Guinea Pigs Required!

Discussion in 'Calibration, Help, and Troubleshooting' started by Greg Holloway, Mar 10, 2016.

  1. PsyVision

    PsyVision Moderator
    Staff Member

    Joined:
    Sep 4, 2015
    Messages:
    798
    Likes Received:
    248
    Looking sharp rob!
     
    Rob Heinzonly likes this.
  2. Chase.Wichert

    Chase.Wichert Well-Known Member

    Joined:
    Mar 1, 2016
    Messages:
    775
    Likes Received:
    208
    Hmm, when I first ran the profile it tried to use the right one... Now I'm not so sure what's going on...
     
  3. Alex9779

    Alex9779 Moderator
    Staff Member

    Joined:
    Sep 4, 2015
    Messages:
    2,411
    Likes Received:
    735
    Ok now my box is almost running, except the heated bed, I had a look at the profile(s) and have a few comments.

    1. I try to set up my profiles so that they are a bit failsafe. I don't like to have to do something in advance on the machine. Thus I like to add the auto homing G28 to the start script.

    2. I have some questions/problems with the tool change script. Apparently you based the profile on some generic dual profile, that's why you have the drop down boxes. If you make the error like someone before and me too, to switch to "Extruder 1 only" and not to carefully check the scripts you run into a crash because the start script is cleared out but the tool change script remains. So no offset is set for Extruder 1 though it will be used once in the tool change script in the first docking sequence.
    To that first docking sequence is my question: why do you dock tool 0 if new tool is 1 and vice versa?

    3. The steps per mm settings in the tool change script... Wouldn't it be better to have that in the start script?

    With 2. and 3. you would have a generic tool change script working in almost all circumstances. What you have to change when you have to change something would all be in the start script. Still thinking about how user friendly this will be... I am more fan of some generic profiles just selecting options in the GUI rather than having to modify a script each time I slice.
    IMHO I would do all material settings on the machine like steps per mm...

    4. Did you actually dig into the FFF profile itself? I did that and it is quite annoying but you get wonderfully generic profiles using the drop downs. Like when switching to Extruder 1 only to remove the heating for extruder 2 and so on...
     
  4. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,546
    Likes Received:
    481
    The adventure begins

    [​IMG]
     
  5. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,546
    Likes Received:
    481
    OK. Print one was hilariously bad:

    [​IMG]

    This happened because I didn't use the Dual Extrusion Wizard Properly; I failed to select the profile to use during the wizard set up. Quickly killed that print, reset up and:

    [​IMG]

    The scarring to the right of the base along with a slightly fat first red layer are the result of the right nozzle being a fraction too low. Clearly I also need to adjust X difference for the nozzle too. I guess as it hasn't moved far enough left, I've failed to get the number high enough? Another print will clarify. It's only a fraction away, so I'm thinking it's maybe only 0.5mm. I'll give it a full mm in order to try to get it right in as few moves as possible. It's possible I've got Y wrong too. I probably should fix the height first though as undo and retightening the right nozzle could throw out both X and Y.

    It would also be interesting to know how high people have their wiper. One thing I noticed during the tool swap was that priming doesn't always lead to a nice falling line of material. Sometimes it seems to build into a ball around the nozzle, and then you're really just hoping that the wiper will get it. I think my wiper is a little too low and it's not always wiping the tip as clean as it could.
     
  6. Greg Holloway

    Greg Holloway Administrator
    Staff Member

    Joined:
    Sep 4, 2015
    Messages:
    967
    Likes Received:
    619
    1: Fair enough, a personal choice.

    2: You're absolutely right it is based off a generic profile. We are still refining things. This is experimental so attention to detail is important. The extruder offset is calculated from extruder 1, so if things are not setup correctly it will crash. The left extruder is extruder 1 (T0), and the right extruder 2 (T1).

    3: Marlin only accepts one value for steps-per-mm on the extruder, if you have two extruders it is highly unlikely they will be the same and so to increase the quality I added the steps-per-mm change. This script sets the correct value for the extruder as it changes tool.

    4: Nope. It's modified from a generic one sent by S3D for us to develop on.
     
    PsyVision likes this.
  7. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,546
    Likes Received:
    481
    In this profile, for the dual, the bed size is too large and could permit you to ram the head into the right I guess.
     
  8. Alex9779

    Alex9779 Moderator
    Staff Member

    Joined:
    Sep 4, 2015
    Messages:
    2,411
    Likes Received:
    735
    3. Ah ok didn't realize that. I am a bit influenced by my UM2 which uses a Marlin clone but I use the TinkerGnome branch which has a lot of tweaks in it. Especially lots of settings for dual what I do not have on my UM2 are shown on the menus and split for both extenders. I thought this was a problem of the UM2 firmware not of Marlin itself. Well sad sad..

    4. Just open it in a text editor it is a simple XML. Once you get the way it is built it is very easy. But fair enough this may be interesting in the future when the base settings have been sorted out...
     
  9. Alex9779

    Alex9779 Moderator
    Staff Member

    Joined:
    Sep 4, 2015
    Messages:
    2,411
    Likes Received:
    735
    Just a simple suggestion: why not just throw the profile(s) in a Github repo? I manage my profiles with Git. Because the FFF profiles are only XML it is very nice to see differences and do merges and stuff...
     
    amdbuilder likes this.
  10. PsyVision

    PsyVision Moderator
    Staff Member

    Joined:
    Sep 4, 2015
    Messages:
    798
    Likes Received:
    248
    @Spoon Unit like the cone! My kind of colours too!
     
  11. Paul Seccombe

    Paul Seccombe Well-Known Member

    Joined:
    Oct 22, 2015
    Messages:
    90
    Likes Received:
    30
    Yet another cone successfully printed with the experimental Simply3d profile.

    [​IMG]

    I too had the problem with the right extruder going in reverse when I tried to run the extruder calibration from the sd card. I reversed it in Marlin and ran the calibration. When I came to print though it went in reverse so I put it back and the print worked. I will have to look at the calibration g code to see what is going on. I also noticed that manual control of the two extruders doesn't work properly. It is obviously in absolute mode but extruder 1 and 2 are using the same parameter so if you move 1 when you try to move 2 it displays extruder 1 position and always goes in reverse.
     
    #31 Paul Seccombe, Mar 12, 2016
    Last edited: Mar 12, 2016
  12. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,546
    Likes Received:
    481
    Really close now. Finally have the nozzle heights aligned. Just some very fine adjustment to make on X and Y.

    [​IMG]
     
    Pierce, Greg Holloway and PsyVision like this.
  13. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,546
    Likes Received:
    481
    For those who have Octoprint, and thus access to a console, you can pimp your tuning process further by submitting your suspected offsets as movements. I haven't take pictures of this, so will try to describe.

    1 Move the left nozzle over the head of the pin using the LCD and the dial. Note down the X and Y
    2 Move the right nozzle over the head using the LCD and the dial. Note down the X and Y
    3 Calculate the differences
    4 Go to the console in Octoprint and execute G91, which sets upcoming commands to relative motion instead of absolute.
    5 Execute a G0 command like this adjusting the numbers for your situation. If your Y number is negative, don't worry about it, just pop that in.

    G0 X33.5 Y0.1​

    6 In a perfect world you just moved the first nozzle directly back over the pin head and you can go and get the tea on
    7 If you want to nail this to the micron, continue to step 8, else take your numbers and get them into your GCODE start script
    8 First move back to where you were by reversing the numbers from your last move

    G0 X-33.5 Y-0.1​

    9 Now based on what you notice, i.e. you overshot or undershot slightly in one direction, adjust your number a fraction and resubmit

    G0 X33.4 Y0.2​

    10 Return to step 7

    This should help you get a really good first approximation.
     
  14. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,546
    Likes Received:
    481
    I think I'm done with tuning this. I kinda wish I'd started out with the model chosen by @Rob Heinzonly , as the traffic cone needs a number of layers before you get the first tool change that gives you feedback. I don't have a totally perfect cone, but the last one above was pretty close and, if you don't have it too close, it looks great. The camera always shows fine differences if you want to be picky. So, having changed calibration prints, I think this is about as close as I'm going to get. Aside from a little extrusion, this is probably going to be fine for my needs.

    [​IMG]
     
    PsyVision, Pierce and Greg Holloway like this.
  15. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,546
    Likes Received:
    481
    @Greg Holloway , is there a reason for the different extrusion amounts per extruder in the purge script?

    ;purge
    {IF OLDTOOL=1}T1
    {IF OLDTOOL=1}G1 E25 F400
    {IF OLDTOOL=1}G1 E-14 F4000
    {IF OLDTOOL=0}T0
    {IF OLDTOOL=0}G1 E27 F400
    {IF OLDTOOL=0}G1 E-14 F4000

    It would probably also be really instructive to hear about the thought process behind this script as it clearly comes from the result of some thinking, and also some knowledge of the e3d v6 hotend. The hotend instructions state that the v6 works best with between 1.5mm and 2mm retraction, but here we see a restration of 14. So I guess this isn't really retraction but ejection. The other interesting thing about the purge is that you notice the extruded filament double, maybe even triple in thickness at the end of the push, and infact on my machine, you hear the extruder motor click back as the system fails to click back. I almost imagine you sitting there tweaking that amount until you heard that single click, but clearly this could be different per material. The added weight of the thick extrusion seems to help the filament hang vertically and I guess it may also give a greater surface area to increase cooling and thus lead to the easy removal as it passes the wiper.

    So far this all me just trying to understand why you've done what you've done and guessing, so it would be great to hear if i'm on the right track or not, and how you thought through this.
     
  16. Greg Holloway

    Greg Holloway Administrator
    Staff Member

    Joined:
    Sep 4, 2015
    Messages:
    967
    Likes Received:
    619
    @Spoon Unit They should be the same, that's my bad. Set both to 25.

    There has been a lot of thinking behind it and the initial concept came from Sanjay. I don't want to say anything more at this time, but what I will say is wait until the Midwest RepRap Festival (http://midwestreprapfest.org/) next weekend. We'll have some information to spread then.
     
    Paul Seccombe and Spoon Unit like this.
  17. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,546
    Likes Received:
    481
    Sure. I appreciate the response. Long term, in order to learn how to think about how to use the Bigbox, it would be great to understand the thinking behind these mechanics; in many ways, we can learn from the results of the thinking as laid down in the GCODE, but it would also be great to gain an understanding of the thinking itself. I await the 'information to spread' with interest !
     
  18. Rob Heinzonly

    Rob Heinzonly Well-Known Member

    Joined:
    Sep 20, 2015
    Messages:
    448
    Likes Received:
    174
    Hello. A weekend full of dual print testing is over, and not all guinea pigs have survived. First the good news. here's my traffic cone:

    [​IMG]

    Extrusion was perfect, alignment was (almost) perfect. And now for the benchy, the dual benchy. While the purge script worked perfectly with the traffic cone, I had a couple of crashes with the benchy just like Alex9779 described. After I figured out what the problem was, the first 100-200 layers extruded perfectly. But when printing the doorpost of the little vessel, way too much filament was extruded (see the orange blobs). While observing the process, I noticed this happened only directly after the purge in the ooze bucket. @Greg Holloway : can there still be some pressure left in the extruder ? Or did the molten filament work it's way up along the shaft of the filament during the purge process ?

    [​IMG]

    Finally the orange extrusion came to an end after the HobGoblin dug its way into the filament during one of the tool changes, Guinea pig three died in the process :-(
     
  19. Greg Holloway

    Greg Holloway Administrator
    Staff Member

    Joined:
    Sep 4, 2015
    Messages:
    967
    Likes Received:
    619
    @Rob Heinzonly Check the extra length on restart, it might be that.
     
  20. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,546
    Likes Received:
    481
    @Greg Holloway , I think one of the issues showing up on prints is layer to layer inconsistency and I'm sure this is at least partly to do with retraction + lift. I struggled with this at first on the BB because I absolutely needed both retraction + lift on the M3D, but this resulted in zits and inconsitency on the BB. After looking through the settings on your pre-delivered Benchy I realised you didn't use lift at all. I was really amazed that the benchy printed so well without left, though I think perhaps a slightly colder hot end would help further to reduce some of the stringiness I saw from a lack of lift. Then, this dual profile you've delivered has lift. It seemed to me the only real reason for the lift was to try to prevent the ooze that might be there on the freshly primed hot end. Did you experiment with a Z lift only on the tool change script I wonder?
     

Share This Page