Question Inconsistent Layers

[Vorkosigan]

Printer Model

V0.1

Extruder Type

Other

Cooling Type

Other

My V0 prints great for the most part, however, I've noticed some inconsistencies in my prints that I haven't been able to fix and I could use some input. I have seen other people ask about seemingly identical issues, but never did see any follow up to say they fixed it.

I'm including a picture and I've circled in red the problem areas. On the cylindrical part, there are printed threads on the inside that seem to line up with the inconsistent layer lines. The printed the cylinder on the left and the block on the right are just basic shapes I printed to see if the problem would still be there, but it wasn't.

https://imgur.com/a/NVk7tvc

Hoping that someone has some suggestions on what to try.

 

[NoGuru]

Well, I would start by making sure the belts are tightened and have as close tot the same tension as possible.
Next I would probably rebuild the extruder. I know no one wants to disassemble but to me if everything is not perfect in the extruder you will see the issues in the prints. Take a lot of time to make sure it goes back together and everything is snug and fits just right.

One more thing, maybe you have a clog or slight clog. Make sure the extruder is clear or change the nozzle.

 

[Vorkosigan]

I will check the belts again. I'm not happy with the process of using a sound app and trying to match the frequency. It's been very finicky for me.

I'm using the mini afterburner for my extruder. While I have seen many many people have problems with it, I haven't noticed any problems with mine. It has been apart a couple of times and I do try to pay attention as I put it together. One thing I am not too sure on is setting the tension of it. It is an LDO kit that had the short screw that I replaced with a longer one.

Doesn't seem to be a clog. If I print the same part, those inconsistencies will be in exactly the same place.

I was very tired when I made this topic, so I forgot to include some details.

I've had this problem from the beginning with this printer. It is mostly a cosmetic issue. Parts seem to all be dimensionally accurate. It does seem to have something to do with the geometry of the part being printed.

I followed Ellis' tuning guide. I can not remember exactly what slicer profile I started off using (it's been too long). I have tried to adjust speeds and temperatures with barely a noticeable difference.

I can't tell if it's a mechanical issue or a slicer issue.

I mostly print things with functional users and it's been great.

 

[NoGuru]

That's a tough one. I agree, I do not like using plucking to produce sound for something that requires tension. There is no way to be super consistent that I have found except tightening 1/4 or half turn at a time on each belt, and feel.

I know you said you have had the extruder apart but maybe reprint some of the parts. Judging by your description you have checked most areas that "should" effect it.
The reason I suggest the extruder is I printed my extruder 3 times in the last year and it was finally printing what I would call perfect.

Wait, you said if you reprint the inconsistency's are in the same place on the part? Try a different slicer, if you still see the errors, its the printer.

Lastly, what is the the Controller and Micro Controller you are using?

 

[Vorkosigan]

SKR Mini E3 V2.0 and a Raspberry Pi 3B

I'm hoping to have time tomorrow afternoon to work on this. (Work does seem to get in the way.)

I've had this 2.4 kit sitting here for months just waiting on me to print the parts, but I want them to be really nice. Then I want to reprint all the parts (for a better color match) of my 0.1 and upgrade it to a 0.2.

 

[Max Pickle]

Hi Vork,

If we make some assumptions based on the clean prints

1. Your rails are running smooth - you cannot feel any grittiness on them.
2. You have dimensional accuracy and are not experiencing layer shifts so the belts check out.
3. You have tried this with other (dry) filaments and the results are the same.

I'm a big fan of the Ellis guide and the first later/extrusion multiplier tests are winners for me but your prints would check out for me in those clean print tests.

So we are left with the more complex shapes causing issues. I think you said geometry seemed influence results.

Have you done pressure advance and resonance compensation? The resonance compensation graphs can sometimes point out unwanted oscillations which you can then track down. The smooth path of the simple shapes doesn't cause a problem but the more complex movements associated with the thread and V. cube could indicate something there?

 

[FkYeahScience]

...harsh lighting is reality.

Setting Expectations

Ellis’ Print Tuning Guide

ellis3dp.com

Filament-Width-Comp-Experiments/README.md at main · Deutherius/Filament-Width-Comp-Experiments

A couple of results from filament width compensation using a single mechanical/hall-effect based sensor - Filament-Width-Comp-Experiments/README.md at main · Deutherius/Filament-Width-Comp-Experiments

github.com

I assume that most of my cosmetic issues (V0.1) are related to filament quality (diameter) and not at all to mechanical issues, moisture or slicer settings.
Right now I want to experiment with filament width measurement & compensation. Still looking for two analog input pins on my SKR Mini e3 v3, all of those I could verify are already in use...

GitHub - drspangle/infidel-sensor: Firmware, mechanical and electrical design files for the Inline Filament Diameter Estimator, Lowcost (InFiDEL).

Firmware, mechanical and electrical design files for the Inline Filament Diameter Estimator, Lowcost (InFiDEL). - GitHub - drspangle/infidel-sensor: Firmware, mechanical and electrical design files...

github.com

Filament Width DIameter Sensor hall based V2 by test3210

Principle of operation : two ss49e hall sensors work in differential mode, temperature drift same for all sensor. Temperature compensation not needed. Frame on the sensor for the dust filter Used parts two ss49e hall sensors https://aliexpress.ru/wholesale?SearchText=ss49e one 5x2mm magnet...

www.thingiverse.com

 

[FkYeahScience]

That's a tough one. I agree, I do not like using plucking to produce sound for something that requires tension. There is no way to be super consistent that I have found except tightening 1/4 or half turn at a time on each belt, and feel.

I know you said you have had the extruder apart but maybe reprint some of the parts. Judging by your description you have checked most areas that "should" effect it.
The reason I suggest the extruder is I printed my extruder 3 times in the last year and it was finally printing what I would call perfect.

Wait, you said if you reprint the inconsistency's are in the same place on the part? Try a different slicer, if you still see the errors, its the printer.

Lastly, what is the the Controller and Micro Controller you are using?

Since there is no reference "sweetspot" frequency provided for the v0.1 (...or to be found by me...?) I adjusted the tension with a spring scale



One side of the scale is fixed to an aluminum extrusion, where the door would be, the other one hooked to the belt(s) for measuring (more equalizing)

 

[NoGuru]

Since there is no reference "sweetspot" frequency provided for the v0.1 (...or to be found by me...?) I adjusted the tension with a spring scale
One side of the scale is fixed to an aluminum extrusion, where the door would be, the other one hooked to the belt(s) for measuring (more equalizing)

Can you show a pic of this? I am super interested.

 

[FkYeahScience]

You asked for it, so here's some ghetto engineering;
inb4: I am well aware that this is less precise than measuring the frequency and I am in no way under the illusion that this is remotely scientific.
I still feel it is a better solution for my printer then frequency measuring since I had way better results quality-wise / determining the relative (!) belt tension (between belts A and B)

I cut the scale to fit to in betweet the x axis extrusion as well as the spring mount inside.
The pretension of the spring / force is at 0N to start with (incidentally, again).




with y=0 I put the hook onto the desired belt



...and pull the 'meter' to sit flush on the x axis extrusion. There's the required force for belt A, repeat for B. Tension adjusted if neccessary.


@NoGuru

Charme und Silikon ersetzen Präzision.

 

[NoGuru]

I am impressed. While not a perfect solution, neither is the plucking. I think this could be a better solution since it will be more consistent.

 

[FkYeahScience]

my first idea was using a defined weight on a non-elastic string/cable hanging on the belt and over the x rail and measure the distance the belt gets pulled forward, this would be easier and without much financial effort. this way an absolute value (for a certain belt type and manufacturer) could be defined

 

[Vorkosigan]

So right after I made this topic, I got sick and then life got in the way and I am only just now able to get back to this. Sorry about that.

...harsh lighting is reality.

I assume that most of my cosmetic issues (V0.1) are related to filament quality (diameter) and not at all to mechanical issues, moisture or slicer settings.

It is, but those are not the inconsistencies I am talking about. The harsh lighting inconsistencies are seen in the photo alongside the inconsistencies I am talking about.

I really like your spring scale idea for the belts.

While getting ready to do some testing based on the ideas so far, I noticed this in SuperSlicer using the Speed view of the sliced model:



Does that look to anyone else like it lines up with the inconsistencies I circled in the photo of the actual prints? My speeds are mostly directly set with just a few being set as a percentage and Auto Speed is off. I am not great at adjusting speeds. I usually just throw stuff at the wall to see what sticks.

 

[FkYeahScience]

try reducing accellerations and not speeds, I feel like this had something to do with cheap AB belts that stretch more during high accelerations.

 

[Vorkosigan]

I've been printing the parts for my 2.4 and I have not been seeing any inconsistency problems. When I view the parts in SuperSlicer using the Speed view, they are a much more solid color, no lines, compared to the attached image above.

try reducing accellerations and not speeds, I feel like this had something to do with cheap AB belts that stretch more during high accelerations.

Are the belts that come with the LDO kits cheap belts?

 

[FkYeahScience]

Are the belts that come with the LDO kits cheap belts?

I cannot comment on their particular quality, I bought some budget ones on ebay once and regretted it dearly. I assume that they are good since I never read about issues with LDO kits.

 

[NoGuru]

LDO uses Genuine Gates belts.

 

[igiannakas]

I’ve been having the exact same problem. Did you manage to fix it in the end?

 

[Vorkosigan]

I’ve been having the exact same problem. Did you manage to fix it in the end?

I did not. Where I left it was thinking that it's a slicer settings issue and my plan was to scrap my SS profile and start over from scratch. I got as far as deleting my SS profile and then just ended up using my 2.4 to print things.

Part of the reason for that it that I plan on upgrading my 0.1 to a 0.2 and that will include reprinting all the parts (previous parts were not exactly the right color and were printed on an Ender 3, the 2.4 should do a better job) and adding a few quality of life mods and then I will see about a new slicer profile and see if the issue still exists.