Layer Shifts / Multi Material 50% Under Extrusion

Forums The Beast V1, V2 and Elevator 3D Printer Support Layer Shifts / Multi Material 50% Under Extrusion

  • This topic is empty.
Viewing 14 posts - 61 through 74 (of 74 total)
  • Author
  • #1394

    Generally copy the code to a new txt file and save it with the extension .gcode then simply load it onto an SD card or into repetier host and hit start print button.


    Is it better to print off of the SD card or over USB, and does it make any difference, especially as far as layer shifts?


    Absolutely none. USB vs SD card has nothing to do with layer shifts. As mentioned earlier layer shifts are caused by the following :

    A)Over extrusion, nozzle catching on extra material in the Z space

    B)Stepper drivers skipping steps and temporarily shutting down – running too hot, insufficient cooling

    C)Slop in gantry assembly

    D)Z Axis moving less than required. – IE steps per mm not set correctly or microstepping jumpers not set correctly – printing thinks its moving say 1mm down but actual moves half or double that – causing over extrusion see above

    E)X or Y stepper motors “skipping steps” (usually because of A or B). This should be evident in the layer shift – has it shifted in X or Y?

    The way to diagnose – show me an entire print which has skipped steps from side on noting which is X+ and Y+ so I can make an accurate diagnosis – or just look yourself. For each individual shift – does the shift occur in one direction? Which direction. You’re unlikely to get something which appears in both at the same time. Look closely at your print.

    I didn’t see any pictures posted here of the most recent tests i requested a few messages ago. You were going to print a test cube without heated bed from memory.

    Especially for diagnosing shifts, some clear side and front photos of the entire object from side view are critical.


    The big print that failed and caused me to write this thread had a final shift in the Y axis of a couple mm, but before that there were micro-shifts, that I only caught at a later inspection, that seemed to be at quite random directions.

    I am not going to print without heated bed since I use nano polymer adhesive, and it’s designed to stick at high temperature and detach at low temp, so the first layer would never stick without a heated bed.

    I sent you pics of the first test cube on your email and asked for some advise on how to fix a few of the issues I see. Maybe my last email ended up in your spam folder? I didn’t print anything else after that, so far.

    Anyway, I am now good on the microsteps for the extruder too, now. While waiting for the revised 2 in 1 firmware, I can use the single hot end one.

    I have calibrated the stepper drivers Vref as well, so that I’m sure overheating is not a problem. I’ll print a few more test cubes trying to fix those issues I told you about on the email, and them try to reprint the big object.

    Since it makes no difference, I’ll do that through SD card, so I can use my laptop in the meantime.


    I responded to your email and mentioned you had some z banding which looked like they were caused by your bed thermostat. I asked you to turn the bed off and reprint and send pics.

    I also said that if you must, leave the bed on for the first few layers, then turn it off and let the print finish.

    There is some very sound logic in this process. I’m not just asking random questions. Each time I get an answer, it helps me diagnose your issues. FYI your test cubes showed no layer shift at all. Just z banding. Once we’ve sorted the z banding, we’ll move onto bigger prints which will likely shift. Again I’ll need photos of those too.

    There’s no point trying to diagnose multiple issues at once. We start with the primary issues and move through the rest one by one until you’re getting perfect prints.

    Could we keep all correspondence regarding this in one place (here)? You can attach images to your messages by clicking the attach button below the text entry field.


    By the way. What material are you printing with? I’m assuming you’re starting with PLA? If so. I’d suggest hair spray, Kapton tape or blue painters tape on a cold bed until we’ve figured out this layer shift issue.




    So PETG? Great filament. But a little warpy and definitely likes a heated bed. You don’t have any PLA to test? It’s very important we get answers to these questions. If you can’t turn off your bed it’s going to make diagnosing your banding nigh on impossible. One print is all I need without the bed on.



    By the way. Have a look at this.


    Recognise this issue? Can you now see why I want you to test without the bed on? I need to cancel the most likely cause (being bed thermostat).

    I’d be very interested to see the photo of your cube without the bed on. And trust me. With this printer, that’s based on a shit load of experience.

    Other potential cause is warps in Lead screws, although 2 features you may have noticed; the larger than necessary hole in the top of the 12mm rod holder for the lead screw as well as the leads being tied together with the 2020 aluminium bed extrusions are designed to negate this effect. (You have attached the 2020 extrusions to the bed mounts with screws and T slot nuts haven’t you) . In fact since implementing these features, the only place I’ve seen z banding to the degree you have experienced is with bed thermostats hence my interest in seeing a print with the bed off all together.


    I don’t even have a bed thermostat. I just use the variable speed controller. If you mean that the knob might make or brake the connection, even if set at maximum, I can tell you that’s not the issue. I would hear those clicks.


    Ok. Let’s move on.

    In that email with your test cubes, you mentioned you had the extrusion multiplier set to .47 but first layer width set to 200%. That’s why your first layer and solid layers are sparse. Please return and keep you extrusion multiplier to 100 in the slicer. Generally you want to keep it within 10% +- but for this purpose, leave it at 100 for the moment. Please set your first layer width to 100% also. If you need more first layer adhesion, increase first layer height, not width.

    During the first layer, make sure you use the Z Baby Step option on the LCD to jog the bed to the perfectly squashed height.


    No, the .47 was for a previous print done with the 2 in 1 firmware. I’m not using that firmware anymore, for the moment.

    The last pics I sent you (with the sharpie) were done with multiplier = 1 and single hot end firmware.

    Thanks for the tips about the first layer.


    In that case. Consider minor over extrusion. Worth playing with the extrusion (flow) multiplier from the LCD during print. Perhaps reduce by 5-10% and re post pictures.

    In terms of the “static retraction” you mentioned. Simplify3D may be your culprit here. Try turning off “Force retraction between Layers” under the advanced tab although IMO this feature doesn’t necessarily work.

    Other settings that may help with over extrusion are “Coasting Distance”, “Wipe Nozzle”, “Extra restart distance (use with caution – especially for prints with lots of retraction)” and when you’ve got a full weekend purely for firstly understanding the subject and repetiers implementation – Linear advance which must be enabled in firmware configuration.h file “#define USE_ADVANCE 1” before it can be used.

    In regards to linear advance, despite what many have said about bowden systems, I was able to get excellent results with long bowden tubes and Linear advance. My figures and settings will be of limited use to you as I used Marlin firmware which uses different settings. For what its worth – as a yard stick though my settings in Marlin were LIN_ADVANCE_K = .95 for a 700mm bowden tube with 1.75mm filament. FYI if it helps marlins default for a direct drive is .22 – meaning my setting was roughly 4.3x that.

    From what I can gather from Reps limited documentation on the subject its useful to know that Repetiers “ADVANCE_L” appears to be the equivalent of Marlins “LIN_ADVANCE_K”. Also that Marlins M900 is Repetiers M233. The scale of this factor is also different between the systems. In Marlin 2.0 a typical scale would be 0.0 – 1.8. In repetier, the advance L range would be more like 0-200 from what I can see. (Ref https://forum.repetier.com/discussion/1200/advance-option-questions )

    The only way to know what works for your system is to run some specific test prints where the range increases line by line, looking at the result and making a judgement as to which line is best, then translating that lines position into your “advance L” in eeprom.

    Some useful reading on Linear Advance (this is made for an older Marlin implementation v 1.x, however the test code can be easily converted for repetier by opening the test code in Notepad++ and doing a search and replace to replace “M900 K” with “M233 Y”) :



    If all of this is a bit advanced for you at the moment, seriously, leave it. There are other ways of controlling flow without resulting in as much hair loss. Saying that, im happy to help wherever possible.

    Needless to say. using a .8mm nozzle, you’re amplifying just about every unwanted artifact 3D printers can produce. Large nozzles always require significantly more tuning that small ones.

    One final thing. Don’t be afraid to try Slic3r. (use advanced mode). It gives you much more granular control of different parts of the print – and preventing unwanted retraction actually works. I personally prefer it over Simplify3D (which IMO is buggy, poorly supported and rubbish – god knows why so many people love it).


    Forgot to say. You may find some value in NOT using S3D’s Automatic extrusion width, especially with large nozzles, can result in significant over extrusion. Try setting to manual and reduce to actual nozzle width (.8 in your case).

Viewing 14 posts - 61 through 74 (of 74 total)
  • You must be logged in to reply to this topic.
Facebook Instagram YouTube