Diagonal artifacts showing up in new delta setup
-
Converted a Tevo Little Monster to use the Duet Ethernet. All has been going well, starting to dial in my config. This is my first time with a delta, but I've been using RRF on a custom core-xy for a few years now.
The last main issue I'm having is, these cris-crossing diagonal lines are showing up in my print. I'm going to presume this is directly related somehow to the delta geometry/movement? Doing searching, I hear people calling it "salmon skin"?
Couldn't figure out how to embed the pics here, so here's some links showing the issues:
All at 200 micron, which is a full step for my steppers (1.8 deg stepper, 20 tooth pulley, gt2 belt)
https://drive.google.com/file/d/1BKoP9j1zwhaWNLLdRAV_qjLCzjtQmT–/view?usp=sharing
https://drive.google.com/file/d/1_h39kgmSfMrC39F0hY8pawxn-FhG90qB/view?usp=sharing
I was doing a lot of interactive tweaking on these (speed, temp), so the quality is sort of all over the place, but the diagonal lines seem consistent. And it also illustrates they show up consistently despite speed or temp.I've found several links specifically for the TLM, and this artifact, but that's using the stock MKS SBase board, running DRV8825's (which I've replaced with the Duet) : They suggest adding a couple 'tl smoothers' to help resolve this issue, and printing a a 'special golden-ratio layer height'. It seems to me the TMC2660's wouldn't need any of this though? And they don't really seem to explain how to determine this 'golden ratio' layer height.
Here's some bits from my config.g that probably relate:
[[language]] M906 X1700 Y1700 Z1700 E1700 I30 ; Set motor currents (mA) and motor idle factor in percent ; Stepping -------------------------------------------------- M350 E16 I0 ; Extruder: Configure 1/16 microstepping, no interpolation M92 E418.5 ; Extruder: Configure steps per mm at 1/16 microstepping M350 X16 Y16 Z16 I1 ; XYZ : Configure 1/16 microstepping with 1/256 interpolation M92 X80 Y80 Z80 ; XYZ : Set steps per mm at 1/16 microstepping ; Speeds -------------------------------------------------- M566 X2000 Y2000 Z2000 E3000 ; Set maximum (jerk) instantaneous speed changes (mm/min) M203 X18000 Y18000 Z18000 E3000 ; Set maximum speeds (mm/min) M201 X2000 Y2000 Z2000 E2000 ; Set accelerations (mm/s^2)Any thoughts on how to improve this appreciated.
-
The "golden ratio" is .1618 layer height. Give it a try.
-
A fine moire on flat surfaces is normal with delta printers. It can be reduced by using 0.9deg motors instead of 1.8deg. See https://miscsolutions.wordpress.com/2015/08/01/upgrading-the-large-delta-printer/ for the difference that makes.
However, your second photo shows something much worse. I suspect a problem with your extruder. What extruder are you using? Is the hobbed part that drives the filament really round, and not clogged with filament debris?
-
I've been considering .9 deg steppers, thanks for the link.
The extruder is a knock-off Titan that came with it. Plan to replace someday. The hobbed gear is totally cylindrical (visually at least, I've not taken it apart and checked with my calipers): no indentation like I'm used on on my Makerbot or Bondtech. But, it looks nice and clean, I only have a few hours printing on the whole machine. However, again, that was a crazy print test, with speeds up to 300mm/sec with acceleration\jerk to match (just to test it), so I wouldn't be too concerned about it looking bad. Just testing the limits of the machine. I showed it simply because it had the same diagonal lines as everything else.
I plan to replace that whole thing eventually, one step at a time.
-
The "golden ratio" is .1618 layer height. Give it a try.
Derp: I'm familiar with the Golden Ratio, but didn't actually remember what the number was (nor did I even google it, sigh). Yep, that guys post used the exact golden ratio as layer heights. I'll try it in the future too.
Thanks! -
I read that the layer height should be a multiple of the distance matching 1 full step motor…
-
Reprap forums is being stupid right now, but if you jump on and search for "removing vertical lines" there is a some good info on it.
-
I don't believe the golden ratio is at all relevant to the z axis of a delta printer. Moving the z axis up and down is as simple as increasing or decreasing the step position of all motors equally – so, as fma said, using a distance that is compatible with the full-step resolution (or half, or quarter) would be, theoretically, ideal.
-
One cause of vertical lines in a print is DRV8825 and similar drivers failing to implement low-current microstepping. Changing the layer height so that it is not an exact number of full steps causes the problem to occur at different positions in each layer, so the positions do not line up and the lines are much less visible.
The symptom being described at the top of this thread appears to be different, although if the layer height for these prints is not an exact number of full steps then it could be related.
My guess is that it's extrusion-related. If so, then if the print is repeated with a different extrusion width, then the spacing of the lines will change in proportion.
