What is this?
So earlier today, I was thinking that I was having a Pressure Advance issue. I have since run many A PA test, tower and pattern method. It is tuned just right. I tried running with a factor of 0.0001 and also 0.0025. No joy. I still get these repeating artifacts along the edge. They are wavy, and cyclical, occurring every 8 layers or so I'd say.
Any ideas? Happens on 3 edges that are not the seam. If I print the seam in the middle of the part, happens on all 4 edges.
57 Replies
Does it happen in vase mode?
It actually does.
??? those numbers look like an order of magnitude too small
TUNING_TOWER COMMAND=SET_PRESSURE_ADVANCE PARAMETER=ADVANCE START=0 FACTOR=.0025
And I also tested using 0.0001 instead.
This for tower mode. For the pattern, I just dload the code and run it. My PA is set to I believe 0.005 for ASA. I'm currently back home, so I can't really check till tomorrow. Either way, it doesn't seem that PA does anything to the print.
Maybe z wobble related?
Let me change filament and profiles to make sure it isn't material related. I've had problems with ASA since the start, though usually it was just with AzureFilm ASA. Is there some way to "test" for z wobble? I mean, other than printing the tall round hollow towers, that is.
Before changing the filament, I decided to do a round tower test. Came out looking great.
So it seems that it only affects sharp corners. The sharper, the more pronounced. Round edges are not affected. Very sharp corners even show some small holes. Images to follow.
Your PA is probably way too high
Where are you setting it? Do you also have it in custom gcode for the filament in the slicer?
(which is the right place if you ask me)
Also what is your square corner velocity and acceleration?
@elcoj Yes, I am currently setting and altering PA when needed in the slicer. I have a profile for each material I print. I have run the pattern tests quite a few times, now, and still getting the same results.
what's your filament start gcode?
copy paste it here to check
As for the acceleration, I have backed down from 15k to 7.5
yeah that's very low. I use 0.020. Shouldnt cause this effect
I just noticed something...
Ahhh, never mind.
My acceleration was at 15k but that was before the print actually started.
I think because you print with 0.6 line width and a 0.6 nozzle, you don't have much layer squish, so the line doesn't stick that well
Ok.
and when it is pulled around a corner, it cuts the corner
try 0.72 line widths to see if that makes a difference
Sounds plausible. So I should Increase the thickness, right?
Ok
Will do.
Just finish out this part and will change it and report back.
Thanks for the tips!
in general I prefer to use 120% of nozzle diameter for line width
that guarantees that the filament is pushed against he previous layer to expand outwards before being dragged anywhere
Accross the board?
yeah, or 120% for spacing
I think I use spacing
I think that's easier to reason about, because that sets the actual distance between the lines
so when designing I can take it into account
Ah, I see.
expand the help for the difference
So if I set 120% spacing, in effect it gives me a 0.6*1.2=0.72 line, right?
play around with it a bit, I see I currently use a slightly smaller width
but have used 0.72 for a long time
top infill smaller than nozzle to have some ironing on the top layer
but as a first test, use 120%, and 120% for infill also makes sense
What about those zero values?
if the problem goes away with fat lines, you can try reducing it later
check the mouse over
Does the slicer assume 0.6 on those?
zero means default
Got it.
Thanks.
Will test after this part completes and come back with results.
Much appreciated.
is your nozzle diameter correct in slicer and klipper?
check the mouse over on the default width by the way
increased temperature will also work with layer adhesion of course
and lower acceleration for (outer) perimeters
but I think adhesion in the corners being not good enough for the acceleration is plausible
they also seem to curl up a bit
if you see them curling up during print, then the solution might be more in cooling
or upping chamber temp
which filament is this?
Diameter is correct in slicer and Klipper. Wasn't sure about the last one, but yeah, checked printer.cfg and we're all good there. The default width is at zero right now because I have everything else filled in. So it is greyed out for now. I assume that if I go back in at a later time, I'll set it to something like 0.72.
I tried varying temperatures from the low 190 up to 260. The filament suggests a range from 200 to 240. No changes to be felt other than at lower temps I seemed to see a bit more deformation. Also, the sheen of the parts, which I was expecting.
I'll change the acceleration for the perimeters next, and bring it even lower.
You are right, I see the corners curl up, then settle down then curl up and over and over it goes.
I am printing Spectrum ASA 275 in black.
How do I set corner acceleration in the slicer? I can't see any setting in there for that.
what's the chamber temp?
When I've seen this effect it usually means I need more temp & more cooling
higher chamber temp with more air flow is usually better for warping
@sherbs @elcoj I'll try both. In the meantime, I got the print to an acceptable level, though I had to reduce the perimeter acceleration all the way down to 1000 mm/s2. Don't know what temperature I have in the chamber, though I have the bed of my 500mm printer set to 70 at this point. I'll try upping the temp and then upping the cooling.
that's far too low for asa. I do 110 for ABS and it should be similar for asa. Get a cheap B3950 sensor and add it as a chamber sensor
Will do. As for the temps, yes, I would also go higher with something like AzurFilm. With this one in particular, I was just following the manufacturer recommendations. In any case, I'll do a temp tower or something.
yeah I find manufacturer's recommendation are usually for slower less capable machines. Good as maybe a starting point but I wouldn't be strict in following them
Makes sense... Slowing down 8s also a means of getting more time under heat and more time under cooling if all else is equal...
So yeah... Upped the print temp to 270. Plate is all the way up to 110. Fan is now at 40%. My widths are all at 120% save for top layer at 0.5mm and 80% flow with 70% solid fill overlap.
While I see improvement, nothing comes close to what I saw with the acceleration decrease.
I just can't figure out what changed for the parts to start doing this. I already had more than a quarter spool spent at the very same settings.
Bottom to top. Most recent print at the tip. Improvement is visible mostly in the left side. Right side seems to be mostly the same.
For comparison, this piece below was printed in the same material less than 24 hours before the issue started. Absolute same profile. While the parts had issues of its own, the corners were not one.
have you tried drying your material?
Yes. I keep the filaments I am currently using in a drybox.
I am no expert and this is an old topic, but how’s your belt tension?
As far as I could tell at the time, seemed to be pretty decent, though in retrospect, I have since tightened it up a bit, so it could be that it had something to do with the issue.
Because from the pictures it looked as if when you change directions the tool heads hangs in one spot for a small time thus creating this corner problem
I slow-mo recorded the print process to see if I could tell something was up. From what I can see, I can't discern any slow down in the corners other than the deceleration.