3D printing

[Kevin_NH]

same thing that happened to drones, pretty much, as soon as DJI monster emerged and most of the self-made models became simply irrelevant, financially. the whole concept of a self-made drone for entrepreneurship became simply obsolete.
there is still some very limited market for those parts and assemblies, but it is nothing like it was 10-15 years ago. life just took over. and now it is just a DJI.

How DJI "won" is a story in of itself.

While they dominate the ready-to-fly market, DJI products make up just a bit over 2/3 of all drones sold in the US (55% of commercial/industrial sales). The new DoD blacklist announced one month ago isn't doing them any favors.

i think this industry will close up quite fast, with major big dogs to provide their own closed platforms that feed on their own custom consumables.
an era of a wild tinkering that takes place now will go, it is inevitable. all those amateurish slicers, build-it-yourself marlin firmwares, all that will be as good as gone very soon.

Yes, there is a market for plug-n-play devices, mostly businesses and consumers with no interest in how the thing works, just want to hit "print".

Following that analogy, look at color printing, small-form-factor computing, home automation -- sure, there are mainstream closed platform products, but still plenty of tinkering ongoing.

 

[paul73]

For drones you need to source precisely made parts for it to work
A 3d printing printer is like a lathe in that it can make functional copies of itself AND if care is used those copies are less crude.
So as long as filament is available, 3d printing as a hobby will continue.

The free slicers are far from amateur and open source firmware is what allowed the industry to flourish - FDM has been around for decades. Once the patents ran out that's when the hobbyists took it to the next level and drove the industry

well, yes and no.
as i just went over whole 9yds here and all impressions are still raw - i can tell, it is extremely amateurish and resembles precisely what a drone firmware realm was at, say, year 2005 or so.
like to take me - a software engineer pro - i could not make a working build without your help there.
i can do it now, after i mingled with it for a month, but, boy oh boy, it is not a commercial grade product, not even close.

and a rig i run now - octoprint on a T95 box, with those plugins and hook to the octopod - it is a classic amateurish realm. i am not criticizing it, just qualifying. Creality could do a LOT more there, if they had DJI ambitions, but, they do not, so far, it seems.

open source rules until industry is in the rapid growth stage - but as it gets toward more professional use and commercialization, other factors start to play more - support, reliability and, well, simplicity above all.

frankly no 3d printer should require more than a direct link to that fusion 360, to think of it. all those slicers, monitors, etc - it is all supposed to be a black box provided by the product itself.
i would not be able to afford such a product for home use, of course, but i cannot afford now my own home CNC machine neither. and i bet it is going same way with 3d printers where CNC went.

 

[pastera]

well, yes and no.
as i just went over whole 9yds here and all impressions are still raw - i can tell, it is extremely amateurish and resembles precisely what a drone firmware realm was at, say, year 2005 or so.
like to take me - a software engineer pro - i could not make a working build without your help there.
i can do it now, after i mingled with it for a month, but, boy oh boy, it is not a commercial grade product, not even close.

and a rig i run now - octoprint on a T95 box, with those plugins and hook to the octopod - it is a classic amateurish realm. i am not criticizing it, just qualifying. Creality could do a LOT more there, if they had DJI ambitions, but, they do not, so far, it seems.

open source rules until industry is in the rapid growth stage - but as it gets toward more professional use and commercialization, other factors start to play more - support, reliability and, well, simplicity above all.

frankly no 3d printer should require more than a direct link to that fusion 360, to think of it. all those slicers, monitors, etc - it is all supposed to be a black box provided by the product itself.
i would not be able to afford such a product for home use, of course, but i cannot afford now my own home CNC machine neither. and i bet it is going same way with 3d printers where CNC went.

I deal with this crap all the time
Once you get past the marketing BS of a lot of commercial offerings it's not much better than open source crowd supported solutions.
Very few CNC solutions are turn key.
You buy CAD, CAM and a a machine and try to get them working with your process. On the surface the basics work well enough but once you try to get fancy the gremlins get discovered.

Back to your point, fusion can link to your printer and do the slicing internally.
I've just never set it up because I don't feel it would offer anything extra for the effort

 

[paul73]

Back to your point, fusion can link to your printer and do the slicing internally.
I've just never set it up because I don't feel it would offer anything extra for the effort

me neither, i get that. i was saying it more from the commercial workflow perspective - as if you are to run a production, that slicing process in the middle is really an entity a manufacturer of the printer is supposed to provide, to guarantee the outcome according to provided quality requirements, i would think.

 

[pastera]

me neither, i get that. i was saying it more from the commercial workflow perspective - as if you are to run a production, that slicing process in the middle is really an entity a manufacturer of the printer is supposed to provide, to guarantee the outcome according to provided quality requirements, i would think.

Many of the commercial FDM printers come with their own slicer tuned to their printer and single source supplies.
Designs still need slicing though.

If you tune your hobby printer a d stick with a single source of decent filament the process becomes very much push a button and go.

 

[Kevin_NH]

well, yes and no.
as i just went over whole 9yds here and all impressions are still raw - i can tell, it is extremely amateurish and resembles precisely what a drone firmware realm was at, say, year 2005 or so.
like to take me - a software engineer pro - i could not make a working build without your help there.
i can do it now, after i mingled with it for a month, but, boy oh boy, it is not a commercial grade product, not even close.

and a rig i run now - octoprint on a T95 box, with those plugins and hook to the octopod - it is a classic amateurish realm. i am not criticizing it, just qualifying. Creality could do a LOT more there, if they had DJI ambitions, but, they do not, so far, it seems.

open source rules until industry is in the rapid growth stage - but as it gets toward more professional use and commercialization, other factors start to play more - support, reliability and, well, simplicity above all.

frankly no 3d printer should require more than a direct link to that fusion 360, to think of it. all those slicers, monitors, etc - it is all supposed to be a black box provided by the product itself.
i would not be able to afford such a product for home use, of course, but i cannot afford now my own home CNC machine neither. and i bet it is going same way with 3d printers where CNC went.

I agree -- Creality is not ready for prime-time, but you only need to go that route if you are cheap, really into the hobby, or a masochist (but I repeat myself).

I know several engineers and makers with home CNC routers (and not solely the Ghost Gunner). Some are willing to sell their soul to Autodesk and use the Fusion cloud toolpath generator, others go with various options which run locally, closed source or open.

me neither, i get that. i was saying it more from the commercial workflow perspective - as if you are to run a production, that slicing process in the middle is really an entity a manufacturer of the printer is supposed to provide, to guarantee the outcome according to provided quality requirements, i would think.

And on the more expensive printers, they do -- if you focus on the turnkey "prosumer" models they all come with the functionality mentioned above --ready built firmware and slicers, cloud-slicing, etc. They mostly also come with a price tag well north of $1K.

There are also mid-priced (+$500) models where they ship with all the silly fiddly stuff done for you, but you are still able to do any upgrades and reflashing you might choose, no vendor lock-in. That's the direction many of the FOSSCAD gun folk generally go -- they want local slicing and local control without anything phoning home, for obvious reasons.

 

[paul73]

Creality is not ready for prime-time, but you only need to go that route if you are cheap, really into the hobby, or a masochist

the path of a cheap masochist is the only true path - the path of the true samurai!


i liked an idea of creality for what it was before i got it - an open end open source open everything with a very ow start price point, as, frankly, i did not expect ender 3v2 to be as good as it worked out to be. i did not think it will work well at all.

i can see now how their newer ender 3 S1 pro addresses most of the issues i see in my cheap v2 - but, well, i can live for now with the hybrid i got.

 

[paul73]

This print is doing it consistently to me if aligning long side along X axis. If rotated 90deg - no issues.

I checked all mechanicals, do not see anything wrong. Any ideas?
I do not understand why it keeps shifting like that. It must be a hit on Y axis to make a belt skip, but, how and why?
looking at the middle section from the top - it seems like it did a skip, then did 2 layers, then skip, then 3 layers fine, then skip again.
it makes no sense.

 

[pastera]

This print is doing it consistently to me if aligning long side along X axis. If rotated 90deg - no issues.

I checked all mechanicals, do not see anything wrong. Any ideas?
I do not understand why it keeps shifting like that. It must be a hit on Y axis to make a belt skip, but, how and why?
looking at the middle section from the top - it seems like it did a skip, then did 2 layers, then skip, then 3 layers fine, then skip again.
it makes no sense.

Check your z lead screw and nut - clean and lubricate
Make sure the guide wheels on both uprights are clean and adjusted.

Assuming you have already checked that at that z height none of the cables can catch on anything and you filament path is good

 

[paul73]

Check your z lead screw and nut - clean and lubricate
Make sure the guide wheels on both uprights are clean and adjusted.

Assuming you have already checked that at that z height none of the cables can catch on anything and you filament path is good

yeah, checked all that. some days ago spent evening playing with eccentric nut on other side from lead screw - to all loose, then all tight - no use.
it is annoying as it never did this before, and i am not sure when/why it began. i only noticed it on this specific print as i was playing with it a lot lately.
odd.

 

[pastera]

Are you running a high speed profile?
Videos show the successful prints using those profiles but they don't show that some prints find the resonances in the frame that induce failures.
Try using z hop when traveling and place the part on the left hand side of the bed so the head is closer to the supported side of the rail.

 

[scheming]

Get a cam, record print and see what happens when it goes wonky. Not sure if octoeverywhere has that feature. I use a cheap wyze cam for my setup and has solved a lot of failure enigmas.

 

[paul73]

Are you running a high speed profile?
Videos show the successful prints using those profiles but they don't show that some prints find the resonances in the frame that induce failures.
Try using z hop when traveling and place the part on the left hand side of the bed so the head is closer to the supported side of the rail.

probably is a resonance, as all support tests and other things print just fine with that profile. hmm.
i wonder if it makes any sense to get this contraption or not, it does not look like it would do much good, to be honest, and some reviews are also not too great.
difficult to tell how much factual rigidity will this thing add, other than to tilt the vertical columns backwards a bit.
realistically, it is not wobbling back and forth there at all.



Amazon product ASIN B07RDB6X5NView: https://smile.amazon.com/CHPOWER-Support-Accessories-Compatible-Ender-3/dp/B07RDB6X5N

 

[pastera]

It's a bed slinger - those supports won't really increase Y stiffness.

If you move the print nearer to the Z-axis lead screw and the problem goes away it may just be that you need a dual leadscrew setup to get things moving that fast AND add corner brackets to the upper cross brace.

Your picture doesn't give a lot of info without knowing what the model looks like
From what I can gather the part is sitting on a large amount of support and failed when the arm connected to the cylindrical part started printing.
Settings are tuned for speed given the thickness of the layer height and high temp of the extrusion.

Printing large flat supported areas needs really good cooling to keep the floor flat and attached to the support roof - if the floor disconnects while still hot it will curl and catch on the nozzle during a move which can knock the part off the build surface or cause a stepper motor to loose a step (layer shift). The faster things are moving the less torque a stepper has so fast speed/acceleration coupled with a interfering model will drop steps easily.

 

[paul73]

It's a bed slinger - those supports won't really increase Y stiffness.

If you move the print nearer to the Z-axis lead screw and the problem goes away it may just be that you need a dual leadscrew setup to get things moving that fast AND add corner brackets to the upper cross brace.

Your picture doesn't give a lot of info without knowing what the model looks like
From what I can gather the part is sitting on a large amount of support and failed when the arm connected to the cylindrical part started printing.
Settings are tuned for speed given the thickness of the layer height and high temp of the extrusion.

Printing large flat supported areas needs really good cooling to keep the floor flat and attached to the support roof - if the floor disconnects while still hot it will curl and catch on the nozzle during a move which can knock the part off the build surface or cause a stepper motor to loose a step (layer shift). The faster things are moving the less torque a stepper has so fast speed/acceleration coupled with a interfering model will drop steps easily.

It just did it again. I had video on a screen but got distracted and missed that moment.
It printed one model fine, in left corner next to a screw, but a next print in a same spot got that collision again. Go figure, I give up. All the rest prints fine, test figures print fine. Annoying as hell.

 

[pastera]

Can you flip that part?
Looks like a lot less support needed if printed with the arms down

 

[paul73]

Can you flip that part?
Looks like a lot less support needed if printed with the arms down

if flipped it leaves bad grid marks on the bottom of the flat parts, looks bad and difficult to clean off.
the whole deal is - it prints just fine when aligned along Y axis, which makes me quite annoyed, as it is difficult to understand, it should not be any collision there. .

 

[FancyGunz]

It could just be belt slippage. Double check all belts for correct tension.

 

[paul73]

It could just be belt slippage. Double check all belts for correct tension.

it is proper, they are tight. well... will see if it will get worse or anything else will start showing such fails. so far it is one only design that did it - consistently enough.

 

[paul73]

Can you flip that part?
Looks like a lot less support needed if printed with the arms down

I think this shit happens at same exact height. Just moved all cables around, tied ‘‘em up, started job again.

I flipped the part, one printed, next test got same failure. It is mechanical for sure.

 

[pastera]

Pull your belts and check that both x&y only tight enough to rove wobble without a lot of friction. Clean the wheels and extrusions to get rid of dirt causing gritty motion.
Pull the z lead nut do the same.
Check the z lead screw alignment - early Ender 3's needed a shim behind the z-motor to align the leadscrew.
Make sure the lead screw and z motor shaft aren't bottomed out in the coupler.
Completely tightening down the leadnut can lead to z binding

 

[Kevin_NH]

I think this shit happens at same exact height. Just moved all cables around, tied ‘‘em up, started job again.
I flipped the part, one printed, next test got same failure. It is mechanical for sure.

There are multiple possible causes, can be a tough nut to crack.

1. Have you tried printing from SD card instead of via octoprint?
2. The "bedslinger" printers (including your Ender 3v2) are prone to collisions on taller prints, reducing speed and tuning jerk/acceleration can help
3. Belts could be damaged -- we have a delta printer which always "jumps" at the same height, no matter what is being printed.
4. When you say "test figures print fine", I assume the test figure is taller than the failure point, but likely not a similar geometry as the print which fails?

 

[paul73]

There are multiple possible causes, can be a tough nut to crack.

1. Have you tried printing from SD card instead of via octoprint?
2. The "bedslinger" printers (including your Ender 3v2) are prone to collisions on taller prints, reducing speed and tuning jerk/acceleration can help
3. Belts could be damaged -- we have a delta printer which always "jumps" at the same height, no matter what is being printed.
4. When you say "test figures print fine", I assume the test figure is taller than the failure point, but likely not a similar geometry as the print which fails?

yeah. tests were not same geometry nor same height. ity was difficult to replicate.

i think it was either - probably - one for the cable catching the side of the enclosure - or - the camera mount, also catching the side of the enclosure.
something somewhere was catching up or getting stuck at this particular height, even though it was not apparent to have any issues.
for now i rerouted cables, re-tied them a bit differently, ran it up and down, lubricated z screw, and so far 3 times printed same crap and it came out fine.
with a filament roll at $13 i can do it all day, that model is, like, a 29g. will see.

 

[paul73]

There are multiple possible causes, can be a tough nut to crack.

1. Have you tried printing from SD card instead of via octoprint?
2. The "bedslinger" printers (including your Ender 3v2) are prone to collisions on taller prints, reducing speed and tuning jerk/acceleration can help
3. Belts could be damaged -- we have a delta printer which always "jumps" at the same height, no matter what is being printed.
4. When you say "test figures print fine", I assume the test figure is taller than the failure point, but likely not a similar geometry as the print which fails?

A rather peculiar shit - all was fine and then, again. I looked at it while it was happening- a single print head hit at previous layer, consistently.

Will get dual z kit delivered tomorrow, will see if that will do any good. Such a total BS out of nowhere. and if it is a sporadic mechanical thing - how can it be repeating layer after layer once it began?

 

[paul73]

so, this is interesting. i rotated model 180deg - and the issue flipped direction. seems like it is a slicer, to be specific - an influx of arc welder that i added in there.
the hit occurs in the transitional area between G1 and G2/G3, in the specified spots, on both sides - one feels harder than the other and causes the shift.
holy crap.
now i need to understand why. the hit was exactly there, as it would proceed to draw out the outer wall. to an eye there was no visible diff in the height, but, obviously - it was hitting something on its path. damn, damn, damn.

 

[paul73]

looking at slicer - nothing indicates a possible issue, but hit was exactly where highlighted, like the walls of G2/3 were higher than G1. WTF?
looking back the addition of arc welder aligns perfectly with the time when it started happening - i just could not associate one with another. one more lesson to write down and track any changes made to a perfectly well working setup. what a fecking POS.

 

[paul73]

so, i made a test slice of 20 layers where issue was happening - all the 5 attempts with arc welder would have a hit/thump in the transition spot.
as i saw from 5 samples - not every hit would make a layer shift, but it will eventually happen as hit gets harder, as the transition area seems to raise there from layer to layer.

printing all the same with arc welder disabled - the thump is gone. WTF, really... so sad. it was cutting print time quite well. unbelievable.
one more jab at the state of this tech, like, is it for real?

am i the only one who had issues with arc welder? i just did 3 prints with it disabled - the hit sound was gone, prints are perfect. so odd.

will try printing full size one again with it disabled - will see what will be happening on a layer 112.

 

[paul73]

and no shit, look at it. dated by some jan 2021. well, at least it is not a product of my imagination. holy grief.
f#ck it, the best fix for it is not to use it.

Arc Welder causing Layer Shift · Issue #113 · FormerLurker/ArcWelderPlugin

Hi @FormerLurker, Apologies if you have already fixed this as I know you have done fair bit of work since the first release. I wanted to report it in case you have not. I noticed layer shifts on a ...

github.com

 

[scheming]

Curious what was the time savings with arc welder enabled?
Seems like the linked thread ended with octoprint/pi declared the culprit, with G2/G3 compounding the issue. Do you use octoprint as well?

 

[paul73]

Curious what was the time savings with arc welder enabled?
Seems like the linked thread ended with octoprint/pi declared the culprit, with G2/G3 compounding the issue. Do you use octoprint as well?

Yes, I use octoprint. Time saved differs upon model shape, maximum effect is on tubular, pipes etc. visually effect is not too huge, surface is a tad smoother with g2s, but not too dramatic to care.

I think the issue was with plastic buildup at the spot of transition from g2 to g1, but, who knows. All prints fine again.