Homeβ€Ί JGAurora A5 & A3Sβ€Ί Getting Started & Troubleshooting

Problems getting a good print, poor quality and looks terrible.

DIYglennDIYglenn Posts: 31🌟 Super Member 🌟
Hey!

I have an issue with my A5 (well it has never been working properly since I purchased it almost a year ago).
The problem now is that the extruder stepper is making a loud grinding noise when retracting. It also seems to be messing up the first layer because of this. The bed leveling etc. is fine, but f.ex. when I have multiple small objects on the first layer (a square, bunch of circles) it strings/drips when moving between them, and doesn't always have filament ready when drawing the circles, ending up with a half circle, or maybe just a line. Almost as if the filament is delayed, and rather spread out when traveling afterwards.

These doesn't stick to the buildplate, and are often ripped off the next time it comes around to do a perimeter.Β 

Not sure if these problems are related, but it seems logical.

I have 8mm retraction at 50mm speed atm.Β 
Post edited by DIYglenn on
Β«1

Comments

  • Samuel PinchesSamuel Pinches Posts: 2,997Administrator
    edited August 2018
    Welcome @DIYglenn!

    - check that the extruder gear teeth are not clogged with filament
    - check that the grub screw on the extruder gear is tight
    - check that the extruder motor cable is in firmly
    - check that the PTFE tube is firmly inserted into the bowden connector, and is not loose.

    Can you perhaps record a video for us to see the problem?
    Thanked by 1DIYglenn
    Post edited by Samuel Pinches on
  • DIYglennDIYglenn Posts: 31🌟 Super Member 🌟
    edited August 2018
    I removed the extruder motor, and cleaned it. It had some "dust" in it, so it might've been a problem. The screw was tight, no problem. The cables were OK, and I took the PTFE tube out and made sure it was all the way in.

    I disabled Z-hop, and increased retraction distance to 8mm, and speed to 50mm/s.Β 

    The sound is gone, but the extrusion is worse than ever. I also think there is something wrong with the X/Y calibration, as the perimeters are separated, and infill is not connected to the outer perimeter.

    My current print is larger and forgiving in its design, so as long as it is made, the first layer can be ugly. It's a 120mm fan vent, but I can't even print parts that are smaller and have a "complicated" design.

    I changed initial layer print AND travel speed to 15mm/s, but no difference.

    I'll post some photos later, and a video if I get the noise issue again so you have something to work by.

    JGAurora has tried helping me the last 8 months. I've finally got all my broken parts replaced, but my printer still doesn't perform as I'd like. Parts are not strong enough to be usable, not printable, or just inaccurate.

    I really hope you guys can help me out, I've given up my printer the last few months, but after reassembling the whole thing, I'd finally like to make it work as intended.

    Attached is my currently printing STL (probably 12-15 hours).Β It can't be attached for some reason...
    I'll just let it print, even though the part won't be perfect. This is second largest print I've done, and since it's square and I made corner discs, I hope it won't curl up like the last large print.

    I need four of these for a large project, where two of them will be more complicated. I just want to get my printer working, so fed up with it, but I still want to use it.

    I got a Wanhao i3 V2 at work which outperforms this thing 10 times when it comes to accuracy and strength of prints, which is sad, the 300x300mm bed is so awesome on the A5, and I'd love to utilize it.Β 

    For "mass production" the A5 is great, just put each print alongside Y axis, and print one by one.Β 

    Man... I just want this thing to work. Sorry for the rant.

    Print is messy with a lot of threads:

    The skirt is printed nicely, except maybe the first 10cm which typically creates a bunch of dots while being primed.

    The X-axis "squiggly" line is probably only supposed to be travel. If it is supposed to be extruded, then it is done poorly. Notice the line from the circle, the nozzle seems to be oozing.

    Oozing continues for every time the nozzle comes back and adds a circle (strange how cura doesn't slice it to do it all in one go (three circles/perimeters)

    Notice how the skirt and outer perimeters look fine, but the circle is underextruded, and not sticking to the bed properly - bed leveling is not the issue.

    Another angle. Messy circle and small details.


    Hardly a circle... This came off and is missing on the finished print.
    Stringing is bad. The circle perimeters aren't connecting, and it seems like the filament that SHOULD be there is spread on the traveling lines. Also notice at the bottom, there is a slight gap on the skirt, and it is pretty bad on the perimeters of the model. That happens especially on the X+ side of the bed, no matter what I print - always a gap between the perimeters.

    Messy after travelling. Also notice the missing filament in the perimeters on the bottom of the honeycomb pattern.


    Notice the missing filament in the corner discs. Doesn't matter, but for a model that would be awful... A lot of stringing going on.

    First circle when starting the print, and then a looong line where it's supposed to just travel.


    I'll update other posts with the "finished" print.
    Post edited by DIYglenn on
  • DIYglennDIYglenn Posts: 31🌟 Super Member 🌟
    (Not sure if my STL attachment is uploaded? It's 18.5MB, is there maybe a limit...?)
  • Samuel PinchesSamuel Pinches Posts: 2,997Administrator
    I’ll investigate the failed upload later on.

    Try slow down the print to 35mm/s and see what changes.
  • DIYglennDIYglenn Posts: 31🌟 Super Member 🌟
    edited August 2018
    I'm printing at the suggested 30mm/s from JGAurora. They say this is max.

    I'll upload some photos later when I'm back home.

    Edit;Β 
    Photos from the finished print:

    A LOT of stringing going on. The layers look OK, except the corner disks which are bad, and the first layer as well...





    From underneath. Notice the gap between perimeters and infill, and a rather ugly first layer... There was also a bit of edge lift, even though the bed is as level as I can get it.





    Removing the disks reveals the biggest problem with my A5 printer - very weak prints. Instead of tearing alongside the edge, it rips the entire layer off. I'm printing at 215C and 65C bed, maybe I need to increase it even more?





    The infill underneath looks bad as well...


    As you can see the infill has a gap around the hole, it is not fused together.


    Post edited by DIYglenn on
  • DIYglennDIYglenn Posts: 31🌟 Super Member 🌟
    I tried the profile that was linked in the Wiki, and while it is set to 210/60C the printer still prints at 215/65C...Β 
    Anyways, this is much faster than my other profile - twice as fast actually, and it seems to be performing pretty much the same as before. So at least we can probably say that the print speed isn't an issue. Actually faster printing seems to clutter less than before.Β 

    The only difference I did was increasing 0.8 to 1.2 on wall thickness (3 lines) and top/bottom. I also turned on Z Hop.

    There is still some filament being spread around though, so maybe retraction and temperature settings needs to be adjusted. The nozzle is oozing when not printing.Β 

    However the Wiki says 6.5-8mm retraction at 75-150mm/s speed, while the profile linked has 6.5mm and only 2.5mm/s!

    Anyways, the biggest issue except missing filament and oozing is perimeter separation on X+ side of parts.
  • Samuel PinchesSamuel Pinches Posts: 2,997Administrator
    I think with that profile, the first layer temp is 215C, but the print itself should still be 210C.
    Maybe try dropping temp by 10-20C? I'm assuming you've installed the community firmware? Some photos would be really helpful - uploading of photos seems to work, but tracking down this file upload problem is keeping me stumped at the moment!
  • DIYglennDIYglenn Posts: 31🌟 Super Member 🌟
    Heh, I get your frustration.Β 
    Added some photos now. Will update with more from the finished print as well.Β 
    Yes, definitely using the the community firmware. Fixed a 11.x% under-extrusion for me :wink:
    Thanked by 1Samuel Pinches
  • Samuel PinchesSamuel Pinches Posts: 2,997Administrator
    edited August 2018
    looking at your photos:

    Β (1) try slow your first layer to 12mm/s
    Β (2) drop temp 10-20C

    Β Perhaps turn retraction off for the moment, retraction is not really important for this problem, and I don't think that profile is really tuned properly... ask me later on once you have this problem solved, and I can share some other profiles (from other people) with you.
    Post edited by Samuel Pinches on
  • DIYglennDIYglenn Posts: 31🌟 Super Member 🌟
    Hi!Β 
    Please see the post further up. I've added photos from the added print.
  • DIYglennDIYglenn Posts: 31🌟 Super Member 🌟
    Here are the result from the print using the profile on the wiki. You were correct, only 215/65 for the first layer, and then 210.Β 

    It seems to be better (and much faster!) than before, but still not a very good print.Β 

    215/65C for the first layer.


    Still weak circles, and stringing/dripping when moving

    Separated perimeters in the circle, messy around it.


    Notice the outer line (down, X+ side) has separation between the perimeters, but not the left side. Messy around the circle, filament that should've been elsewhere.


    Perimeters are much better on the X- side (top)



    Helper disks has a initially poor infill. Seems like the printer doesn't wait for the extruder to be ready with filament?


    Perimeters separated at the bottom. Infill seems to overlap with inner perimeter.



    Large infill/perimeter separation on Y- (?) axis. Also separated between circle and infill.


    There are so many problems with the print... I cannot understand what's wrong. I've never had any problems with other printers
  • DIYglennDIYglenn Posts: 31🌟 Super Member 🌟
    edited August 2018
    Seems like an issue with Cura is part of the culprit as well. No retraction on initial layer!
    https://github.com/Ultimaker/Cura/issues/2476#issuecomment-357624835

    Horrible with prints I usually do; Cases with vents/holes in them.Β 

    Post edited by DIYglenn on
  • DIYglennDIYglenn Posts: 31🌟 Super Member 🌟
    Tested with combing "off". DEFINITELY helped me with printing small parts, no more ripping off printbed.
    Unfortunately this means A LOT of retracts, super slow prints.Β 

    Will wait and see if the fix which allows this setting for first layer will be pulled in, or if I'll have to compile it myself.
  • DIYglennDIYglenn Posts: 31🌟 Super Member 🌟
    edited September 2018
    Testing with "XYZ Calibration cube" from thingiverse.Β 
    It lifted in a couple of the corners when printing with 205C/65C. Will see if 200C/55C is a better temperature.

    It did however reveal some calibration issues:
    Z: 20.04mm (+0.04)
    X: 19.87mm (-0.13)
    Y: 19.95mm (-0.05)
    This is when slightly pinching the calipers diagonally (not touching top/bottom layers) so the measurements are accurate. Also tried rotating slightly when measuring to confirm the measurements are correct.

    This is probably not too bad, but I guess enough that some larger prints would be inaccurate. This is with the newest community FW.

    Also, when connecting with Repetier-Host, I get no information under "Firmware EEPROM Settings" (M503).Β 
    Isn't this set in the custom firmware?!

    I also have a ~7mm under-extrusion when measured (using Repetier-Host to feed).

    I don't understand... Could there be something wrong with my FW?
    Post edited by DIYglenn on
  • DIYglennDIYglenn Posts: 31🌟 Super Member 🌟
    Sorry for spamming here, just posting my thoughts and debug.

    So I could not get M503 to work. But I KNOW I have the new firmware, because the printer is acting differently, faster movement, then very slow movement when homing - so the FW is the 1.1.8 one.

    I reuploaded the same 1.1.8C, and ran the commands - And guess what, M503 works...
    Unplugged, tried again - Same.

    Will do some more tests (like e-steps etc) and see if it has been corrected.
  • DIYglennDIYglenn Posts: 31🌟 Super Member 🌟
    I followed this tutorial again

    I ran M502 - Then M500 as the suggested after connecting the LCD.


    When issuing "M503" this is the result:

    20:25:08.361 : N14 M503*16
    20:25:08.363 : echo:  G21    ; Units in mm
    20:25:08.364 : echo:  M149 C ; Units in Celsius
    20:25:08.366 : echo:Filament settings: Disabled
    20:25:08.367 : echo:  M200 D1.75
    20:25:08.367 : echo:  M200 D0
    20:25:08.368 : echo:Steps per unit:
    20:25:08.370 : echo:  M92 X80.00 Y80.00 Z800.00 E100.00
    20:25:08.371 : echo:Maximum feedrates (units/s):
    20:25:08.374 : echo:  M203 X500.00 Y500.00 Z15.00 E25.00
    20:25:08.375 : echo:Maximum Acceleration (units/s2):
    20:25:08.377 : echo:  M201 X1000 Y500 Z100 E5000
    20:25:08.379 : echo:Acceleration (units/s2): P<print_accel> R<retract_accel> T<travel_accel>
    20:25:08.382 : echo:  M204 P800.00 R800.00 T1000.00
    20:25:08.386 : echo:Advanced: S<min_feedrate> T<min_travel_feedrate> B<min_segment_time_us> X<max_xy_jerk> Z<max_z_jerk> E<max_e_jerk>
    20:25:08.388 : echo:  M205 S0.00 T0.00 B20000 X8.00 Y3.00 Z0.30 E5.00
    20:25:08.390 : echo:Home offset:
    20:25:08.391 : echo:  M206 X0.00 Y0.00 Z0.00
    20:25:08.391 : echo:Mesh Bed Leveling:
    20:25:08.392 : echo:  M420 S0 Z0.00
    20:25:08.394 : echo:  G29 S3 X1 Y1 Z0.00000
    20:25:08.395 : echo:  G29 S3 X2 Y1 Z0.00000
    20:25:08.396 : echo:  G29 S3 X3 Y1 Z0.00000
    20:25:08.397 : echo:  G29 S3 X4 Y1 Z0.00000
    20:25:08.399 : echo:  G29 S3 X5 Y1 Z0.00000
    20:25:08.399 : echo:  G29 S3 X1 Y2 Z0.00000
    20:25:08.400 : echo:  G29 S3 X2 Y2 Z0.00000
    20:25:08.401 : echo:  G29 S3 X3 Y2 Z0.00000
    20:25:08.403 : echo:  G29 S3 X4 Y2 Z0.00000
    20:25:08.404 : echo:  G29 S3 X5 Y2 Z0.00000
    20:25:08.405 : echo:  G29 S3 X1 Y3 Z0.00000
    20:25:08.406 : echo:  G29 S3 X2 Y3 Z0.00000
    20:25:08.408 : echo:  G29 S3 X3 Y3 Z0.00000
    20:25:08.409 : echo:  G29 S3 X4 Y3 Z0.00000
    20:25:08.410 : echo:  G29 S3 X5 Y3 Z0.00000
    20:25:08.412 : echo:  G29 S3 X1 Y4 Z0.00000
    20:25:08.412 : echo:  G29 S3 X2 Y4 Z0.00000
    20:25:08.413 : echo:  G29 S3 X3 Y4 Z0.00000
    20:25:08.414 : echo:  G29 S3 X4 Y4 Z0.00000
    20:25:08.415 : echo:  G29 S3 X5 Y4 Z0.00000
    20:25:08.417 : echo:  G29 S3 X1 Y5 Z0.00000
    20:25:08.418 : echo:  G29 S3 X2 Y5 Z0.00000
    20:25:08.419 : echo:  G29 S3 X3 Y5 Z0.00000
    20:25:08.421 : echo:  G29 S3 X4 Y5 Z0.00000
    20:25:08.422 : echo:  G29 S3 X5 Y5 Z0.00000
    20:25:08.423 : echo:Material heatup parameters:
    20:25:08.425 : echo:  M145 S0 H205 B70 F255
    20:25:08.425 : echo:  M145 S1 H240 B110 F0
    20:25:08.426 : echo:PID settings:
    20:25:08.428 : echo:  M301 P35.30 I4.35 D71.57
    After testing the extruder I got 11.76mm of under-extrusion.
  • DIYglennDIYglenn Posts: 31🌟 Super Member 🌟
    These are the settings I get from Repetier-Host:

  • DIYglennDIYglenn Posts: 31🌟 Super Member 🌟
    OK, so I don't know why the custom firmware has 100 steps/mm?Β 
    Is this correct for everyone else?
    My calculations gave me E113.33 (-11.76mm correction).
    I used this guide to figure it out.

    Weirdly enough, it's still 8.25mm under extrusion afterwards?!

  • DIYglennDIYglenn Posts: 31🌟 Super Member 🌟
    I did the same again, using the new information, E123.52. And no matter how perfect my measurements were, I still get under extrusion! Now 3.81mm.Β 
    Try again...
    E128.6...
    Did slightly more than 100mm now...Β 

  • DIYglennDIYglenn Posts: 31🌟 Super Member 🌟
    edited September 2018
    I need help...
    This doesn't make sense...
    Is Repetier-Host doing something wrong when feeding?

    The last time I tried now, it started grinding...
    Post edited by DIYglenn on
  • DIYglennDIYglenn Posts: 31🌟 Super Member 🌟
    edited September 2018
    OK, so I used the (MUCH) slower extrusion program on the actual printer. It now only under-extrudes 3.46mm.

    After a couple of tests, using the JGAurora built in extrusion program (concluding that Repetier-Host was pushing to fast), I find that E103.59 gives the correct amount.
    Might have to fine tune and test more though...Β 
    Post edited by DIYglenn on
  • DIYglennDIYglenn Posts: 31🌟 Super Member 🌟
    OK, I get it. I shouldn’t mess with X Y calibration and calibration cubes because the setting is purely mathematical. Teeth’s and pitch of belt.Β 
    I’ll still do some measurements to double check (move axis 200mm and measure).Β 

    There’s probably reason to suspect multiplier here, check shell thickness.Β 
  • Samuel PinchesSamuel Pinches Posts: 2,997Administrator
    Sounds like the extruder stepper driver is underpowered and is skipping steps. I would advise checking the tuning of the VREF on the E motor driver.
    https://jgaurorawiki.com/a5/stepper-drivers

    Thanks for posting your journey as progress things - its definitely helpful to see what you've tried and what problems you've faced.
  • DIYglennDIYglenn Posts: 31🌟 Super Member 🌟
    Thanks!Β 
    That's what I thought. Maybe some others will see the same issues.Β 

    Actually I don't get that much grinding anymore. I think it was skipping because the temp was a bit too low and Repetier-Host a bit too "aggressive" when feeding. When using the display on JGAurora it never skips a step.

    I ended up with E103.10 as the correct steps to get 100mm fed through. At least that's as close as it will get atm.

    I also increased flow to 105%, because my 1.20mm walls were 1.15mm (and I rounded up a bit).Β 
    I'm now VERY close to 20x20mm on the cube, so this seems to be the right thing to do. My prints are a bit uneven though. That is, my cubes have rough sides in addition to some waves.

    The waves are probably from vibrations (jerk/acceleration issues?)
    The rough sides/lines could be that the filament is inconsistent, so I'm ordering some new spools from a known vendor.Β 
  • DIYglennDIYglenn Posts: 31🌟 Super Member 🌟
    So my last print finished. In addition to the 105% flow, I increased from 200C to 210C. I also lowered the inner wall speed to match the outer wall better, holy crap that did a difference!!

    My part is a 100% perfect 20.00x20.00x20.00mm, but the walls are now 1.30mm instead of 1.20mm (8% too thick).
    This is strange when I initially had walls that were almost 5% too thin! The temperature seems to have a big impact, and I will have to investigate, trying with 100% flow again.

    But I might ignore it as well, 1.20mm vs 1.30mm doesn't really matter, and the dimensions of the cube is perfect.

    The big difference here is the STRENGTHΒ of the part. Holy crap, I've never gotten anything that looks this great that isn't easily ripped apart.Β 

    Probably because:

    1. The 5% increase in flow fills better between the perimeters. Everything is simply 5% thicker in all directions than before.

    2. The increased temperature reached a "magic" point where it fuses the layers and perimeters better. With the added plastic from the increased flow, this also gives a much better bond.

    3. The speed is now 30mm/s for outer wall, 40mm/s for inner, and 60mm/s for infill. The inner wall had 60mm/s before, which gives a difference of 100% between outer and inner wall - this is not a good idea if the walls are to properly fuse together.

    Will continue testing, temporarily concluding that the filament is the source for the lines in my print (inconsistancy in filament thickness) and that the vertical waves/ripples/ghosting is because of vibrations or jerk/acceleration settings that might need to be adjusted.Β 

    Anyone else that have managed to get rid of ripples in their prints? Especially visible on flat sides or sides with text..Β 
  • DIYglennDIYglenn Posts: 31🌟 Super Member 🌟
    I've set acceleration/jerk to what this blog suggests on the Wanhao i3 we have at work, which pretty much removed all ripples/ghosting on prints. But the settings on the JGAurora A5 are already as low as these and still struggles.

    How low will I have to go with the JGAurora?!Β 

    Suggestions? The belts are tight, so that shouldn't be the issue. I've replaced the springs on the bed - they are stiffer than ever.
  • Samuel PinchesSamuel Pinches Posts: 2,997Administrator
    Do you see ripples on all sides of the print? Can you please post a photo of what you are seeing? Could it be salmon skinning?
  • DIYglennDIYglenn Posts: 31🌟 Super Member 🌟
    Sorry for my late reply. I've had some time late evenings now, and tried to calibrate further. No matter what I do, I can not get rid of the vibrations and uneven layers.Β 

    Here's an example of a cube from Thingiverse.

    What I can see is underextrusion (top layer), wobble/vibrations on the sides, and you might also be able to see the ringing I'm experiencing.Β 


    Here are some pics which I should've posted earlier:
    I've continued printing the 20x20mm cube I made with different settings.

    This is from my post further up, the strength issue. I went with a 105% flow, and after also increasing to 210C from 200C I got a very strong print, but not prettier.

    200C, 100% flow at left, EASILY torn apart, and 200C, 105% flow at the right - possible to break, but not easy.




    210C, 105% flow.Β 

    Part is "unbreakable", but still not a great looking part. You can easily see both horisontal lines and the vertical ringing.



  • DIYglennDIYglenn Posts: 31🌟 Super Member 🌟
    To compare, here's the same filament on a Wanhao i3 printer I've been using at work, it was printed at the same speed as the A5, a bit faster than supported, so it does have some ringing. But the layers are EXTREMELY good compared to the A5...



    The white part is from the A5


  • DIYglennDIYglenn Posts: 31🌟 Super Member 🌟
    I spent a lot of time last night, trying to calibrate further. I did a PID calibration of the hot end, the result was approx (50, 7, 150). The temp is a steady 210/210 throughout the whole print.

    I also calibrated acceleration and jerk, starting at 500 acceleration and 10/10 jerk, and working my way down to 300 and 5/5. Still not good though.

    Here's my progress:

    Left to right;
    1. 105%, 210C. Part has 1.30mm walls instead of 1.20mm
    2. PID calibrated, Acc/jerk changed
    3. Acc and jerk changed to current (300 and 5/5).
    4. Reduced flow to 100%. Walls are approx 1.27mm
    5. Filament measured, varies from 1.72 to 1.75, using 1.73 as filament diameter in Cura, flow 95% - Walls are 1.21mm
    6. Reduced flow to 94%, walls are 1.20mm, as the design.Β 

    I also lowered the bed as much as possible, and moved the Z axis stop lower. That way the springs have more tension then before, but this didn't remove the ringing. The belts are nice and "plucky".

    As the amount of plastic is slightly reduced, the part is a bit smoother. While the difference is not that apparent here, it has improved between print 1 and 6.Β 

    Please bear in mind that I have never done any calibrations on Wanhao i3, just printed. I find it strange that this printer cannot produce a straight wall....



    I'm running out of ideas. I just cannot get a good looking print. The gcode moneycat printed "OK" when I got it, and it did look much better when I sliced it myself, but the part was very weak, so I stopped using the included filament, as it always had issues.

    After now reducing my flow settings etc, I am afraid that I might see some weak parts again, but I'm focusing now on trying to get rid of those lines before printing anything larger.

    Can anyone else test this on their printer? It take approx 22 minutes to print. (0.2mm layers)

    My print speed is very low (30mm/s outer, 40mm/s inner, 60mm/s infill) so I can't see that this should be a problem either.Β 

    Here's some photos of my latest print (the rightmost)Β 








Sign In or Register to comment.