3D printing

[pastera]

i am printing a resonance tower now after finally fixing the bed, it is now at 4500mm/s acceleration and still shows no obvious oscillations. it cannot be right....
it jerks, though, like it is real.

ahh, ok, at 5500 it started showing. now need to sort it out what to do with it all.

this shit is just out of this world. the instructions are:

1. Compute the ringing frequency of X axis as V · N / D (Hz), where V is the velocity for outer perimeters (mm/sec). For the example above, we marked 6 oscillations, and the test was printed at 100 mm/sec velocity, so the frequency is 100 * 6 / 12.14 ≈ 49.4 Hz.

from what sign in the sky am i supposed to know what was the the actual velocity of the print? the klipper screen shows max velocity in the center of screen as 300.
as print still goes - the speed in left upper status window fluctuates from 120 down to 30. wtf?

Look at the wall speed in CURA - that's what it would have been
example attached is 90mm/s

All of those varying speed and flow rates tell you exactly what is happening after all of the tuning and acceleration calculations are done.
Keep an eye on the flow - if it hits over 13-14mm^3/s your running up against the max flow rate with a standard nozzle.

 

[paul73]

yeah, the process to setup it all is not for the faint-hearted for sure.

but it seems to be worth it. i am gonna convert old one to klipper this evening, as i have a working reference now after whole day of this masochism.
prints quite well after all the tunings. i have it now at the 300/5/4000/4000.
kinda wonder what is the effect of the first 300 there for 'velocity' as my walls are set to 120 right now, may be 150 would work too, but it did not do very good with marlin, so i may be pushing it already.

Here is the test cube with these settings rendered like those from yesterday, an actual part I print now looks smooth. Better than marlin one.

 

[richc]

So I received the Qidi i-Fast printer. Did a couple test prints that worked. That did a print of my own design and it worked well.

I needed to make some changes and did so, and then everything went to hell in a hand basket. Print wouldn't adhere to the platen. Just weird problems.

The mfr suggested washing the platen and taking the print head off and making sure there was not damage. I did both and also leveled and adjusted the nozzle gaps.

Still not working... and then I saw it.

I guess when making changes to the design I accidentally nudged the object down below the zero point. And nudged an object or two out of alignment. It was trying to start printing in midair.

Now we are happily printing again. But I also learned a fair bit about taking the damn thing apart!

 

[pastera]

So I received the Qidi i-Fast printer. Did a couple test prints that worked. That did a print of my own design and it worked well.

I needed to make some changes and did so, and then everything went to hell in a hand basket. Print wouldn't adhere to the platen. Just weird problems.

The mfr suggested washing the platen and taking the print head off and making sure there was not damage. I did both and also leveled and adjusted the nozzle gaps.

Still not working... and then I saw it.

I guess when making changes to the design I accidentally nudged the object down below the zero point. And nudged an object or two out of alignment. It was trying to start printing in midair.

Now we are happily printing again. But I also learned a fair bit about taking the damn thing apart!

It's all about the first layer.
Ohh, and really nice printer.

 

[richc]

It's all about the first layer.
Ohh, and really nice printer.

I bought a QIDI printer 7 or 8 years ago. Became a hobbyist with some practical uses.

I decided on a gag gift for my son and two son-in-laws for Xmas. Sadly the old QIDI was beyond repair. So I bit the bullet and broke the budget.

The state of the art has come a long way. I see lots of very cool improvements and enhancements. Someone was paying attention. Like the platen being held in place with strong magnets, but removable. Generally I'd have to wait some amount of time to do a second print while the platen cooled down. Now I can just take it out and replace it with a fresh one.

The heated print cavity in a nice touch too. Build quality is amazing. Plus the breadth of materials one can use is much improved. I got a spool of something that is supposed to approximate wood. Something to play with.

It's all about toys! Being a crusty old software guy I'm wondering if there are some interesting business applications out customers might find useful.

 

[paul73]

i've done some more additional tuning/massaging and now it runs at 150mm/s walls, 500/5/4500/4500 speed. klipper is indeed amazing. and this ender platform is pretty capable too.
but, it would not be so easy to switch to it with no marlin experience. all in its good time.

 

[bigblue]

What’s the largest build size on that?

 

[FancyGunz]

10min 54 seconds, PETG

 

[paul73]

10min 54 seconds, PETG

That was last calibration cube I printed at the 500/5/4500/4500, .24mm layers, 150mm/s lines. Above it results were getting much crappier.



But all this calibration bs matters less, the actual model print was excellent.

 

[bigblue]

Lot of gaps in there?

 

[paul73]

Lot of gaps in there?

nope. it is no gap, it is how seams - layer ends look like with random placements. this stuff should not be looked under a microscope.
i am not saying it cannot be improved even more, but, it is pretty damn good for that speed.

or may be i screwed up something in settings of the profile when that one was printed. go figure. test cube does not seem to have those. will see, will print one more.
an amount of fine tuning this shit requires is endless.

 

[pastera]

10min 54 seconds, PETG

What was that done on?
Nozzle? Layer Height?

 

[pastera]

nope. it is no gap, it is how seams - layer ends look like with random placements. this stuff should not be looked under a microscope.
i am not saying it cannot be improved even more, but, it is pretty damn good for that speed.

or may be i screwed up something in settings of the profile when that one was printed. go figure. test cube does not seem to have those. will see, will print one more.
an amount of fine tuning this shit requires is endless.

Get rid of random Z seam - extra travel movements and not needed if pressure advance is set properly. Reduce retraction to <1mm
Also make sure coasting is turned off if pressure advance is on.
If coasting is off then your pressure advance is set too high - rerun the calibration print or drop PA by 30%
Print a cylinder with the seam aligned - if the seam dips reduce PA; seam is pronounced: increase
Iterate at half the previous change to get a perfect value (30%->15%->7%) or use your judgement to SWAG based on how big of a difference the last change made.

And they are visible in your test cube - the divots are at the sharp corners so they are harder to see

 

[FancyGunz]

What was that done on?
Nozzle? Layer Height?

Same printer as you, with a 0.6mm nozzle and 0.45 layer height at 60mm/s. I'm limited by my nozzle, I have a CHT clone on its way so hopefully I can dramatically increase my speeds. Anything faster right now and my flow drops off a cliff.

The advantage to a larger nozzle right now is high output while keeping speeds low, getting really nice lines.

nope. it is no gap, it is how seams - layer ends look like with random placements. this stuff should not be looked under a microscope.
i am not saying it cannot be improved even more, but, it is pretty damn good for that speed.

I strongly disagree, that is awful quality and I would not be OK with that. Those aren't Z seam artifacts alone, the gaps are a result of a lot of underextrusion and possibly too high coasting settings or incorrect pressure advance. You are pushing your machine way beyond its current capabilities with those unnecessarily fast speeds and acceleration values, and your print quality is suffering as a result. I recommend using a tool like this to determine your optimal volumetric flow rates, and stick to that:


View: https://m.youtube.com/watch?v=lBi0-NotcP0&feature=youtu.be

Hotend Flow Test Gcode Generator

Generate Gcode to test your 3D printer's maximum hotend flow rate at different temperatures.

hotend-flow-tester.netlify.app

 

[paul73]

I strongly disagree, that is awful quality and I would not be OK with that. Those aren't Z seam artifacts alone, the gaps are a result of a lot of underextrusion and possibly too high coasting settings or incorrect pressure advance. You are pushing your machine way beyond its current capabilities with those unnecessarily fast speeds and acceleration values, and your print quality is suffering as a result. I recommend using a tool like this to determine your optimal volumetric flow rates, and stick to that:


View: https://m.youtube.com/watch?v=lBi0-NotcP0&feature=youtu.be

Hotend Flow Test Gcode Generator

Generate Gcode to test your 3D printer's maximum hotend flow rate at different temperatures.

hotend-flow-tester.netlify.app

this is a klipper induced issue and your baseless assumptions are useless.
if you have any factual idea to offer - offer it, otherwise, avoid statements that are not based on any facts.

here is the sample of same z-seam gap klipper does, it is printed from standard 0.2 cura profile at 25 mm/s with everything turned off in it. pressure advance settings were already tried out and are not changing this. nothing of that was happening with old profile on marlin 2.1.1.

pushing machine my ass. of course i am pushing it, it is the only reason i am doing all that with this klipper install, otherwise i would not be even considering using it.

i see some similar complaints on reddit from people moving to klipper, so, it has to be something rather generic, but, cannot see a solution yet.

 

[pastera]

Same printer as you, with a 0.6mm nozzle and 0.45 layer height at 60mm/s. I'm limited by my nozzle, I have a CHT clone on its way so hopefully I can dramatically increase my speeds. Anything faster right now and my flow drops off a cliff.

The advantage to a larger nozzle right now is high output while keeping speeds low, getting really nice lines.


I strongly disagree, that is awful quality and I would not be OK with that. Those aren't Z seam artifacts alone, the gaps are a result of a lot of underextrusion and possibly too high coasting settings or incorrect pressure advance. You are pushing your machine way beyond its current capabilities with those unnecessarily fast speeds and acceleration values, and your print quality is suffering as a result. I recommend using a tool like this to determine your optimal volumetric flow rates, and stick to that:


View: https://m.youtube.com/watch?v=lBi0-NotcP0&feature=youtu.be

Hotend Flow Test Gcode Generator

Generate Gcode to test your 3D printer's maximum hotend flow rate at different temperatures.

hotend-flow-tester.netlify.app

I'm using a Tronxy X5SA Pro with a Creality Sprite extruder not my Ender 3.

Concur with all of your assessments

 

[paul73]

I'm using a Tronxy X5SA Pro with a Creality Sprite extruder not my Ender 3.

Concur with all of your assessments

And here is same sample from working 150mm/s profile. What are the quality issues there?

 

[pastera]

this is a klipper induced issue and your baseless assumptions are useless.
if you have any factual idea to offer - offer it, otherwise, avoid statements that are not based on any facts.

here is the sample of same z-seam gap klipper does, it is printed from standard 0.2 cura profile at 25 mm/s with everything turned off in it. pressure advance settings were already tried out and are not changing this. nothing of that was happening with old profile on marlin 2.1.1.

pushing machine my ass. of course i am pushing it, it is the only reason i am doing all that with this klipper install, otherwise i would not be even considering using it.

i see some similar complaints on reddit from people moving to klipper, so, it has to be something rather generic, but, cannot see a solution yet.

View attachment 707692

look at your print in this post - 3D printing
That print has divots and blobs that look like a similar z-seam issue

 

[paul73]

look at your print in this post - 3D printing
That print has divots and blobs that look like a similar z-seam issue

That was whole different issue fixed by an error in the profile with hardcoded into it 6mm long retraction setting. It was fixed and everything was fine after that.

i am pretty sure extrusion is set correctly too in the klipper cfg, as it was measured, but, i can play with that. do not think it is the extrusion amount.

 

[paul73]

well, i need to get back to work and stop dealing with this, and it all just makes me extremely annoyed.
retraction on/off and values are not it. pressure advance is not it. extrusion is not it, at least seemingly, may look more into it.
what seemed to make some difference was altering z seam alignment values in cura - changing it to 'shortest' produced a significant reduction of the gap.
in the photo the shortest is in the middle, the user specified on top, sharpest corner is lowest. looks quite different from what i am used to see in marlin, it was not making such an effect there.

did not mean to sound harsh here, but i am extremely annoyed with all this. may be i should change to some other type of the slicer program, not sure. it is some pure BS that happens.



if this image also shows some signs of underextrusion, please be so kind to educate me of where exactly is it in there.

 

[paul73]

Ok, seems like a some level of an improvement. 8:57 print, same .24mm layers and 0 infill.
Seams are not really visible anymore unless hunted for specifically.

 

[pastera]

Did a bunch of prints
Issue is high speeds/accelerations, non-optimal pressure advance and retraction.

Result is that you are going to have an anomaly at layer transitions that gets worse with higher speeds - you can tune to reduce it for a given speed but at high speed it's going to be there.

 

[paul73]

Did a bunch of prints
Issue is high speeds/accelerations, non-optimal pressure advance and retraction.

Result is that you are going to have an anomaly at layer transitions that gets worse with higher speeds - you can tune to reduce it for a given speed but at high speed it's going to be there.

beats me. mentally i agree with you, ONLY - i have done it all turning off all of the above in all combinations. i am pedantic with equipment...
i may still be missing something, of course, but, alas. for now i will have to give up. dunno what to do anymore.

the PA param i use now is minimal as is - it was measured as 0.032, and i tried all values from 0.09 to 0.005 - effect is visible, but not for the z-seam issue. the best print quality is still at the measured PA.
there also a 'smooth time' thing in there - not quite clear what it does, as now it sits at 0.04 and reducing it to 0.01 produced no difference at all.

accelerations change also did not produce gap sizing difference. in the samples i did i enforced variable speed for first 20 layers also - zero difference. don't know, it won, for now.

the best and only observed effect for now was to set 'z seam alignment' to 'shortest' and corner preference to 'smart hiding'. it makes issue less visible, but, does not remove it completely.
still, it did close the gap for 50% if not more. it is still there, clearly visible, at ANY speed.

 

[paul73]

Just to keep ball busting going - latest state of affairs- if you can’t fix it - hide it.

It’s not gone but you almost can’t see it now.

 

[FancyGunz]

this is a klipper induced issue and your baseless assumptions are useless.
if you have any factual idea to offer - offer it, otherwise, avoid statements that are not based on any facts.

here is the sample of same z-seam gap klipper does, it is printed from standard 0.2 cura profile at 25 mm/s with everything turned off in it. pressure advance settings were already tried out and are not changing this. nothing of that was happening with old profile on marlin 2.1.1.

pushing machine my ass. of course i am pushing it, it is the only reason i am doing all that with this klipper install, otherwise i would not be even considering using it.

i see some similar complaints on reddit from people moving to klipper, so, it has to be something rather generic, but, cannot see a solution yet.

View attachment 707700


if this image also shows some signs of underextrusion, please be so kind to educate me of where exactly is it in there.

I'm not sure why you're yelling at me. I apologize if I was too "direct" or whatever, but I stand by what I said. You posted a cube with consistent underextrusion (aka not extruding enough), it looks like at the start of each layer. Maybe too much retraction, do you have "retract at new layer" turned on? Maybe it's oozing and emptying the nozzle so it skips when it starts? Reprint at the same speed with pressure advance and input shaper turned off, even though you said you checked it. Either way you have large gaps which are not normal. In the cylinder above, the bottom / intiial layer (initial layer) looks underextruded, see how the lines aren't even touching. Previously you posted another cylinder with lots of holes and gaps. These look to me like underexrusion.

 

[paul73]

I'm not sure why you're yelling at me. I apologize if I was too "direct" or whatever, but I stand by what I said. You posted a cube with consistent underextrusion (aka not extruding enough), it looks like at the start of each layer. Maybe too much retraction, do you have "retract at new layer" turned on? Reprint at the same speed with pressure advance and input shaper turned off, even though you said you checked it. Either way you have large gaps which are not normal. In the cylinder above, the bottom / intiial layer looks underextruded, see how the lines aren't even touching. Previously you posted another cylinder with lots of holes and gaps. These look to me like underexrusion.

Man, not yelling at you.
I hear you. Trust me. You said all the right things.

It is I tried them all and nothing works. Just some stupid bs. Pastera btw is getting same on his test prints. he thinks it is acceleration driven, i tried reducing it and got it all the same, so, done.
Bs is as bs does.

I gonna pause on it now as I am out of ideas, will wait it over a bit

From last print - surface is good now at least. seams bullshit is now tucked into the armpits where it is almost invisible.
f#ck it.
i never had marlin do that to me, really - the only time was the whole layer clog when the retraction was set to crazy amount and filament was literally stuck cold. but nothing like that.
it seems like a deliberate accurate stop of the layer. no blobs, no errors. just a gap.

and gap size/length DOES NOT change with any speed or acceleration or PA setting change for what i tried, again, who knows if it was tried in the right combinations, but, i did try it.

 

[paul73]

i know what pisses me off.

as an engineer i do expect a logical predictable 'action-reaction' sequencing when i deal with any functionalities.
but when you deal with an amateurish bullshit what all this 'cura' and 'klipper' and 'marlin' really is, the way it is written, documented, maintained and supported by the horde of amateurs on github - you just get a monster that is not rationally responding with an anticipated reaction to your inputs, so you have to descent into an endless combination game, to find of what specific set of input parameters, not even related to the issue may or may not produce the anticipated outcome.

rant is over. it prints good enough for me now, so, i will keep it like this, unless someone somewhere will find what the f#ck causes klipper to do those z-seam gaps.

 

[paul73]

I'm not sure why you're yelling at me. I apologize if I was too "direct" or whatever, but I stand by what I said. You posted a cube with consistent underextrusion (aka not extruding enough), it looks like at the start of each layer. Maybe too much retraction, do you have "retract at new layer" turned on? Maybe it's oozing and emptying the nozzle so it skips when it starts? Reprint at the same speed with pressure advance and input shaper turned off, even though you said you checked it. Either way you have large gaps which are not normal. In the cylinder above, the bottom / intiial layer (initial layer) looks underextruded, see how the lines aren't even touching. Previously you posted another cylinder with lots of holes and gaps. These look to me like underexrusion.

as of underextrusion - i double checked the settings, this one is set to the value producing of extrudering 101mm of the material upon the 100mm request. i don`t want to do more than that. pic is same as of cube before, just other side. Zseams are in upper right corner.

Klipper says at printing that side it barely does 8 mm3/s. Cannot be exceeding anything there.

 

[FancyGunz]

as of underextrusion - i double checked the settings, this one is set to the value producing of extrudering 101mm of the material upon the 100mm request. i don`t want to do more than that. pic is same as of cube before, just other side. Zseams are in upper right corner.

Klipper says at printing that side it barely does 8 mm3/s. Cannot be exceeding anything there.

View attachment 707744

That one looks pretty good, though!

Have you tried disabling combing? I have to agree that the pictures with the gap lined up nicely pretty much points directly to a software issue.

Do you have something weird like Arc Welder plugin enabled in Cura, which I don't believe is supported in Klipper without modifying the config file?

 

[paul73]

That one looks pretty good, though!
Have you tried disabling combing? I have to agree that the pictures with the gap lined up nicely pretty much points directly to a software issue.
Do you have something weird like Arc Welder plugin enabled in Cura, which I don't believe is supported in Klipper without modifying the config file?

i let this shit to get the best of me, just need to be calmer. everything that was supposed to be disabled was disabled. nah, nothing weird was there, just odd outcomes.