Wiper Designs

Discussion in 'Getting Started' started by Andy Cohen, Sep 26, 2019.

Tags:
  1. Andy Cohen

    Andy Cohen Well-Known Member

    Joined:
    Aug 23, 2019
    Messages:
    217
    Likes Received:
    57
    So My TC is running great. All 4 extruders print well. Toolchanging is now as close to 100% reliable as I can get. So now I am trying to print a simple multi-color PLA 3D Benchy and DOH! The hotends have so much drippage they eventually knock the Prime Pillar (S3D) right off the bed. I tried 4 times. Next I'll try with an Oooze Shield instead... But... How about wiper designs? The idea of replacing all the nozzles with steel replacements only because of the use of a wired brush can't be the only way to do this. I have a batch of the thick felt strips that Aleph Objects uses in the lulzbot mini for nozzle wipes and I will see if I can insert a piece of it into the wiper part that is on the Github. It should be fine with brass hotends. But there must be other approaches.

    While at MRRF I saw the 3 beta E3D TCs. One had a kind of bladed thing in front of each tool such that when the hotend passed over it the blade knocked off any oozing. There's nothing like that I could find on the Tverse. I think it was on Hackney's machine

    Then there's how you get the wipe to happen??? I see the "Brush.g file... Do I insert gcode in a tool change script (S3D again)???
     
  2. Paul Arden

    Paul Arden Well-Known Member

    Joined:
    Aug 28, 2019
    Messages:
    162
    Likes Received:
    56
    I feel your pain :)

    [​IMG]

    To be honest this is just not enough reading around yet and taking in all of the different ideas people have. For the out of the box wiping though...

    If you look in the tfreeX.g and tpostX.g you'll see a commented out call to the purge and prime macros. Then the actual moves are in those macros. I see some people however are also moving this logic elsewhere but that is where the default ones E3D give you live.
     
  3. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,546
    Likes Received:
    481
    Multi-color benchy is a challenging print. I have a feeling that setting starting points closest to the center of the model is probably required. For non-abrasive brushing, I'm using a silicone brush. I've created a pattern for the prime and purge steps that have two goals; firstly prime the hotend in theory to the point of readiness, and secondly to catch the primed material in the brush and knock it out. The prime steps seems to have relatively little success with knocking the primed material out, probably because it just got attached and was hot and sticky at the time. So in my purge step I revisit it with the fast moving nozzle at which point I have about a 90% hit rate for knocking it free. The thing that must be avoided is any chance of picking up the strand(s) on the way back to the print head. Even if the strands aren't knocked free first time, there's a good chance of it happening on subsequent purges.

    I love that picture. That's a good indication of the challenge we face.
     
    garethky, Paul Arden and Ntesla66 like this.
  4. Andy Cohen

    Andy Cohen Well-Known Member

    Joined:
    Aug 23, 2019
    Messages:
    217
    Likes Received:
    57
    I found where the bladed knock off design is and the designer (Micheal Hackney) will be sending the files to me later tonight.
    So far everything I've tried eventually gets knocked off the bed... Purge pillar, ooze shield... I even tried a 1 layer raft to hold the ooze shield down... but the number of crashes by nozzles with long strands hanging are too much.

    For sure the 3DBenchy is a challenge. It's LOTS of toolchanges in a small object making for lots of strands crashing. I guess I should try something with less changes in a small area in Z.
     
  5. Andy Cohen

    Andy Cohen Well-Known Member

    Joined:
    Aug 23, 2019
    Messages:
    217
    Likes Received:
    57
    I think I have part of a solution... a bungee cord strung from the screws of the furthest Titans and at a height so that when the nozzle passes it gets wiped by the bungee cord. It kinda works, easily knocking off the longer drools... but I think it will work even better if I purge right before tool PU or perhaps purge after the tool is released... the later so that to drool by the nozzle is soft while the length is hard making it easier to knock off.

    Also, if you are putting your temps at the suggested 220C you're way too hot. I'm down to 205 and could probably go lower. Less heat, less drool.
     

    Attached Files:

    #5 Andy Cohen, Sep 27, 2019
    Last edited: Sep 27, 2019
  6. Andy Cohen

    Andy Cohen Well-Known Member

    Joined:
    Aug 23, 2019
    Messages:
    217
    Likes Received:
    57
    For sure the bungee cord helped a lot. I position it so that as the nozzle passes by the bungee rolls down the slope to the opening of the hot end, it wipes the opening then rolls up the slope of the other side. The porous cloth helps the soft PLA to stick and the hanging bits make it stick even more. Eventually it builds up and falls off the bungee. Not a perfect solution but it does make a big difference. A purge before a tool PU will no doubt make it work even better.
    How's that for a simple solution? So a a "G1 E100" in each tFree file before the PU should be all I need. No need for the macros or motorized Purge/wipes assemblies! Although it may work quite a bit better if the cord is extended forward so that when the tool is waiting to heat up it passes it afterward rather than before.
     

    Attached Files:

    #6 Andy Cohen, Sep 27, 2019
    Last edited: Sep 27, 2019
  7. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,546
    Likes Received:
    481
    Looks OK. The mass stuck to the bungee infront of T2 is a bit of a concern. I suspect a simple advance here would be have shorter bungee runs in order to provide more tautness for each tool, and also to permit a different bungee height per tool.

    You've now got me thinking about pickup up a cold tool to ram/brush content off the brush/cord, which is probably just a slicer challenge.

    It's clearly a big improvement over the first result you had, but I think I can probably get a better result with the brush. I'll try a three colour one later so that we can compare results.
     
  8. Andy Cohen

    Andy Cohen Well-Known Member

    Joined:
    Aug 23, 2019
    Messages:
    217
    Likes Received:
    57
    So (when I get back to the shop) I need to figure out:
    1. How to get the extruder to full temp before it's picked up
    2. Where to do a G1 e50 before the pickup and after the heat up
    3. Where to put the G1 e50 after the tool is released
    This way I'll get 100mm of drool with the bottom hard the top 50mm still a bit soft. The weight and length won't get past the bungee.

    If I can do all of the above I have no doubt the bungee cord will work well.
    Btw... I had been assuming that the place where the bungee crosses over the Z axis rail is a temporary compromise until I made a better fit setup... however, after watching it I see that the diagonal suspension actually helps the wipe giving it a lot more area and not snapping the cord as it passes. I suggest trying it as shown first. The cord rides down and up the slope of the nozzle making for a very reliable wipe.
     
  9. Andy Cohen

    Andy Cohen Well-Known Member

    Joined:
    Aug 23, 2019
    Messages:
    217
    Likes Received:
    57
    Tried in each of the config files... tFree, tPost and tPre
    None let me do what I wanted to do...
    That is... Purge after the nozzle is hot enough and BEFORE the tool is removed from the parking place.
    Any ideas?
     
  10. Andy Cohen

    Andy Cohen Well-Known Member

    Joined:
    Aug 23, 2019
    Messages:
    217
    Likes Received:
    57
    The next thing I will try is to preheat all 4 extruders in my start gcode using the S3D variable names and the S3D temp settings... For example...
    "
    G10 P0 S[extruder0_temperature] R[extruder0_temperature] ; Set tool 0 operating and standby temperatures
    M109 S[extruder0_temperature] ; set extruder temp and wait
    G10 P0 S[extruder1_temperature] R[extruder1_temperature] ; Set tool 1 operating and standby temperatures
    M109 S[extruder1_temperature] ; set extruder temp and wait
    G10 P0 S[extruder2_temperature] R[extruder2_temperature] ; Set tool 2 operating and standby temperatures
    M109 S[extruder2_temperature] ; set extruder temp and wait
    G10 P0 S[extruder3_temperature] R[extruder3_temperature] ; Set tool 3 operating and standby temperatures
    M109 S[extruder3_temperature] ; set extruder temp and wait
    "

    Then put a G1 E25 in the tPre
     
  11. Andy Cohen

    Andy Cohen Well-Known Member

    Joined:
    Aug 23, 2019
    Messages:
    217
    Likes Received:
    57
    Well that did not work. LOL!
     
  12. Andy Cohen

    Andy Cohen Well-Known Member

    Joined:
    Aug 23, 2019
    Messages:
    217
    Likes Received:
    57
    Next thing I tried was to add another cord that goes across where the bed starts. I changed the gcode in the appropriate file so that the nozzle stops while warming up right in front of the cord... but CRAP none of the wait for temp codes seem to work!!! So I can't get it to purge when I want it to.
    This is staring to be a PITA!
     
  13. Paul Arden

    Paul Arden Well-Known Member

    Joined:
    Aug 28, 2019
    Messages:
    162
    Likes Received:
    56
    Since you're using G10 shouldn't you rather use M116 than M109?
     
  14. Andy Cohen

    Andy Cohen Well-Known Member

    Joined:
    Aug 23, 2019
    Messages:
    217
    Likes Received:
    57
    Tried M116. It did not wait. The start gcode with the s3d variables did not work at all.
     
  15. Andy Cohen

    Andy Cohen Well-Known Member

    Joined:
    Aug 23, 2019
    Messages:
    217
    Likes Received:
    57
    Totally nailed it.
    So the way it works is there are 2 bungee cords. The 1st which is strung from the farthest left upper Titan bolt to the farthest right, knocks off drools which were the result of sitting in the parking place. I strung a second bungee cord from braces I designed and printed which sits right above the bed when the tool is just picked up and I moved the tool waiting spot 20 mm further back in Y so that the nozzle sits right behind the bungee. I rem'd out all the movement lines of the Prime/Purge files and removed the ";" from the M98 call in the tPost files. So that right before the tool moves in to print it primes then moves past the bungee which wipes it. If anyone here wants more details I'll post in another thread.
    Time to start printing!
     
  16. Paul Arden

    Paul Arden Well-Known Member

    Joined:
    Aug 28, 2019
    Messages:
    162
    Likes Received:
    56
    Strange. How are you calling it? This is what my start G-code looks like (for PrusaSlicer not S3D but you get the idea):

    Code:
    M104 ; included to suppress PruceSlicer automated Gcode
    G29 S1 ; use mesh bed compensation
    T-1 ; drop active tool if any
    ; set tool active and stand-by temperatures
    G10 P0 R{first_layer_temperature[0] + standby_temperature_delta} S{first_layer_temperature[0]}
    G10 P1 R{first_layer_temperature[1] + standby_temperature_delta} S{first_layer_temperature[1]}
    G10 P2 R{first_layer_temperature[2] + standby_temperature_delta} S{first_layer_temperature[2]}
    G10 P3 R{first_layer_temperature[3] + standby_temperature_delta} S{first_layer_temperature[3]}
    ; turn on the tools but do not execute tool change macros
    T0 P0
    T1 P0
    T2 P0
    T3 P0
    M116 ; wait for tools to reach temperature
    
    In my case it is definitely waiting for the temperatures to come up after calling M116. I did have to add the lines to actually turn the tools on though since otherwise they would be in whatever state they happened to be in when the print started (so if in an off state they wouldn't warm up).
     
  17. Andy Cohen

    Andy Cohen Well-Known Member

    Joined:
    Aug 23, 2019
    Messages:
    217
    Likes Received:
    57
    Oooo... Never tried the "P" parameter with the "T" command! Thanks.
     
  18. Greg Holloway

    Greg Holloway Administrator
    Staff Member

    Joined:
    Sep 4, 2015
    Messages:
    983
    Likes Received:
    622
    make sure you add a T-1 P0 otherwise it will think it has a tool picked up.
     
    Killercds and Paul Arden like this.
  19. greenlee

    greenlee Member

    Joined:
    Nov 9, 2019
    Messages:
    5
    Likes Received:
    5
    here is the wiper I'm using.
    Grab a silicone sheet from mcmaster https://www.mcmaster.com/5787t33
    you'll also need a 3x~8mm bolt to clamp the sheet into the gap
     

    Attached Files:

  20. Andy Cohen

    Andy Cohen Well-Known Member

    Joined:
    Aug 23, 2019
    Messages:
    217
    Likes Received:
    57
    So after thinking I had it this nailed by using the dual bungee cords I switched to PETG. Not.
    PETG is just too sticky and getting a straight down line of extrusion is simply unreliable it stuck when the extruder flew past the cord.
    So I combined my arm design which holds the bungee with the E3d brush holder. I chopped a silicon basting brush for the brush. The brush is not meant as a wipe for the nozzle. Rather it is meant to catch strands which did not release after I ran my Prime script. The tool gets picked up and moves past the 1st Bungee cord which knocks off the drool the came out while parked. It moves out to right before the second cord (held by the arms on the sides) and puts out a long strand at a higher feedrate, then a short strand with a lower feedrate. It then pauses for 2 secs allowing the strand to stretch. Then the tool moves to where the Bruch holder is. In the process it passes by the second bungee cord diagonally and in most cases knocks off the long prime strand. Next the tool goes through the brush using E3d's motion of back and forth in X. I do it 3 times with the idea that any strand that did not get knocked off will get tangled in the brush. Works great! Not one miss in a 7 hour print. The only down side is like with most designs the strands pile up. So... I mirrored the Arm/Bracket design so that I could install it on the right side of the case and wipe tools 2 and 3 over there which would split the places where the strands pile up giving me a lot more excess capacity.
    BUT.
    OUCH!
    After installing the mirrored print I find out that the Toolchanger's limit in the X axis will not allow one to do this! IMG_6210.jpg
     

Share This Page