S3D profile

Discussion in 'Calibration, Help, and Troubleshooting' started by Spoon Unit, Mar 9, 2016.

  1. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,095
    Likes Received:
    322
    I feel like I have my S3D profile nailed for PLA now. If you do decide to use my profile, your first layer is still always going to come down to the bed levelling process and your flow control. I personally don't seem to grasp the filament esteps calibration, so I'm calibrating flow by watching like a hawk as the first layer goes down.

    I had to unlearn quite a bit to get here. This printer is not my old printer. I tried all sort of retraction options on the way to this, but noticed the benchies people were posting on G+ were better than mine. I'd struggled with the first benchy, likely due to differences in the fan positioning, the fan effectiveness, and possibly even my heater effectiveness. I just died almost on layer 5 to 7. So I'd then turned to S3D and used knowledge learned on another printer, with lots of fairly horrible results. Eventually I decided to open up the supplied benchy gcode and see how it had been sliced (handily printed into the file at the top by S3D). I began setting up my profile like theirs, and noticed they don't ever lift on retraction. The speed of movement of the carriage, plus a little retraction of filament seemed to be enough. I also noted they had accepted that the best overall shell comes from giving in to having a seam somewhere. FYI when slicing with my profile, the seam position is set on the far right of the bed in the center, so if it's important to you, re-orient your model so the seam will be on the at the right most point.

    Alas, I tweak the profile all the time, so can't 100% guarantee it, but these two should be almost identical to the one that printed this, and I've created one for printing with each of the extruders on a dual (still yet to figure out running them both on the same print. If yours is a pro or lite, you only need to alter the machine dimensions to open up the rest of your bed.

    I hope this helps someone.

    [​IMG]

    [​IMG]
     

    Attached Files:

    Alex9779, JohnEsc, Mike Kelly and 2 others like this.
  2. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,095
    Likes Received:
    322
    Just wanted to mention after doing a little more calibration that the cooling implemented in these profiles could well be a little too careful. I implemented them based on my observation of the heat fluctuations of the extruder. So while I think these are probably actually still good settings for an untuned PID, a more aggressive increase on the cooling fan will probably be possible after PID tuning. The details of my PID tuning session are here if you need convincing that it's worthwhile going through PID tuning:

    https://forum.e3d-online.com/index.php?threads/pid-tuning.731/#post-5397

    The part cooling fan has benefits for a range of things; for sure bridging is improved by cooling as well as stringing. More experienced printers can probably explain a range of other benefits. So it seems that the perfect cooling solution would be to have 100% cooling from level 1. Reality disagrees however, and cooling on the first layer can hurt bed adhesion, leading to lifting. So, given that we want to start on zero coolling, the next idea would be to have 100% on layer 1. Reality also says don't this either, as the fans on the bigbox for part cooling are monsters. Ramping up to 100 bathes the nozzles in too much cool air and the heating element can't keep up with the drastic change in conditions. I have a feeling that with PID tuning I might well be able to return toward the settings used by e3d on the pre-supplied benchy slice, which starts the fan on layer 3 at 30% and then ramps up 10% per layer. Given the small size of each early layer this didn't give my machine time to adjust when untuned. I'll implement cooling like they did again this morning for another test and let you know what I find.
     
  3. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,095
    Likes Received:
    322
    Well. Things are looking good. Just returned to my print to notice fan speed at 80%. I increased it 10% every 2 layers from a starting speed of 30% on layer 3. So clearly it held nicely once it hit 80 with the new PID settings. Just for fun I knocked it up in two steps to 100% and it's now on full speed. I have sneaking suspicion that the speed increase isn't truly linear, just from the audible volume, so it seems that 100% is probably not a full 20% additional rotation speed. More likely it's a reference to the total voltage allowed in. In any case, the good news is that 100% fan is achievable with a tuned PID. @Greg Holloway, given the potential for this to impact first prints on the supplied GCODE (which will no doubt morph), is there any way PID tuning could be incoporated into the commissioning phase without access to a GCODE terminal (i.e. from the front screen alone)? If so, it could lead to a little more delight for those printing the benchy for the first time.
     
  4. Greg Holloway

    Greg Holloway Administrator
    Staff Member

    Joined:
    Sep 4, 2015
    Messages:
    565
    Likes Received:
    397
    @Spoon Unit Great Post!

    As far as I know there is no method to do PID autotune without a terminal connection. I seem to remember that the PID information is accessible through the LCD but there's no option to autotune. It would not be a bad idea to put in a feature request on the Marlin GitHub if that hasn't been done so already | https://github.com/MarlinFirmware/Marlin

    My best advice for the fans when printing the benchy is to lower them down. I'll look at adding in some instructions to the first print section.
     
  5. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,095
    Likes Received:
    322
    You could probably add into your Marlin fork a menu item to do the tuning, but I think the problem would probably still be actually reading the results. It wasn't clear to me whether the autotune step just reported its findings or actually also set the values. Either way, the values need to be made solid somehow, and I've chosen to do that in starting GCODE, and in order to achieve that you need to see the numbers on the screen.
     
  6. Greg Holloway

    Greg Holloway Administrator
    Staff Member

    Joined:
    Sep 4, 2015
    Messages:
    565
    Likes Received:
    397
    M134 will permanently write the PID values to Marlin.

    Check here for a list of all the available codes and commands | http://reprap.org/wiki/G-code
     
  7. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,095
    Likes Received:
    322
    I see. I've been fiddling with M500 too. Using that I can save bed levelling numbers, bed PID and nozzle PID. M501 to review saved settings.
     
    mike01hu likes this.
  8. Greg Holloway

    Greg Holloway Administrator
    Staff Member

    Joined:
    Sep 4, 2015
    Messages:
    565
    Likes Received:
    397
  9. R Design

    R Design Well-Known Member

    Joined:
    Oct 7, 2015
    Messages:
    918
    Likes Received:
    183
    On cooling and bridging: I noticed that on square parts with bridges on each face the issues were occurring on the faces parallel to the motion of the air and on the ends of the bridge that were nearest to the fan (PLA).

    So I sliced the part such that all the faces were at 45 degrees to the fan.

    It worked!
     

Share This Page