printer I've struck the little fan many times with various objects, amongst which the metal screw driver might have been the worst. No problem though, never had a blade break off.

    I have however noticed that over time (months/year) the blue bracket becomes more loose. Not a big deal though is it?

    Your PTFE sounds like it was pulled into the hotend for some reason? Filament too thick?
     
  4. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,546
    Likes Received:
    481
    I think Simplify will handle different nozzle sizes. It doesn't really care about the nozzle size and just creates a path for the current slice based on the nozzle. I think that does mean that the edges will be fine.

    My idea with the different nozzle sizes was indeed build volume and detail for the two nozzles respectively. The first proper test I'll make once I've got the offsets sorted out again will be a plant sign, where the majority of the plant sign is built fast, but the lettering uses the smaller diameter. In fact, from S3D, it looks like a 0.6mm nozzle will give a good result.

    Could well be a filament blob that dragged in the PTFE. Hadn't considered that. It does seem to me that the PTFE is a little bit tighter than it needs to be against the plastic.

    The only problem with the loose blue plastic is that the no de-bladed fan is vibrating a lot and due to the loose plastic, thats turning into a very unpleasant noise.
     
  5. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,546
    Likes Received:
    481
    It looks like simply doubling the retraction is not, alone sufficient. I may need a longer push too. I just noticed the right hand, 0.6mm nozzle with a massive trail of ooze. Fortunately it doesn't really matter much for the traffic cone. I think that same is true to a lesser extent with the 1.2mm nozzle. While it's idling I do see a small blob of plastic hanging on the end.
     
  6. Kick2box

    Kick2box Well-Known Member

    Joined:
    Mar 15, 2016
    Messages:
    198
    Likes Received:
    27
    For retraction I would try just to add the difference in length of the nozzles which is 9mm, as the heatbreak is the same.
     
  7. R Design

    R Design Well-Known Member

    Joined:
    Oct 7, 2015
    Messages:
    944
    Likes Received:
    190
    What layer height(s) are you experimenting at?

    I found that at 0.25mm (as Richrap uses) my 0.6mm Volcano was pretty sweet. And at higher layer heights is good too, but then I get artefacts on rightangled corners (the test cube).

    Experimented a bit with coasting (S3d) but not convinced.

    How far between the Volcano block and the fan duct? Material X is not bothered?

    What was your reasoning in choosing to put the PT100 towards the fan duct (and not the heater cartridge)?
     
  8. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,546
    Likes Received:
    481
    @Kick2box I already added 10mm, as I thought that was the difference. The problem is, this doesn't work as well as with the standard nozzles. When we tried guessing previously at how purge worked, some suggested that the nozzle-ramming-speed-extrusion was designed to get a lot of mostly un-fully-melted filament into the nozzle, before whipping it back out. So, not only do we need to increase retraction, we also need to guess (unless we get help) what sort of numbers might make sense for the purging of the nozzle itself.

    @R Design good question. I did a little research on the net and came up with some experience from others that layer height should be <= 80% of nozzle width. I ran my dual volcano traffic cone with 0.5mm layer for both the 1.2 nozzle (i.e. well within 80%) and also with the the 0.6 nozzle (slightly over the 80%). The traffic cone actually printed pretty well. I also printer a small vase with 0.6mm layer height, which was nice, and a larger vase with 0.4mm layer height, which also looked great. With the 0.6mm layer height, every thing looked great, but the 0.4mm had one particular issue on the print where some part of one layer had pulled away while being laid.

    I don't think coasting is going to be necessary. Seems like the e3d equipment does retraction really well and we're using a really high number for that so I think that we have extrusion motors to cleanly cut the cord, thus making coast slightly unnecessary.

    The volcano block and the fan duct are about 3mm apart. I'm actually amazed that there's no impact on the fanshroud at all. I've had the heat block up to 240C so, yeh, MaterialX looks pretty amazing.

    I put the PT100 closed to the exit in order to minimise bending of the pt100 wire. I think looking back on it, there was actually no other way, because two volcano blocks the other way round would overlap.

    @Chase.Wichert I've been thinking about the seam issue. It seems like this will be a natural issue as the nozzle size and layer height increase. Looking inside my traffic cone, I can see that scar is closer on the 0.6mm nozzle than the 1.2mm nozzle. Just thinking about the physics of this I thought about a stick of liquorice (sweet tooth) hitting the floor and then being laid over and pulled a bit (more likely with toothpaste than liquorice as I probably ate the liquoice before it hit the floor). So you've got this cylinder of toothpaste, hits the floor sticks, but then gets pulled over and dragged a little, with the drag affecting the top more than the bottom (because the bottom is more stuck). I think that's what's resulting in the scar and so it seems like a solution will be to have some extra restart distance which is either the layer height, the nozzle width, or some combination of the two. In the first instance, I'm going to make my next test with 50% of layer height. Hopefully that will provide enough extra material to minimise the effect of drag as there should be a touch of pressure in the end of the nozzle immediately pushing to keep that first bit in place.
     
  9. Kick2box

    Kick2box Well-Known Member

    Joined:
    Mar 15, 2016
    Messages:
    198
    Likes Received:
    27
    with Slic3r I purge like that
    G1 E10 F400
    G1 E-5 F4000
    and it's working perfectly, there is NOTHING coming out after that.
    I reduced in order to spent less filament during toolchange and because I had jammed hotends with higher values.
     
  10. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,546
    Likes Received:
    481
    You're doing that with dual volcanoes @Kick2box ?
     
  11. Kick2box

    Kick2box Well-Known Member

    Joined:
    Mar 15, 2016
    Messages:
    198
    Likes Received:
    27
    No, these are my values for normal dual. Just wanted to state that it's working with much less retract.
     
  12. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,546
    Likes Received:
    481
    @Kick2box It looked to me like the purge process had some thinking behind it from @Sanjay. I didn't want to tinker with the numbers on that until they explained the thinking behind it, which they seem to be holding close to their chests. Unfortunately this means we can't fully understand what we've bought from them and I really don't want to just fiddle with the numbers on the dual myself without the understanding. The chance of getting the much larger melted zone up into the heat break and jamming just feels a bit too high to risk, which is then a lengthy fix operation, with the attendant re-work to level the nozzles.

    Still, if it's working for you with less retract, then great. They may have overdone the numbers to be sure; hard to tell.
     
  13. Sanjay

    Sanjay Administrator
    Staff Member

    Joined:
    Feb 27, 2014
    Messages:
    126
    Likes Received:
    119
    The ram-purging technique isn't top secret! It's just a bit tricky to explain and I need to do a blog post with diagrams etc to really explain it properly.

    What's happening is that you ram filament into the hotend just slightly faster than the hotend can melt it. This causes all the melt to be purged out, and inside the nozzle there is no liquid filament, just soft stuff that you can completely pull out cleanly.

    The process is essentially:
    1. Finish printing, move to dump zone.
    2. Extrude as quickly as you can at a speed *just* below the speed at which extruder skipping or filament grinding occurs.
    3. Do this for as long as it takes to purge all melt.
    4. Retract immediately and rapidly as possible to a distance such that all the retracted material is out of the melt zone.
    5. Tah-dah, no ooze.
    The variables here are the ram-purge speed, the ram-purge length, the retraction-speed (as fast as your extruder can cope with), and the retraction length.

    These variables require much tweaking, trial and error etc to get just right. To achieve this with volcano you will need to be able to ram-purge and retract extremely quickly, as you are effectively racing your extruder against your hotend, and trying to beat volcano is a hard task. You will also have to retract out of that melt zone very quickly and to a distance that is totally outside of the long volcano melt zone. If your extruder is man enough, I think it should be possible, but will require much tweaking and high speeds.
     
    JvdP and Spoon Unit like this.
  14. Henry feldman

    Henry feldman Well-Known Member

    Joined:
    Sep 5, 2015
    Messages:
    969
    Likes Received:
    391
    @Sanjay is this a feature that is easily accessed on the BB? (e.g. can we add a g-code script into Octoprint to perform this action [or add to the post-print script])?
     
  15. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,546
    Likes Received:
    481
    No diagrams necessary here. That completely clears it up. It also explains why the extruder gear skipped at the end of the purge; this would be the point where unmelted filament was at the bottom of the nozzle and couldn't be pushed further as it wasn't melted yet. I'll give this a bash on my next run:

    ;purge
    {IF OLDTOOL=1}T1
    {IF OLDTOOL=0}T0
    G1 E35 F500
    G1 E-23 F4000

    This increases the push a little and the push speed, and increases the total retract by 9mm to account for the longer nozzle. I did already notice that the heavy weight of filament from the purge sometimes cause it to fall clean away, and on the volcano this can probably be guaranteed at a certain nozzle size and extrusion volume.

    Thanks for taking the time to explain it clearly!
     
  16. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,546
    Likes Received:
    481
    @Henry feldman the technique is already encoded into the script Greg gave out in the post entitled "Guinea Pigs Required".
     
  17. JvdP

    JvdP Well-Known Member

    Joined:
    Aug 3, 2014
    Messages:
    60
    Likes Received:
    24
    I think you're really going to struggle with the Volcano and this technique due to the enormous melting zone.

    However your large nozzle might actually help you here as it causes less back pressure. This means you will need less torque on your stepper to move the filament down during the extrude push.

    If you still find the extruder to be skipping steps then remember you can always crank up the power a little by increasing the voltage on your drivers to 0.7v or higher.

    This will be really interesting to follow along with, so I'm looking forward to your findings!
     
  18. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,546
    Likes Received:
    481
    Giving up actually on the purge on the volcano. I was close, but kind tired at the same time after a long day. Perhaps someone else could pick up the baton. This was where I got to:

    ;purge
    {IF OLDTOOL=1}T1
    {IF OLDTOOL=0}T0
    G1 E36 F1200
    G1 E-23 F4000

    Firstly, I'm currently running two 0.6 volcanoes.
    The large melt zone on the volcano allows the fast and long push of E36 F1200 to just reach the slipping point on the extruder and then the rapid retraction takes place. F4000 is just hopeful however as after about F1800 it really doesn't get any faster. It could well be that more speed is needed to get the filament out before melting occurs.
    At the point I gave up, I had actually run one successful print, but the next failed. In this case, failure consists of the extruder being unable to be re-fed because the filament has become too wide. It's simple to fix the issue - you just pull the filament out and cut the blob off then refeed - but you've still just spent all your time attempting a print.

    To get this close to correct in the first place, I wrote a simple purgetest.gcode containing only the two commands:

    G1 E36 F1200
    G1 E-23 F4000

    I played with a number of ideas. Once that had run, I waited a moment to give the filament now pulled into the heatbreak a chance to cool and harden, which will happen on most dual prints, then attempted to run

    G1 E23 F80

    I chose this slower speed with the thought that approaching reasonably slowly might give the filament time to slowly re-melt, and I altered the prime script to also use this slowed speed, but it still seems a little hit and miss. I did have success for one full print though, so maybe the numbers are close but not completely solid.

    In fact, before running the second re-feed, you can pull the filament out and see what sort of blob you have. If it gives a lot of resistance to pulling, the blob could be too wide to fit through the PTFE at the top of the extruder assembly. In quite a few cases, this is what I found after pulling it out:

    2016-03-31 20.04.45.jpg 2016-03-31 19.59.43.jpg

    That slight thickness on the end is enough to cause the extruder to be unable to repush the filament into the hot end.

    It seems entirely possible that the volcano is simply not able to work with dual printing. It also seems possible the numbers just need a little more tweaking. Perhaps ramming the filament in even faster than F1200 is the key. That should at least be tested at the max speed.

    I think any attempt at calibrating for ram purging will actually require slightly different numbers depending on the nozzle width.
     
  19. Kick2box

    Kick2box Well-Known Member

    Joined:
    Mar 15, 2016
    Messages:
    198
    Likes Received:
    27
    I suggest you first try to reduce the amount of retraction, try with -15 or -16. I had these problems with the normal E3D, as already mentioned.
     
  20. JvdP

    JvdP Well-Known Member

    Joined:
    Aug 3, 2014
    Messages:
    60
    Likes Received:
    24
    I was going to suggest what @Kick2box said. Those wide bits after a cold pull (or sometimes even a hard pull) are quite common even on the v6. I've seen it many times on my hotends. Perhaps you could strike a good balance by pulling back less.
     
    #40 JvdP, Apr 1, 2016
    Last edited: Apr 1, 2016

Share This Page