Diagonal lines in prints



  • I bought a used Ultibots D300VS and have spent a few months tightening things, measuring, replacing parts, etc… mostly trying to fix these diagonal lines and I'm at a loss as to what it could be. Maybe it's my power supply? Missed microsteps? Problem with the duet?

    This print is a single wall flow rate test sliced in KISS : 3 mm^3/s/ at bottom to .5 mm^3/s at top (ringing at the bottom from the higher speed)
    -PLA
    -No heated bed
    -Hot end temp (according to duet) doesn't fluctuate more than .1c.
    -Diagonal lines are fairly consistent angles regardless of speed,

    Image may be showing weird in the forum compression so here's a link
    http://i.imgur.com/ODJ0uuT.jpg

    Printer details
    -Ultibots D300VS - Duetwifi board v1.0 running 1.19 firmware.
    -Titan Aero for the extruder
    -Prints are accurate in dimensions
    -Arms + magballs are even lengths
    -Replaced carriages and effector with more diminsionally accurate prints
    -Belts are tight
    -Carriages roll smooth - polycarbonate wheels
    -Some pinched spaced between magballs at the carriages .4mm difference between carriage ball spacing and effector - due to tightening carriages to the rail
    -Power supply cranked up to 24.9v as I noticed a repeating voltage drop every so often on my multimeter..maybe .5v to 1v.. can't remember. Increasing doesn't seem to help.
    -Increased wire gauge to hot end heater and from the power supply to duet.
    -Tried a few different brands of filament - currently this is AtomicFilament which I've read is highly regarded.
    -Twisted wires on just about everything now and spaced them out so there's less chance of emi

    The lines are generally consistent angle regardless of the print size, although one side always looks a little less like lines and more diagonal lump patterns

    My house outlets aren't grounded… could that be a problem? Can I drop a rod in the ground and run a line to the ground on the power supply?


  • administrators

    What are the tower steps/mm, and what layer height did you use?



  • .2mm layer height
    .4mm nozzle
    800 axis steps/mm
    Motor currents 1000mA

    Not sure if it it helps but here's the autocalibrate settings from my config-override
    M665 L360.310 R207.855 H446.611 B140.0 X-0.017 Y0.172 Z0.000
    M666 X-0.231 Y0.166 Z0.065 A0.00 B0.00

    M307 H0 A136.3 C499.7 D2.5 S1.00 B0

    ; Z probe parameters
    G31 T1 P500 X0.0 Y0.0 Z0.70
    G31 T3 P500 X0.0 Y0.0 Z0.70
    G31 T4 P500 X0.0 Y0.0 Z-0.14



  • Left print here was sliced with cura, spiral mode, 20mms http://imgur.com/a/C6KWy

    Also included a snapshot of M122 while that was printing.

    I guess that rules out the slicer


  • administrators

    Nothing wrong so far that I can see. What type of extruder does it have, and what E steps/mm and microstepping do you use? Is the E motor 0.9 or 1.8 deg per step? How wide is the print?

    PS are you running the bed heater in PID or bang-bang mode?



  • Are your pulleys running smoothly? I bought one pulley that had a non-centric bore, as the pattern on the parts is regular that's perhaps worth checking.



  • Extruder stepper is 0.9° E3D NEMA17 12.7N-cm - Titan-Slimline
    Extruder itself is the e3d titan with aero for the hot end

    M350 X64 Y64 Z64 E16 I1 (Steppers 0.9° steppers for all steppers including extruder

    Print is 25mm wide

    No bed running for these, but I run it in PID when I do

    I retested my extrusion and noticed it was slightly short so I bumped it from 812 to 820 steps/mm

    Reran the print and now the angle of the lines has changed! They're less vertical
    http://imgur.com/4QNf2ID

    Maybe I should try to redo the wiring for the stepper?



  • Tried a few things with simplifying, rerouting the wires for the extruder stepper but didn't have any luck.

    Extruder stepper ratings :

    Rated current: 1.4 A/phase
    Rated voltage: 2.7 v
    Inductance: 2.0 mH/phase

    I had this set to 1200ma (85% of the 1.4A rated should be about right?)

    Dropped it to 1000ma, then 850 - didn't help.

    I wonder if my stepper shaft could be bent…



  • @MiR:

    Are your pulleys running smoothly? I bought one pulley that had a non-centric bore, as the pattern on the parts is regular that's perhaps worth checking.

    Great idea, though now that Ive had a change in the pattern based on small adjustment of the extruder stepper steps, I dont think this would be it.



  • Pretty sure the shaft on the stepper wasn't bent - pattern is too tight for that to be the problem anyway.

    I tried swapping in a different stepper on the extruder and still get the diagonal lines - same angle too (which might make sense since I halved the steps.

    17HS08-1004S
    Motor Type: Bipolar Stepper
    Step Angle: 1.8 deg
    Holding Torque: 13Ncm(18.4oz.in)
    Rated Current/phase: 1.0A
    Voltage: 3.5V
    Phase Resistance: 3.5ohms
    Inductance: 4.5mH ± 20%(1KHz)

    Set to 850 ma and 410 mm/s


  • administrators

    See what difference not interpolating the extruder steps makes:

    M350 E16 I0



  • @T3P3Tony:

    See what difference not interpolating the extruder steps makes:

    M350 E16 I0

    Thanks for the idea - I tried it but surprisingly it still looks the same.

    I also tried moving the extruder stepper wires, hot end power and thermistor wires over to the second set of plugs - didn't help : /



  • Ran some more tests where I changed the steps per mm to see if the diagonal lines changed angle and turns out they dont - it's just hard to tell sometimes but the angle of the lines is different on some sides of every cube.

    Now I'm much more suspect of the pulleys as MiR suggested - will check that next.

    I took apart the titan extruder and noticed the idler that rides against the hobb gear has a bit of wobble to it - and I think it's from the gear?

    https://goo.gl/photos/hzUhnTGuvzgFeM6t9

    Flipped the gear and still see it
    https://goo.gl/photos/oziQfsHv4FPEwABk6

    Tried with just a piece of plastic to remove the idler as the problem and moved the gear up to try to eliminate any affect from the shaft.
    https://goo.gl/photos/hkQEasigaQgCNcNu7

    Also tried tightening the grub on the round part of the shaft instead of the flat in case it would take up the play better but didn't notice a difference.

    It needs a full rotation to achieve the wobble, and with the lines as tight as they are I suspect that's not the problem.

    To summarize - factor's that don't seem to affect it
    -Print speed
    -Hotend Temp
    -Heated bed on or off
    -Fan speed from 50% - 100%
    -Part position on bed
    -Part rotation
    -Extruder stepper current changes
    -Lowering extrusion factor in web interface to 85%
    -Increasing stepper wire gauge
    -Twisting stepper wires / thermistor wires
    -Extruder stepper interpolation turned off
    -Extruder tension
    -Filament slack
    -Swapping extruder stepper for a different type (started with .9* and then tried a 1.8*)
    -Two different e3d hotends / extruder configurations (Titan and D300vs stock)
    -Layer height (tried .15, .2, .268 - all look the same)
    -Changing the extruder steps per mm



  • @MiR:

    Are your pulleys running smoothly? I bought one pulley that had a non-centric bore, as the pattern on the parts is regular that's perhaps worth checking.

    Aaaaaahhhhhhhhh I THINK YOU WERE RIGHT. One of my pulley's isn't concentric. Gotta order a replacement. Wow thank you so much for suggesting that - I never would have found it. I'll mark this answered if it proves to be the problem.


  • administrators

    I suspect a problem with your extruder. That's why the angle of the lines changed when you changed the E steps/mm. If I am right, it will also change if you alter the extrusion width in your slicer.

    The actual problem is likely to be the hobbed insert not being uniformly round, and so not driving the filament evenly as it rotates. This could be caused either by a manufacturing defect or by the splines on one side of the insert being blocked with filament or other debris.

    The pulley needs to be fixed too.



  • I think you're right about the extruder - here's a clip of the titan idler that rides tightly against the stepper gear.

    https://goo.gl/photos/Yc6R1PegXFZpS4Nb7

    I suspect this wobble shouldn't be here, though it takes a full rotation which is much longer than the space between the diagonal line steps..so I'm not sure. I think tightening the idler impacts the lines a little bit so that's more evidence.



  • E3D is sending me a new pinion gear, but some more interesting developments..

    Someone with a well tuned D300VS printed the same objects I'm having this effect on (single perimeter, PLA) and they're also seeing the patterns, but observed it doesn't seem to impact the actual surface, just the way light is interacting with the print. They were using the stock extruder.

    I also observed when printing a hexagon shape, the sides closest to the towers had the lines more prominent, and the sides on the opposite were hardly visible. Seen here https://goo.gl/photos/hnc4CBySY3rispkYA If I print next to a tower I can see it on all sides.

    I'm not sure what carriages they were using but I know the stock ones use a screw to tighten the carriages to the rails, and doing so creates a .5mm to 1mm difference in magball spacing between the carriage balls and effector's. I have some eccentric spacers I may reprint carriages for…or I might just let it go at this point : )



  • I printed a 40x40x40 single walled object in vase mode. If I look hard enough, I can see the lines you are talking about. But, dang, I do have to look to see them.



  • @number40fan:

    I printed a 40x40x40 single walled object in vase mode. If I look hard enough, I can see the lines you are talking about. But, dang, I do have to look to see them.

    Thanks for giving it a try 😄 Perhaps I've hit a limit of some kind. I wonder if it's delta specific



  • Woooo I fixed it! 😄 Well 95%

    https://imgur.com/KibBGNV

    I guess the difference the carriage warp was making on the mag ball positions was enough to cause the effect. Switching over to eccentric spacers significantly improved it!


Locked
 

Looks like your connection to Duet3D was lost, please wait while we try to reconnect.