[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vm / vmg / vr / vrpg / vst / w / wg] [i / ic] [r9k / s4s / vip / qa] [cm / hm / lgbt / y] [3 / aco / adv / an / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / pw / qst / sci / soc / sp / tg / toy / trv / tv / vp / vt / wsg / wsr / x / xs] [Settings] [Search] [Mobile] [Home]
Board
Settings Mobile Home
/diy/ - Do It Yourself

Name
Options
Comment
Verification
4chan Pass users can bypass this verification. [Learn More] [Login]
File
  • Please read the Rules and FAQ before posting.

08/21/20New boards added: /vrpg/, /vmg/, /vst/ and /vm/
05/04/17New trial board added: /bant/ - International/Random
10/04/16New board for 4chan Pass users: /vip/ - Very Important Posts
[Hide] [Show All]


[Advertise on 4chan]


File: 20240430100006.jpg (3.78 MB, 4000x4000)
3.78 MB
3.78 MB JPG
3D Printed Breakfast Edition

Last Thread: >>2787167

>Your print failed? Go to:
https://www.simplify3d.com/support/print-quality-troubleshooting/

>Calibrate your printer.
https://teachingtechyt.github.io/calibration.html

If that doesn't help you solve your problems, post:
>A picture of the failed part
>Printer make & model
>Filament type/brand
>Slicer & slicer settings

>What printer should I buy? [32/03/90 :detadpU tsaL]
Do your own research, these are just popular and available options.
All controversial printers and brands have been removed from the list for your safety.
DIY: https://reprap.org/wiki/
SLA: >>>/tg/3dpg

>Where can I get things to print?
https://www.thingiverse.com/
https://thangs.com/
https://printables.com/
https://grabcad.com/
https://www.yeggi.com/
https://cults3d.com/
https://www.stlfinder.com/
https://google.com/
T*legr*m

>What CAD software should I use?
Free to anyone: Fusion360, Onshape, TinkerCAD, FreeCAD
Free to me: Autodesk Inventor, AutoCAD, Solidworks, Rhino, Solid Edge
Autistic /g/oobers: OpenSCAD, OpenJSCAD
Mesh free-forming and modeling: Blender
Architects: Sketchup

>What slicer should I use?
For everyone: Cura, PrusaSlicer, BambuStudio for Bambu owners.
For enthusiasts: SuperSlicer, OrcaSlicer
For autists: Pleccer/SuperPleccer, Kiri:Moto, FullControl

Legacy Pastebin (Last updated 1239 days ago): https://pastebin.com/AKqpcyN5 (embed) (embed)
#345
>>
first for chinese data providers
>>
Anyone played with CF-PETG? I'm considering using it for a wearable I'm working on but not sure how much better it is vs regular PETG when it comes to overhangs and support release.
>>
>>2793796
thanks for baking
I made phroggé
>>
I have six million dollars.
What 3d printer do I buy?
>>
>>2793870
ender 3 with 5.99 million worth of mods
>>
>>2793870
fill your home with Bambulab X1C's if you hate having to do any setup and don't mind the spyware
>>
>>2793869
stl?
>>
>>2793893
May as well be asking where to find a Benchy stl.
(2012) https://www.thingiverse.com/thing:18479
https://www.printables.com/model/3116-treefrog
Don't know if the Prusa repost has any changes, might be cleaner, fuck if I know.
>>
>>2793888
okay i did this
my prints are warping what do
>>
>>2793946
sell it
>>
Is there any torrent site for .stl files? Fuck paying for each of these prints
>>
CREALITY K1C Enclosed High Speed Carbon 3D Printer 300°C hotend combined with a hardened steel nozzle tip, 4.3 Color LCD Screen; Automatic Leveling, PEI Flexible Build Plate; 220 x 220 x 250mm Print Size, Built-in AI camera, etc.

STILL on sale at MICRO CENTER (usa) for $489.99
>>
File: PXL_20240307_011052904~2.jpg (1.84 MB, 3072x4080)
1.84 MB
1.84 MB JPG
I've run into an issue with my SOVOL SV06, when I queue up a print, the printer head gets so hot that it melts the plastic to the point it is too soft to extrude. I initially thought it was a heat creep problem and got to the point I even replaced the whole extruder and heater block, but the brand new one has the exact same problem. It's to the point that when it heats up even the circled thumb point used for helping manually extrude is scalding to the touch.

Have any of you encountered anything like this before?
>>
why is everyone printing those stupid dragons?
>>
>>2794015
Normies will pay $30+ for them
>>
>>2794034
why tho? they are stupid and useless and destined for the trash
>>
>>2794036
The thing is that 3d printing is still very much magic to 99% of the population. The gift shop at the Colosseum is selling meh-quality busts for €150. Just go to etsy and type in "articulated 3d print" and you'll see people with THOUSANDS of sales for the most lame stuff.
>>
>>2794007
hotends that have the heatblock close to the motor and frame that holds them together tend to get hot, i had the same problem on an IR3 hotend

Check that your extruder vref isnt cooking it.
What cooling does the hotend have? I dont see any heatsinks on the motor or any part of the frame that isnt just a solid block, no fans to cool the general mass of the hotend?
The other fans; they are running right? Fans get pretty damn hot if they are wired backwards or told to run at a speed to slow to actually turn the fins.
Probably also worth doing a pid calibrate and checking the heater and thermistor are properly secured and not damaged.
Same for any wires that might be shorting to the hotend somewhere.
And if its inside a heated enclosure... well theres that too.
>>
>>2794015
you can't easily injection that kind of articulated part. it's cool and fun
>>
Is charging 3 cents per gram for 3d printing to get started a good idea?
>>
>>2794085
No, not remotely, that's a fucking terrible starting point, are you trolling?
>>
My K1 suddenly won't extrude anymore. I cleaned everything inside, but it would start printing fine for the first couple of minutes, then it would just stop. New extruder maybe?
>>
>>2793990
here:
UGF5IGZvciBvdGhlciBwZW9wbGUgd29yayB5b3UgcG9vciB0d2F0
>>
>>2793870
You can probably get an electron laser sintering setup for that kind of cash. Makes much stronger metal parts compared to laser sintering.

>>2794085
Depends on how easily you can print things, but probably not. Even if you had a print farm that automatically extracted prints off the build plate to start new prints, there's the process of slicing them, cleaning the bed every now and then, handling failed prints, taking the completed prints and packaging them, that's at least a few dollars worth of a flat fee.
I'd probably put the filament price no less than 6c/g, maybe over 10c/g, have a flat ~$2 per part, and additionally a flat ~$5 per build plate if you're not mixing different customer's prints together. So three 100g prints on one build plate would make you $41 gross. A single dinky 10g print would make you $8.

If you have to manually walk over to your printer with a USB stick, have to manually remove prints and/or can't queue multiple seperate prints, or have to do any design work yourself, those prices would go up even further. And FYI you shouldn't print pre-sliced g-code, don't trust customers not to give you shitty g-code that could damage your printer.

>>2794108
Check the extruder open/close lever, I hear it can get stuck not properly closed on the filament.
>>
Am I just asking for trouble by setting my Z seam to randomize? I tried to print a bowl like object and it's come out like a pimply stringy mess, and I'm not sure if it'd had been cleaner if the Z seam being aligned would have made it move and retract less around the bed.
>>
how can i make my PETG prints more scratch resistant ?
>>
>>2794191
try orca scarf seam
>>
>>2794198
I'll be sanding and painting the final product, I'm just trying to find the best way to make it easier to finish since the current prints I've got look like total ass.
>>
File: 1711497366390.jpg (6 KB, 325x309)
6 KB
6 KB JPG
>>2793870
>six million
Just get an X1C with AMS and spend the rest of your money on filament and delousing agents.
>>
>>2793675
Honestly, no. Qidi machines always like the most flimsy plastic laying around thrown at a 3d printer. I have hard believing they're actually good.
>>
>>2794329
What Qidi machines have you owned?
>>
File: deepenededdiscount.jpg (31 KB, 663x731)
31 KB
31 KB JPG
Wew. Got my 2nd AMS for a deep discount.
Had uploaded a model that got 2500 downloads and another model that got 600. Took 6 months but amassed $320 worth of giftcards but I think the steam's running out on those models so went ahead and took the plunge especially after the points change.
I don't print very often so getting a second printer wouldn't do me much good.
But when I do print, I'd like to have 8 filaments on deck.
>>
How should I go about selling my ingenious 3D printable designs for money?
Any websites you'd recommend?

I'm a NEET so I'd like to make some money
>>
Absolutely LAYING it
>>
Why is Cura putting retractions in my spiralized shoe shell?
>>
>>2794451
damn that boy thick
>>
>>2794388
What have you done?
>>
File: F7bX5fpagAATYIQ.jpg (699 KB, 2229x2972)
699 KB
699 KB JPG
Are there any printing services out there where I can send in my custom model and ask for a print? I'm specifically looking for one with resin material.
I'm also concerned if the manufacturer is reliable and will handle it safely because I know resin is a material not to fuck around with.
>>
>>2794192
Spray isopropyl alcohol on it and apply an ignition source Then extinguish after 1 seconds. Walla! PETG is now plastic tempered
>>
>>2794554
>Walla
It's 'Voila!' you amerisavage
>>
>>2794542
JLCPCB and PCBway do, so do a bunch of other more local print farms that will charge more.

>>2794559
i assume they say it that way ironically, it's better for my mental health
>>
File: 1714651411271.jpg (153 KB, 857x1143)
153 KB
153 KB JPG
mornin 'ghetti
>>
File: 2024-05-02--22-21-09.png (697 KB, 1920x984)
697 KB
697 KB PNG
>been meaning to do this for a long time
>wrangling of complex bools
>everything went nicely
>almost too easy
>printer GO!
>sleepy bo-bos time

fingers crossed its still sexy in the morning
>>
File: 1711392367603669.webm (41 KB, 320x318)
41 KB
41 KB WEBM
>>2794594
>JLCPCB and PCBway
Thanks for the recommendation friend, these look pretty reliable.
>>
File: 20240501_182202.jpg (1.63 MB, 1440x1876)
1.63 MB
1.63 MB JPG
>>2794451
Correction needed
>>
File: 1702818725341829.png (2.47 MB, 1920x1080)
2.47 MB
2.47 MB PNG
>affordable Voron with injection molded parts is still a meh dream
Expected more, but maybe shooting for 350 instead of just 300 was a bit much on their end.

>>2794684
Looks snuck, hope it does fit as well.
>>
what glue should i use to semi-securely stick pei coated steel to glass at temperatures of 80+ degrees celsius
let's say it's secured on corners but bows out in the middle
it's going to be a semi-permanent fixture but the option to remove and replace it in case of unforeseen damage has to be there
asking for a friend
managed to stick them together with hair spray for a couple prints but then it violently ejected a 3-hour print right near the end
>magnets
tried, they lose power when heated over 60 degrees
>buy a new
no
>go back to glass
eeeeeeeeeh
>>
File: 1710136675007693.jpg (55 KB, 984x984)
55 KB
55 KB JPG
>>2794706
>>
>>2794725
the plate is already secured with butterfly clips on the periphery, the problem is that the middle of the plate bows up. it doesn't take much force to flatten it, but it's also a very awkward place to apply force to. right now i can put a small steel weight and a magnet in the middle and print in the corners, and as i said, hair gel held it down for a couple prints, but i need something that will hold it properly and more securely.
>>
>>2794706
>>2794733
I've never had to do what you're describing but my first tries would be something like RTV silicone or I'd look for a semi-permanent contact cement.
>>
>>2794706
>>2794733
Put on a 468mp sheet and call it a day?
>>
>>2794705
Its a really good printer if you already know what you are doing and have used klipper before. Like since its using an inductive probe you have to heat soak first. And theres no way it can print at 40k accel with quality on a 350 bed when bambu can only do 20k on a 250 bed. Not a fault of the printer, just a limitation of corexy designs.
>>
>>2793796
Do any filament fusers actually work?
I keep trying but I've never had any success.
>>
File: 2024-05-03--12-38-02.png (215 KB, 447x558)
215 KB
215 KB PNG
instagram keeps showing me interesting 3dp shizz, so i guess i'll post some here
i cant always save them, and I guess most would be too big to post here anyway, so links will have to do
https://www.instagram.com/reel/C4ihfTANd0U
>>
>>2794923
the ones ive seen probably do, but from my own experience with tests i did years ago; its always a bit fiddly and hardly seems worth it when PAUSE_PRINT is an option.
>>
>>2794945
also: i think this would work reasonably well too:
- make this: https://www.thingiverse.com/thing:3338467
but give it a few winds to make it *push* filament.
-then put a Y junction in the filament path, with primary filament going to extruder as usual, and secondary *pushy* filament butting into it.
whewn the end of the primary gets pulled past the Y junction the secondary should start getting pushed to the extruder. Assuming it catches on the extruder gears you should have seamless re-feed.
>>
File: 2024-05-03--13-12-38.png (101 KB, 443x553)
101 KB
101 KB PNG
>>2794943
https://www.instagram.com/reel/C4ydT0FPTnO
>>
File: 2024-05-03--13-22-51.png (205 KB, 453x558)
205 KB
205 KB PNG
"fluffy bag"
https://www.instagram.com/reel/C5ECIGFNodz/
>>
I was designing a sheet metal + 3D printed parts ITX PC case for hours but then i had to make a change to the first part which introduced a few new faces. This part was the basis for all the other parts (it's part binder was used to design everything else) so the entire project got borked and i will have to spend a lot of time fixing it, pretty much having to do most work from scratch
Should I shift to openscad? I am a programmer so i don't think it'll be too much of a jump
>>
>>2794988
Forgot to add, i am using FreeCAD now
>>
>>2794988
OpenSCAD is kinda limiting. You can't do complex transformations of composite shapes like adding chamfers, and there's nothing in there that even knows what sheet metal is. You don't define a hole as being 8mm from an edge, you define it as being at a certain coordinate, if you mess around with variables you can get stuff act as if it's parametric, but it has to be manually set up that way. It's bare bones and not at all optimised for practical design. There are more powerful versions of software like OpenSCAD, but I don't think they'd be much better.

FreeCAD should have a proper parametric design with forward propagation of changes, but it being a buggy mess does tend to get in the way of that. Are you using vanilla FreeCAD, or are you using the RealThunder branch like you should? Ondsel might be more stable, idk.
>>
>>2794993
>Are you using vanilla FreeCAD, or are you using the RealThunder branch like you should?
I'm using vanilla. I'm scared of using realthunder's branch because I don't want to get used to any design patterns which are not standard
>>
>>2794166
>Check the extruder open/close lever, I hear it can get stuck not properly closed on the filament.
i fixed it. Was the gears in the extruder that needed a cleaning and relube
>>
>>2794947
Have you fed filament by just turning the spool? I think the coil on the spool will expand first, and usually there's nothing to confine it.
>>
>>2794993
What i think i mean to ask is openscad any better if I'm a beginner and i need to often make fundamental changes to the base part.
With FreeCAD i spend multiple hours making a base part, then making many parts which depends on it. Then when i see i missed something in the base part i make that final change and everything which depends on the base model breaks. Then i have to redo the entire project but with the change that i did in the end.
I understand that openscad doesn't work that well when external files need to be imported and used for referencing geometry but if every single mechanical feature is known (and i can basically draw the entire mechanical drawing on paper)

Sorry for the long post, I've spent a week remaking the same computer case every single day because i forgot something minor every time in the base part and had to redo it.
>>
I accidentally shattered the hot end fan on my ender 3 pro last night. I'm getting really sick of this printer.
1. Is a direct drive upgrade worth it? Which one?
2. Are there any compatible hot ends where the nozzle is also a bed leveling probe?
>>
Sold my ender, looking at a sidewinder 4 plus. Does anyone know if it goes on sale below 350 frequently?
>>
>>2795072
When you make something in your base, are you using parameters? Like if you're making a tube that has a cylinder go in it, are you using a parameter like `innerDiameter` in all places?
>>
>>2795097
Unfortunately I'm not, but i probably could
Would that make freecad viable? I'm not sure how it would counter problems related to some part very deep in the dependency tree needs a new face or feature
>>
>>2795104
It certainly makes changing things a little easier, but I guess it depends what you mean by things breaking. The kind of breaking I'm thinking of is parts suddenly not fitting or just yeeting themselves off to fucking wherever.
>>
Hi, I am aiming at Ender 3 V3 KE, and it is priced 400e, within country. Of course there are other printers which have various prices but I am unsure which one to get, as this one seems ok with few upgrades like linear rails. I am in process of making tabletop cnc milling machine, and now I am stuck on part where I need plastic adapters and would be good to have 3d printer at home. Which one would you all recommend, that is plug and play, and little bit of tweaking is also ok
>>
File: 20240503_124212.jpg (2.13 MB, 3468x4573)
2.13 MB
2.13 MB JPG
My print started out fine to my standards but near the end it looks like the layers began to shift. Did my belts just become loose at the end or what?

Brand new roll of Overture PLA white, 210 C nozzle, 60 C bed, 99% infill, tree supports, took about 26 hours to print. White filament has always given me shit.
>>
>>2795153
Since it's been a long print it might be something overheating, perhaps the drivers or the heatbreak
>>
File: 20240503_123633.jpg (1.71 MB, 2880x2160)
1.71 MB
1.71 MB JPG
>forget to clear the bed
>step away after hitting start
>come back
>???
>crashed but printing perfectly in new spot
Amazing
>>
why are the prusa mk3 and mk4 priced so similarly? is there that little difference?
>>
>>2795184
There was something like a 90% part difference between the mk3 and 4. That's why the full upgrade kit costs so much too. Just buy a P1S or X1 instead.
Only get the mk4 if you are deadset on getting a Prusa. You'll need to tune the input shaping manually. Prusa is sucking his own dick so hard he genuinely believes every single mk4 resonates at the same frequencies. There is a post on their forums about using a klipper test print and modifying the machine start gcode to override Prusa's "perfect" values.
Prints on my mk4 looked like dog shit until I did this and the frequencies were pretty far off from those "perfect" values. Checked the belts, screw tensions, alignments, nothing helped except manual IS tuning.
>>
>>2795153
this addresses the lines:
>>2794953
and on the overhang at the top: you seem to have shit bridging. Partfan exposure from different angles is always better. Maybe adjust your bridge speed. Organic supports are great for those reacharound-handy support places (so you dont need to put support on top of print).
Again, it looks like the lines interfere with the overhang threshold too.
Not sure how you would solve that, but printing in different parts and/or different orientations might be required.
Similarly: you could always neatly saw through the model at the most fubar lines... print the rest of the model from there on up.. and try grafting them together.
You might want to print reduced-size versions of large prints to find out where they are going to fail, before you print the megalodon-sized version.
>>
>>2795238
PS: It would be good if we could see the entire model. But right now oit looks like a no-brainer that you should be rotating it 90deg to the right and printing it with the bolt hoes(right) at the bottom and the pointy bit (left) at the top.
>>
>>2795153
minimum layer print time before slowdown, might be worth playing with too
maybe also adjust how aggressive your fans are on shorter duration layers... or try one with no fan
>White filament has always given me shit.
thanks for reminding me of this, i will probably still do it, but i have been considering buying 10KG of white petg. Still not sure.
>>
Whats a good way to level up in the 3d printing space? I saw earlier in the thread people were talking about those articulated dragons that anyone can print. Have even seen them at flea market type events. I have a feeling most people with 3d printers are pretty close to this level of skill. Mainly just printing goyslop stl files. Im sort of plateauing. Most of my design work is in tinkercad and I have designed some pretty complicated stuff on it so far, but I know its a bottleneck. Have made about $1200 selling quest 3 headstrap adapters I had to rush to market but would like to go beyond that.

The design process of some of my personal projects has been pretty fun and I like to see all of it come to life even if some of it ends up being a dead end. I have a bunch of tools and im trying to become more proficient with them, and 3d printing has been pretty useful for rapid prototyping, making jigs, and producing products. Mainly just trying to figure out where I start to go from here? I dont want to be just limited to 3d printing and im a welder by trade, but want to have a wider and better skillset. Just want to learn some valuable skills and put them to use the best way possible.
>>
>>2794997
Use realthunder, like any sane person
It's not like step files become incompatible
>>
Dumb question, but what color filaments are good for beginners? I mean more from a practical perspective, it doesn't take a genius to pick your favorite color.

For example, are certain colors easier to paint, have different performance qualities, easier to spot flaws, etc.
>>
>>2795283
>printing goyslop
Now I'm wondering if there are filaments made from s-o-y beans. I mean, if you can make printer ink, candles, and even furniture out of it, I don't see why you can't make filament with it.
>>
>>2795287
black and white. If you need another color then buy as needed. You dont want to end up with a bunch of random spools. Would be more concerned with mechanical qualities of the filament as well. TPU and PLA are worlds apart.
>>
>>2795287
Filament behavior will vary with color, but only very little. The most dramatic ones are usually white or black filaments, as they require large quantities of pigment, far more than other colors do.
>>
>>2795288
PLA is basically printed goyslop. Im sure there is legit based bean filament though
>>
File: quickmodel.jpg (3.47 MB, 4000x3000)
3.47 MB
3.47 MB JPG
Was dispensing liquid 1 syringe at a time and that takes time and effort. Made a real quick model in under an hour then about an hour of printing. Hopefully this little tool can save me dozens or hundreds of hours. Was a good exercise in design. Started out fancy but just dumbed it down in the end. Good lesson in design for me going forward is that I just need to make something cheap and dumb that works unless its something for consumers. Just make it as cheap and easy and efficient as possible.
>>
>>2795283
Get into better plastics, especially ABS. ABS is strong and temperature resistant enough for automotive use, and it’s cheap as. Polycarbonate is the other one I’d look at. Nylon less so, but its flexibility combined with its strength may be handy. TPU and similar stretchy filaments would also be very useful to be able to print, especially with a dual filament printing setup. Knowing how to choose the right plastic for the job is half the battle.

If you’re designing custom parts, then try to design parametric models that can be easily customised to suit a variety of customer requirements. For enclosures and brackets and such. Could do this with proper CAD tools or something like OpenSCAD.

Also get yourself a good selection of heat-set inserts and other things to add into your prints.

That Slant 3D printing channel has some good tips for professional design.
>>
>>2795331
I watch a lot of slant 3d and a few other printing channels. Havent really messed with any of the other plastics besides TPU and a bit of ABS. I am aware of some of their properties though. Interested in Nylon, Glass filled Nylon, and probably filament with glass beads or other exotic materials.

The design is where I feel im limited most. I have been able to design some organic shapes on tinkercad, but nothing advanced like being able to sculpt something like a computer mouse. I know I have to find a modeling solution besides tinkercad. Have been pushing it to near its limits. Hate feeling like im going to be starting from zero. Also im wondering if I can fill in some gaps with 3d scanning and AI in the not so distant future.

The heat set inserts is a pretty good idea. I have designed with dovetails in the past, but heat set inserts can really make a prototype look professional.

I don't want to be too entirely tied down to 3d printing either though. I have a small diode laser I have cut some parts on, and plasma cutting is just a step up from that and routing is adjacent to that. There can be money in all those things and they all basically use a very similar skillset. The current product im personally developing has me working with open face silicone molds for consumer electronics. Mainly just trying to get away from hobby level stuff and move more into actually using these skills to earn a living. The cash on the adapters was a small taste. Would like to see where I could get if I really started to widen the skill and capital gap.
>>
>>2795287
grey
It photographs well therefore it timelapses well
It prints well (unlike strongy pigmented filaments), like black and white. White also blemishes far too easily, even during the print process. Black makes it almost impossible to see any artifacts you want to remove, and too easy to see ones you cant remove. Both are still good colors in their own way though.
It isnt some garish color that you can only feel right printing some things with

Personally if i was current me advising newbie me i'd say buy 10kg of flat grey, and 10kg of miscellaneous colors. Best value for money, not a huge investment.
Or 5kg grey and 5kg colours if you are a total rookie who thinks $300 is a lot for 20kg of filament.

You are probably going to throw away most of that anyway, so grey is a good baseline, with learning to be had by comparison against the other various colours and opacities/surface types.

Dont fall for the $40/kg of filament because its BeTteR ruse. Unless its carbon fiber or some hybrid that you ACTUALLY need; $15/kg is the right price for PLA and PETG.
>>
>>2794811
>Its a really good printer if you already know what you are doing and have used klipper before.
Just like every other printer /with a new board? Since you already listed more negatives on your own you can see why i rather saw it as "meh", albeit not necessarily all bad.

>>2795331
I'd at least mention ASA in that regard. At least when it comes to bigger prints.

>>2795342
>prusament PETG is at 30€/kg
>favorite local PETG is at 20.5€/kg
>polymaker PETG is at 20.40€/kg
Inflation fucked with all those numbers.
>>
File: 2024-05-04--17-44-21.png (2.13 MB, 1920x1080)
2.13 MB
2.13 MB PNG
>>2795359
So then stop buying the shit from asshole resellers.
https://www.aliexpress.com/item/1005006064018961.html
https://www.aliexpress.com/item/1005006007948082.html
Theres plenty of good quality cheap filament if you just start buying direct from the source rather than some opportunist who imports the same shit and warehouses it.
Shit, even kingroon is a 1st level reseller, you could always go direct to the factory via alibaba. I've done that before with good results. But minimum order tends to be ~30KG and they seem to always want to charge the highest DHL/FE rate to ship it. Still, some research and waiting a week or three would get you much cheaper filament.
Buying euro "made" filament is your problem. They probably just import and relabel it anyway.

....currently printing chinese PETG from 2018.
>>
>>2795126
Do you need your parts to be particularly strong? If so I'd recommend a printer that can do ABS out of the box. Also 400 euros seems kinda steep for a KE, that's more Ender V3 price (not the KE).

>>2795340
>organic shapes
Learning blender might not be a terrible idea. It isn't parametric (for the most part), but it's good for more organic shapes, and IIRC it has a plugin that makes it passable for mechanical CAD work. Though most people I've seen use it also learn a proper parametric CAD program, and import the sculpted shape into that once it's refined. If you're not a Freetard like me, go for the free version of Fusion360 or Onshape, that or pirate Solidworks.
>3d scanning and AI
It's getting there, but the landscape is awash with tech demos and expensive scanning hardware, it's not really consumer ready yet. Though maybe there's a youtuber keeping his finger on the pulse of the 3D scanning field that's done the dirty work for you and can tell you the latest cost effective setup. I still think we're a while off being able to scan something in and have it be good enough without any hand finishing though.

Casting is absolutely the way to go, there's lots of custom rubber or composite things that you can't buy, and being the guy who can 3D print a mould and cast one will set you apart. Both things like silicone and urethane, but also forged carbon. Though stocking all the different resins is expensive if you want any sort of variety (e.g. a range of shore hardnesses).
A router isn't as useful unless you can do aluminium, but custom wooden parts might be a good market if you know where to look.
A diode laser can do business cards and engraved art, but I don't think its potential is that great by itself. I use one for PCB making.

>>2795359
>I'd at least mention ASA in that regard
It's ABS but twice the price. If you need something outdoors maybe, but I'd just use ABS and paint it.
>>
>>2795372
Thank you for the reply, so would Bambu Lab X1 Carbon priced at 1900euros be worth it? And yes, I would like to print some parts for car aswell for tabletop cnc machine and other household items
>>
>>2795374
X1C is maybe overkill, you can get a P1S and add a hardened nozzle. Without the botnet, a K1C or SV08+enclosure is probably good enough if you don't mind tinkering a bit. Watch some reviews to see what's worth getting and what isn't, spending 2k on a printer deserves a few hours of research at the least. I recommend Aurora Tech's reviews.
>>
>>2795380
Thank you once again, I will look into them, but what are some things to look out for and recommended upgrades? What about enclosures, cameras, sensors, upgradable steppers, linear rails etc...
>>
>>2795330
Nice results anon
>>
File: IMG_20240504_162500.jpg (899 KB, 1764x3567)
899 KB
899 KB JPG
Going to put some legs on this 2000mm oak counter top I got. /3dpg/ related because my printer will live on the table.
>>
File: 1714583605197673.jpg (7 KB, 250x243)
7 KB
7 KB JPG
>>2793796
i have limited time/access to my school 3D printed (PETG) is it really that bad to print multiple parts in the same batch ?
>>
>>2795434
Depends.
Is the printer enclosed? You'll get parts that are as strong as if you printed just the one.
If the printer isn't enclosed, you might lose some layer adhesion.
You best practice really is to make sure that the part actually works/fits before you print dozens of it.
>>
>>2795449
there will be a single model of each part, i'm supposed to assemble them together like Legos
>>
>>2795451
you'll be ok.
>>
>>2795434
Watch out for stinging. Even with models that dont have issues with that, multiple parts in the same batch will create stringing if the printer/filament is susceptible to that.
>>
>>2793796
Please help me. Should i buy a Artilley Sidewinder, Anycubic Neo or the sovol S06, i'm complete noob and i just want to print small trinkets and other stuff and i don't wanna deal with another eternally unfixable machine just the minimal mingling with software and other technical issues.
>>
File: your welcome.png (133 KB, 1535x740)
133 KB
133 KB PNG
>>2795458
>sovol S06
your are welcome
>>
>>2795461
Thanks. I gonna check that out.
>>
File: beh.jpg (36 KB, 815x497)
36 KB
36 KB JPG
*heavy sweating*
>>
>>2795468
whats the UV for ?
>>
>>2795489
dabbing on pla-lets with superior uv-impervious petg
jk just a default led colour
the right tree support failed and then recovered with an ant bridge of loose extrusions and stringing right before it reached the main body of the part, it worked out in the end
all hail petg and its stringing (and uv resistance)
>>
I cant print anything on my printer without a brim using PLA.
The bed is levelled and cleaned with isopropyl alcohol. Z offset is just perfect: if I increase it even by 0.01 I will get artifacts on the second layer due to squishing the first layer too much.
I am printing on a glass bed.
Printing at nozzle 230 for first layer, 215 after, bed 70 first layer, 60 after.
Any ideas?
I don't want to believe that these printers were made to be used out of the box with a gluestick or hairspray. I do believe that the print should stick.
What I have tried before was using hairspray but I can't stand the smell, though the prints stick.
I tried lowering the Z offset even more, doesnt help, just makes worse.
I tried lowering the bed temp to 55, 60, 65, didn't change anything adhesion wise.
Anything else I could try to make it stick without using anything extra like a gluestick/hairspray etc?
>>
>>2795523
1. use a brim you cunt
OR
1. suck it up and use hairspray faggot
>>
>>2795523
>bed 70 first layer, 60 after.
why tho, you're literally causing it to cool and disconnect.
>>
>>2795523
thicker first layer
>>
File: Untitled.png (11 KB, 665x249)
11 KB
11 KB PNG
>>2795528
I said Im already using a brim to print successfully but would rather not use it to avoid cleanup

>>2795529
I dont know to be honest, I read from somewhere and gave it a try. It's not how I have always printed, before testing it like that I did just plain 65c bed which didnt really work either. This is just what it is at the moment for testing to see what works and what not

>>2795530
It is, pic related
>>
>>2795533
bigger extruder multiplier on first layer
>>
File: prusa724.jpg (444 KB, 2000x949)
444 KB
444 KB JPG
I'd like to make this area stronger.
I learned about modifiers yesterday.
I'd like to make walls 3 layers thick in the designated area but prusa adds a wall at the intersection of my stl model and box modifier.
How do I thick only existing walls and not create another?
>>
>>2795587
You don't, this is a limitation of current slicers that people are actively working to improve.
>>
>>2795523
>dont want to believe
I mean you can keep denying chemical principles all you want, or you could just get some PEI for the print surface.
>>
>>2795592
Best of luck to them then!
Curious that when you configure the modifier to change fill, it works like intended, it adds more infill out of thin air without any wall surrounding it.
>>
>>2795533
after a few failed prints years ago i set my pla bed to 60, and have neveer changed it since.

i suspect glass sucks for pla.
get a magnetic springsteel pei sheet.
or use a layer of pet tape.

its conceivable that you are one of those people who has no idea what clean means, so clean your bed with some ammonia based spray cleaner... windex seems appropriate for glass, but any household surface cleaner should do. Or wash it in dishsoap, if thats how you lean.
Then iso it to make sure theres no fragrances or whatever left over.
...but glass probably just sucks for pla

lowering bed after the first layer isnt too bad for contraction, but its something I stopped doing.
PLA: i use 60C, or 70 if its a large surface area piece.
PETG i use 70C, or 80-90 if its a large surface area piece.

how do you know your 1st layer height is 'spot on'? print any old thing with a 4 line wide skirt, set general layer height to something thick, like 0.3mm, and cancel the print once the skirt completes.
Let it cool a bit and *measure it*. Hopefully you have calipers that arent the usual garbage.
>>
>>2795587
put the modifier only where a wall already is, but a smidge deeper into the model
>>
>>2795619
Glass does suck for PLA, it sucked 15 years ago and it still sucks now. You can greatly improve the performance of a sheet of glass by lightly sanding it to a frosted texture. I don't know why anyone would waste their time doing that when a magnetic spring-steel PEI powder coated build plate is so cheap and just SO FUCKING GOOD.
>>
>>2795587
Alternately; put an array of tiny cylinders through the relevant piece, if they are small enough they wont even be printed as holes, just solid cylinders within the interior of the print.
>>
>>2795626
>tiny cylinders
and by this i mean they DO need to be hollow, or the slicer will just register them as irrelevant and ignore them. Vary their wall thickness to make more robust support, just do a test first to see how small you need to make the interior diameter. something less than 1/2 a line width should be about right.
>>
>>2795362
>I don't have it, therefore no else has as well
Calm your projection. You can literally drive down to Löhne and get a quick tour of Material4Print yourself if you ask nicely. Made in Germany is their stick, so they're more than happy to show it off.
>Shit, even kingroon is a 1st level reseller
*That* is surprise to you?

>>2795372
And a quarter of warping. You know
>At least when it comes to bigger prints.
>>
>>2795637
>Löhne and get a quick tour of Material4Print
so then pay euro overheads, euro energy costs, euro lazygirl tour guides, on top of the menial spool sealer person wage for something that is largely automated and can be done for less in china.
>*That* is surprise to you?
no, thats why i said it....
Dealing with manufacturer is usually cheaper, but larger minimum investment.
Dealing with 1st/2nd tier resellers is a good convenience/consumer protection/price point. With the exception being resellers in a different country to the manufacturer: they add too many overheads.

My point is that you feel you are getting something *worth* more than twice the price by buying german, but in general you almost certainly aren't.
Paying euro prices for chinese raw material with all the energy and labour costs added in germany ... or just paying for 'i want it NOW' convenience is your choice.
...although i do understand that euro customs can be a problem sometimes. Fortunately - for freight/customs - I live relatively close to Asia. So its a no-brainer when deciding on purchases that we dont mass produce locally.
>>
what enclosure will my neptune 4 plus fit in?
>>
So I'm having issues with my Ender 3 V3. Any time I print any type of filament that isn't just normal PLA, currently trying to print marble filament, the extruder just keeps skipping. I've confirmed that increasing temp does nothing, slowing flow rate does nothing, slowing the print speed from 220mm/s to 80mm/s does nothing. So I'm wondering if I'm just missing something. And before anyone says no you can not currently get larger nozzles for this thing because they are proprietary nozzles.
>>
>>2795637
>And a quarter of warping
I did not know that. Is warping still an issue with a properly preheated enclosure?

>>2795652
Is the nozzle actually getting hot enough? If you push the filament through by hand or touch it to the side, does it melt? What temperature does the screen show? How easy is it to force PLA to skip in the extruder if you try to hold it back from feeding? It’s either a firmware problem, a thermistor problem, or an extruder problem.
>>
>>2795669
It should be getting hot. To eliminate some things I took the marble filament out of the extruder as it was printing and skipping. Put normal generic pla in and it started printing completely fine. With the way the V3 is designed you can't push filament through manually. You have to select the option for the machine to extrude or retract filament so that option is not possible. I'm running it at 220F nozzle temp for both filaments. I sent a request into Creality to see what they give me for info too. Maybe it's a firmware update, I bought it the day it came out and haven't updated it but again the fucker should just work out of the box for 300 some fucking dollars
>>
Is the Ender-3 good for terrain?
Also can I just use it after assembly right out the box, or do I need to buy other things?
>>
>>2795673
>220F nozzle temp
lol
lmao
>>
>>2795696
It says to run that for PLA
>>
>>2795703
celsius
>>
>>2795331
Went ahead and bought a small selection of heat set inserts. Going to upgrade some of my prototypes a bit. Also might print a bung for a drain cleanout hole in my front yard. Probably going to start putting together a portfolio of builds and capabilities.
>>
What's the best adhesive for bonding TPU to PLA?
>>
>almost out of filament on the spool
>decide to use the rest on printing simple pen holders in vase mode
>first one prints without flaws
>scale the second one up to 150%
>everything's fine for a while
>suddenly begins to stop at the same place every 2-4 laps around, for a second, creating a girthy seam
what the fuck
>>
File: 1710887358005452.png (168 KB, 481x397)
168 KB
168 KB PNG
>>2795762
like this? https://www.youtube.com/watch?v=ZM1MYbsC5Aw
>>
>>2795763
yeah kinda like that but in a straight line
added M413 S0 to start gcode, hopefully that fixes it
>>
>>2795726
E6000

This stuff I used for the first time to bond TPU to ABS, and holds great. If you live in Murica, I have seen it at crafts store like Michaels, and Hobby Lobby, I am sure you an find it at walmart or target as well.

I would, if I were you, have different types of adhesives for different materials, completely optional.

>acetone
Use this to mix with ABS pieces to make ABS glue (guessing game on the ratio mixture) or just acetone brushed on by itself (I normally do this im just joining piece together to form as one original piece

>Super Glue
Again assuming youre in Murica, harbor freight brand super glue is rebranded gorilla super glue 4-5 dollars for 20 grams, while its 7-8 for 25 grams of gorilla glue brand.

>Weld-on #3, methylene chloride, or diy
This one I havent personally tried it is however something I have been looking into recently, this welds multiple plastics together, like petg to pla or petg to petg (theres a weldon specifically for petg iirc, but its probably using same chemicals) or pla to pla (again weldon does have a pla version, but like I mentioned, could be using same chemicals)

Reason why I mentioned diy your own as a possibility is on the assumption you previously bought acetone, then you can get plexiglass and have it dissolve in some acetone.
https://www.youtube.com/watch?v=iUqscl4EGxQ

>epoxy
I have used this stuff a few times when I wanted to bond hook nad loop to pla when I made my one sanding grip handles.
>>
>>2795828
>acetone and plexiglass are more expensive than just buying the glue
>>
>>2795828
Yeah it's at my walmart, I like the little tubes
>>
>>2795839

Yeah you can do whatever is cheaper, my work has acetone in storage closets and I was able to take some home, I guess I should've added "whatever is cheaper for you"
>>
>>2795711
Yes, sorry I mistyped that.
>>
In the market for a 3D printer for around $1k for more or less professional level prototyping - I'm going to be using my prototypes for robotics and contract work. Unfortunately, getting powder based 3d printing is out of the question (had the best quality prints with that tech) and I don't want to deal with SLA since that crap is temperamental and I will be using my prototypes under compressive loads. This basically forces me to use FDM.

Research said that Bambu's X1C is the current best one on the market but I have had bad experiences with Chinese 3D printers (Flashforge is absolute garbage) and I really don't want to deal with having to lock down my network security.

Prussa was second best but it's just highway robbery at those prices and its open air.

So, what do you guys use? What has worked well.
>>
>>2795828
I asked in stupid questions thread but going to ask here as well since you mentioned it and technically it's about polymers. Any idea if e6000 will behave similarly to super glue used in this article?
https://www.wikihow.com/Fill-Plastic-Holes
>>
>>2795954
I have an X1C. So far I haven't had any issues with mine, though as with all printers some people do. It has largely been a no brainer to use it and I have been able to skip all the tinkering and bullshit that other people once claimed to be an expected experience of 3D printing and I have been able to focus on solely printing and designing. The X1C has been released for a couple years now so I don't know if they intend to release a slightly larger model, if you lean towards them that could be something to wait for.
>>
Been away from 3D printing since the Anet A8 era, before all the hype died down.
Is 3D printing appreciably better or is it just consoom consoom consoom?
Is everyone still using PLA and PETG?
>>
>>2795962
No, the stuff can turn into kind of a mess and like ABS glue whatever is evaporatings causes it to string like cob web thin type stringing and just keeps stretching.

I would look at to see if it mentions what kind pf plastic thay you are working with and s ee what ways would be best to weld or glue together cause baking soda and super glue seems like an odd combination.
>>
>>2795962
oh forgot to mention the stuff fucking stinks and lingers, I mean you can try it, but I have read rumors about ingredients in the glue that could cause cancer and parkinsons, even if its not true it stinks anyways.
>>
>>2795974
a bunch of off the shelf printers can hit 500mm/s, print tpu and ASA out of the box, but it's still not very useful outside of very niche applications
>>
>>2795965
How well does it compare to the PS1? Is the X1C actually worth double the cost?
>>
>>2795669
With enough heat soaking you probably get by, but at some point i just gave up and put in a spool of ASA for everything reaching >400cm2 footprint. It's just not worth my time anymore dealing with misprints.

>>2795954
>prototypes for contract work
So any material requirements?
>>
>>2796135
ABS is more than enough for the vast majority of prototyping work I have done and will ever need besides the possibility of PET/G or HDPE for sliding contact points. Anything that needs a works-like and feels-like job or overmolding prototyping will have to be contracted out for machining or other manufacturing methodologies.
>>
File: spongeadapter.jpg (3.22 MB, 4000x3000)
3.22 MB
3.22 MB JPG
From pain in the ass to fixed product in under an hour. Quickly designed and printed an adapter to keep my dish sponges sitting up in a basket air drying. Next is a bung for the bunghole on the drain cleanout on my house that was burried under cement. Hope to have that one measured and modeled in 10 minutes and printed in tpu in a couple hours.

Starting to really like using this tool. Those ender 3s were hell. The flashforges or whatever they were I dont even want to think about.
>>
File: IMG_20240506_084908.jpg (1.86 MB, 4080x3072)
1.86 MB
1.86 MB JPG
The table is finished and i like it very much.
>>
File: 1683886463824235.jpg (352 KB, 2724x2050)
352 KB
352 KB JPG
>>
Hey guys. I know, at a surface level, bad filament clogs printers, but I go to a makerspace, so everyone is massively autistic. What can I say to them that will fucking stop them from using absolutely retarded dollar store filament? What's really causing the clogs?
>>
>>2796203
any kind of airborne particles

dust and grit from anywhere else... sanding? grinding? windy outside? smoke?

grubby cunts who dont keep filament spools sealed and cart them around in backpacks

filaments generally have a fairly strong static electric attraction, most noticeably when separating print from the bed, but any small particles can be static-sticky too, so if there is particulate shit nearby it will stick to nearby things... like your spool of filament

Does the extruder have any particles of filaments in its gears/nearby surroundings?

If you switch filaments you need to use the appropriate temperatures to purge the prior stuff... ie: switching from (hotter) petg to (cooler) pla may cause blockage problems
>>
File: 01075689.jpg (90 KB, 640x784)
90 KB
90 KB JPG
Is there a way to buy tiny samples of known good filament?
Sometimes I want to see if a problem is caused by the filament without opening a new roll.
>>
>>2796218
yeah, they are simply called filament sample packs, like this one: https://www.3dprima.com/filaments/filament-samples-3mm/primaselect-sample-pack-2-85mm-pla-abs-abs-hips-petg_21898_2932
>>
>>2796189
Looks sturdy. one of those kits from home depot?
>>
File: 20240506_091125.jpg (3.41 MB, 4000x3000)
3.41 MB
3.41 MB JPG
Creality K1C stock settings, printing currently on gray Kingroon PLA but it's happening on white creality PLA as well. I am a dumbass and broke the wires on the original hotend so replaced it with the FlowTech Micro Swiss hotend. Also have a PEI plate. Using Orcaslicer default PLA settings for the KI except changed the temp to 230C instead of 220 as I'm mostly using regular PLA. This was working fine for similar and more advanced prints on the stock hotend, but now this kind of thin, skipped layers seems to happen near corners or in between straight edges from each other, never on the corners or in the center. I let it auto level and calibrate each print, but I've only had this issue since changing to the new hotend. Any ideas or similar experiences?
>>
>>2796203
you're mad because other people are using cheap filament? I don't understand. also printing forward with fucking garbage pasta tier brittle shit filament is fine it's retracting that fucks shit up and breaks the filament. bad filament snaps when you bend it. using the tool hardened end of the roll (e.g. it's been through the extruder) will make it brittle, especially if it's been sitting in the bowden tube. waste a foot of filament each time you start with the printer. wet filament will pop and leave craters in the print.

two other clogs: one from burnt debris in the nozzle, happens when overheating or using filaments with additives. calls for a cold pull. get nylon, do it right. two is a gap between the heat break and ptfe or heat block allowing molten plastic to clog, sometimes occurs only on long prints with heat creep depending on the gap location. symptoms are not being able to pull the filament when cold/cool or not advancing when hot. disassembly will allow you to extract the filament forwards and symptom will be a wider section of filament.

>>2796207
spools that sit a long time and get dusty too, this mainly contributes to burnt foreign matter in the needs a cold pull scenario. pushing hot plastic through a tube with a gear is pretty robust in terms of susceptibility to dust and dirt in my experience. Fibers and shit from a backpack is a new one on me.
>>
>>2796255
Can you tell me how is K1C as I plan on getting one, with stock hotend. According to YT videos, its good machine, but what have you made with yours and what sort of problems did you encounter?
>>
>>2796298
I'm mostly made figurines and they came out well besides the first layer being a bit off sometimes. The pei plate fixed that. It has clogged up on me a few times and the hotend heating block being able to spin around when you unscrew the nozzle (which is what caused the wires to break) definitely sucks.
>>
>>2796306
So, would you recommend K1C printer or nah? I would print pieces for household and maybe carbon fiber enchanced parts for car and abs most of the time, or should I get some other printer?
>>
>>2796307
When it was working fine with the stock hotend it was great, I'm only having problems after installing the "better" hotend. I'd recommend it but be prepared to have to troubleshoot and fix things like any Creality product.
>>
>>2796313
Never owned a 3D printer but I am looking forward to it
>>
>>2796249
I got a 25mm thick countertop in solid oak from a construction store. The legs are witre manutan workbench legs and rated for 500kg each.. It would have been nicer if the coubtertop was 40mm thick instead.
>>
File: Playsmart online.png (682 KB, 1430x854)
682 KB
682 KB PNG
>Ex-gf left her old Polaroid Playsmart at my place when she moved out
>Because firmware was fucked so wouldn't work and no way to reflash
>Took it apart and threw a new pico in there for its brain
>Hooked it up to my Pi 4B that runs my CR10 Frankenstein with Brainlett for a new Klipper instance
>All wiring and pins correctly defined where it homes correctly and senses temp
>Added cooling for board and new standoff mounts for board formfactor
>Got multiple USB cams running on Fluidd
>PID calibrating now.

Not gonna lie, I feel a little badass, even getting a shit JoAnne Fabrics special running again. Can't believe they asked $600 for this a couple years ago.
>>
>>2796189
>>2796194
why did you feel the need to come back 20 minutes later and post the exact same thing
>>
>>2796431
Try again.
>>
File: cc_gp.png (113 KB, 162x501)
113 KB
113 KB PNG
>>2794542
I like Craftcloud, they'll let you quote an entire print without committing to an account and offer a ton of different materials from a zillion different vendors.
>>
File: bracket.jpg (89 KB, 1200x1600)
89 KB
89 KB JPG
>>2794451
hmn that's some juicy layer porn
>>
>>2795359
>just like every other printer /with a new board?
The difference here is the flying gantry and bed size. A 350mm voron is 1500. This is 550. There are many advantages to having a capable flying gantry such as adding a tool changer or using a belt kit for part ejection etc.
>>
PETG doesn't react to isopropyl, right?

Does it react to bleach? google keeps throwing out very vague "yeah maybe, probably idk" answers
>>
>>2796619
No to IPA, bleach might attack some pigments but I'd assume not. Just do a test sample.
>>
redpill me on alfapro
>>
>>2796407
It's always nice when a project works the way it should
>>
File: 1010.jpg (2.53 MB, 4000x3000)
2.53 MB
2.53 MB JPG
What si this problem and how can i fix it?
It's creality white filament, on Ender v3 se
>>
>>2796255
If anyone is having similar issues, I managed to mostly fix it by increasing the flow rate to 110% under expert settings. Still need to calibrate it a bit better but it's much better already.
>>
>>2796677
Vase mode would fix it and it might be suitable for your part. More Linear/Pressure advance might fix it in other parts.
>>
>>2796677
You mean those end of layer blobs, or the horrendous layer shift?
>>
File: 1704001190840271.png (2.15 MB, 1144x834)
2.15 MB
2.15 MB PNG
>advertise machine with six weeks backlog
>show off long anticipated enclosure
>just two more weeks, guys
At this point he gotta be trolling.
>>
File: Stefan Sandladerer.png (438 KB, 875x400)
438 KB
438 KB PNG
>>2793796
Alright, /diy/, you have the opportunity to kill exactly 2 krauts, but one has to endure long, painful death in which they remember every miserable moment of their faggot life, and the other dies quickly with barely any pain.
Which one of those two gets the quick bullet and why?
>>
>>2796761
>horrendous layer shift
The ender doesn't have dual-filament capabilities, i think he just stacked two prints for demonstration purposes
>>
I have an anycubic mono 4k that i haven’t used over the winter so it’s been collecting some dust. Are there any best practices i should do before printing with it again?
Re-lubricate anything, check any particular parts?
>>
File: GLEC0bAWUAA37Na.jpg (142 KB, 1284x903)
142 KB
142 KB JPG
>>2795434
Ignore the enclosure faggot, PETG loves to jam in one of those.
As long as the filament is dry and settings are fine the parts are going to be very strong and have a nice glossy finish.
If it's a bit damp or your settings are shit you might have stringing issues but a quick flash with a lighter usually takes care of those easily.
If it's wet as fuck and you have no fucking idea what are you doing it's going to be worse than no-chamber ABS and will be brittle as shit regardless of layer direction relative to the break.
I print multiples all the time and the stringing is minimal even with a 3-week old spool hanging on my constantly breaking chinesium frankenstein, and i'm certain my room is anything but dry.
Pop it under a filament box with holes on top of it on a bed 10 degrees lower than the bed temperature for a few hours if it seems to go to shit and it will be good as new as long as it's not past the flooding point.
>>
>>2796792
>10 degrees lower than the bed temperature
Fuck i meant 10 degrees lower than the bed in your filament profile.
>>
Are there any "planning" sites or middle school level formula to help me determine the most efficient order to print different items to make the most out of each spool? I'm doing a work project where I need to print 5 different drawer organizers for 60 total drawers, all of them ranging between 100g and 220g each.

For example, one of these files needs 175g of filament, so on a 1kg spool I could get 5 of those printed, but still have 125g left so I could print a 100g file out of it too.

I could sit down for a few minutes and manually figure out the most efficient use of each spool but is there a faster, premade way where I can just plug in "need 60 of these 100g caddies, 60 of these 175g caddies, etc" and have it automatically tell me how many of each to print per spool?
>>
>>2796799
nvm i just remembered filament runout sensors are like $10
>>
>>2796887
>not getting an ams so you can automatically hot-swap filament when it runs out
>>
File: platte pei und magnetfuß.jpg (169 KB, 1745x1024)
169 KB
169 KB JPG
>>2793796
gm kings, I have a question:
Will I die if I use a 235mm PEI steel plate on a 210mm heat bed? Will there be warpage from the overhang?
I own a kingroon kp3s pro s1 with bed size 210x210mm. I am currently using the stock glass bed. I would like to try using PEI with the magnetic sticker, however I can't seem to find any that are 210mm for a reasonable price. I do see 235mm ones for cheaper.
>>
>>2796977
it'll be fine
the magnet might sag though, or at the very least be stucky underneath, so you might want to trim it after you stick it on
>>
>>2796977
bambufags don't want you to know this one weird trick a local anon uses to increase the size of his bed! creality shitters shocked!
>>
>>2796787
left gets it slow because of the more punchable, dysgenic face and especially his cancerous catchphrase
t. kraut
>>
>>2796977
Anon please, official kingroon store has one exactly for your printer for like 25bucks.
>>
>>2796792
Fuck youdumbass communist mother fucker
>>
>>2796978
Thank You.
>>
>>2795974
Things were just copied shit over and over until bambu came along.
Now it's copied decent stuff.
Creality still a shit.
more things change, more they stay the same
>>
>>2795523
glass buildplates are for bundles of sticks

your nozzle and bed are too hot. 190/60, dont need different temps for first layer, just dont turn on part cooling until layer 2/3

havent had an issue since i started using those settings and properly set my offset.
>>
I'm crazy and I want to try to use 3D printed "boards" as a face frame for a tiny shelf.
Have you ever tried to drive a brad nail through PLA? What happened? What would you expect to happen?
How well does polyurethane construction adhesive bond to PLA?
>>
>>2793796
Anybody here tried the amolen dual color PLA? How is it?
>>
>>2797257
>PLA? What happened? What would you expect to happen?
the top, and probably bottom layers would split, and the infill would collapse a bit, any stresses on the member thereafer would contribute to mechanical failure of the fixing.
If you screwed, or maybe even nailed into printed holes it would work better, but probably not much unless the nail had a lot of room around it and the pla.
PLA is brittle.
PETG would deal with crucifixion better, but it'd likely still split.
>>
File: 2024-05-09--13-40-15.png (2.14 MB, 1920x1080)
2.14 MB
2.14 MB PNG
I just unearthed another vintage filament; 2018 TPU.
Never printed this stuff before, but my ridiculously uncommon sink needs a grommet. So once more unto the breach.
>>
>>2794705
I want to maked babby with Aurora
>>
File: 2024-05-09--16-31-28.png (1.41 MB, 1920x1012)
1.41 MB
1.41 MB PNG
>>2797325
well, that worked really nicely. first try. seems like it'll seal adequately too.
fuck yeah 3dprinter

>>2797309
model the holes and screw them down... now thats another matter. PLA becomes brittle in prolonged direct sunlight (like; years), but PETG should be pretty resilient. That said it might be more likely to bow than PLA.
>>
File: 2024-05-09--20-55-48.png (149 KB, 320x425)
149 KB
149 KB PNG
entredasting....voice to print:
https://www.instagram.com/reel/C6r9seFPvF0
>>
File: IMG_20240508_204429.jpg (550 KB, 3840x2160)
550 KB
550 KB JPG
Testing. The nails are holding and the parts are not visibly losing their integrity.
I think this is gonna work.
>>
File: 1704572383984925.jpg (109 KB, 675x900)
109 KB
109 KB JPG
trying the free spool of transparent rose color petg I got from my bro and dialing in the settings (still fuzzy), did a vase print of a pen holder model to see if I got the seam again on my previous vase mode prints, and this is the girthiest yet
I was about to go on a long rant how the feds are somehow doing this to me but then someone whispered in my ear: turn off timelapse in Octoprint
and since I am using octo4a on a crusty old tablet with the cam off, it suddenly makes sense
>>
File: ready.png (183 KB, 815x798)
183 KB
183 KB PNG
what programs do you guys use to split a model into pieces
Im done with this model but after putting it into a slicer the amount of supports for the main body is insane, im thinking of splitting it into 3
I tried in blender but it fucks up the mesh WAY too much
>>
>>2797633
I use blender. How are you trying to do it in blender?
I use a plane with a solidify modifier to give the joint some wiggle room. Then I use a boolean modifier (pointed at the plane) to cut the part.
>>
File: 2024-05-10--10-01-42.png (44 KB, 835x447)
44 KB
44 KB PNG
>>2797640
do you actually use the stuff that stl exporter has?
I used to use it a long time ago, but never really found it very useful. iirc something about the exportname field used to piss me off too.
Returning to 3dprinting recently i just started using the inbuilt one.
Initially i thought id need to error checking of the 3dprint toolbox plugin, but i'm finding that I totally dont, even with linux orca being incapable of mesh repair.

>>2797633
So just slice it vertically down the center from front to rear and lay it on its side to print.

Sounds like you expected blender to have a button that neatly separates and arranges parts. (lol)
Try select it > tab (edit mode) > 1 (vertex mode) > select a vertex > ctrl-L (should select all linked vertexes, depending on the model topology that may be a single component of the mask) > p (separate).
tab ... and now you should have two separate objects that you can examine for ease-of printing.
This method wont necessarily make things easier to print, but it may be useful in making breaking it into smaller parts to print separately and assemble later.
>>
>>2797615
So what was timelapse doing to mess it up, halting for a second or something?
>>
>>2795828
Good tips
>Probably the same chemicals
Maybe you could check the Material Safety Data Sheet?
>>
Any camera recommendation for crisp 1080p print recordings?
For some reason arducam 16mp seems to do better in an enclosed printer than their 64mp version. C922 is around for the c920, but doesnt seem to perform any better in actual image quality? Wyze doesnt seem to be thing here, without import.
Currently leaning towards arducam 16mp due to it's low foot print.
>>
>>2797749
try using a 2-5MPx camera instead of 16
1920*1080=2,073,600 = 2MPx

The 5MPx OV5647 based cameras I use on my raspberry pizero security cams are best at 1296x972 the res is basically the same as 1080HD visually, but the cams ability to output the lower res is more useful.
It can do still images of 2592x1944(=5,038,848), hence the 5MPx specification, but how efficiently the camera chip can spit that out in a mjpg streamable form - either by scaling or cropping - is really what defines 'good' in a webcam. Not the maximum pixel format.

It may also be that the USB is a bottleck at some point: cam chipset, cable, connector, bus, etc
Or maybe the lens resolution is garbage compared to the sensor resolving power, this seems ot be pretty common.
Or maybe the compression the camera or the OS arbitrates for the stream is just shit for 'crisp 1080p' and better for onlyfans.
>>
>>2797767
I'll check em. But from everything i saw yet, the regular 5mp raspberry pi cameras all sucked ass and not the good way.
>>
>>2797769
i didnt mean picams specifically, i meant stop looking at more megapixels as being better
Are you even using a CSI socket pi board that can drive a pi camera?
The skipr board i use in my printer has integrated pi, but no CSI, so i have to use USB cams.
These dont totally suck.
https://www.aliexpress.com/item/1005005510880232.html
But I'm also not 1080p-ing it
v4l2-ctl --list-formats-ext -d /dev/v4l/by-id/usb-170428-_Integrated_Webcam_HD-video-index0
#ioctl: VIDIOC_ENUM_FMT
# Type: Video Capture
# [0]: 'MJPG' (Motion-JPEG, compressed)
# Size: Discrete 1280x720
# Interval: Discrete 0.033s (30.000 fps)
# Size: Discrete 640x480
# Interval: Discrete 0.040s (25.000 fps)

this is pretty garbage, but i AM 1080p-ing it
https://www.aliexpress.com/item/1005002342741767.html
picrel: >>2797325
# [0]: 'MJPG' (Motion-JPEG, compressed)
# Size: Discrete 1920x1080
# Interval: Discrete 0.033s (30.000 fps)
# Interval: Discrete 0.033s (30.000 fps)
# Size: Discrete 2592x1944
# Interval: Discrete 0.033s (30.000 fps)
# Size: Discrete 2560x1440
# Interval: Discrete 0.033s (30.000 fps)
# Size: Discrete 1280x720
# Interval: Discrete 0.033s (30.000 fps)
# Size: Discrete 640x480
# Interval: Discrete 0.017s (60.000 fps)
# Size: Discrete 1920x1080
# Interval: Discrete 0.033s (30.000 fps)
# Interval: Discrete 0.033s (30.000 fps)
>>
>after writing all that
>figures he should find out what an arducam is
oh shi... its a CSI cam.
Ok so you are using a pi then.

I doubt there is a simple way to know what is 'better' until you use them for a while, unless you can find someone who has done v4l2-ctl --list-formats-ext on all of them you cant really know whats likely to be a res/rate that the hardware/firmware will like best.
Buy one and let us know with screenshots.
>>
File: 2024-05-10--14-37-23.png (346 KB, 1024x846)
346 KB
346 KB PNG
lol, WTF
>>
btw, autofocus is probably a bad idea, unless it has a v4l2-ctl control to specify and lock the focus distance
Similarly fixed focus lenses are always focused too far away, so unless you know they havent glued the focus helicoid into the mount thread it's probably also a bad idea. That's probably why youve heard the picams suck; inability to focus them closer.
>>
>>2795770
thats a z seam the puckered herpes version is from randomizing the seam
>>
>>2797776
just use a crappy 10$ usb webcam
>>
File: 2024-05-10--18-47-24.jpg (930 KB, 1920x2952)
930 KB
930 KB JPG
>>2797633
I happened to open thingiverse just now and this helmet was on the page...
Yeah, whoever made this is one of those douchebags who makes 3d garbage, not even designed for 3dprinting, and then tosses everything onto sites specifically for 3dprinted things.
It's not even a fucking .stl, its an .obj.
And the orientation is Wavefront retarded.
Theres even a visor inside the visor...
These people fucking suck.

So the model is composed of 3 easily separable parts; none of which makes it any easier to print.
The visor(s) could be printed without supports fairly easily, but theres no design for where the two supposedly slide against each other..whatsoever. So The only thing that would be worth printing would be the helmet itself, using a shitload of organic support, and then youd have to redesign a visor to actually fit, and rotate in a way that didnt involve clipping through the helmet.

After scaling it to 60% to get it into my 220x220 plate its ~45% support and 13 hours printing.
Cutting it into printable parts would be the way to go, but youd need to spend a lot of time designing useful joints.
Fuck that.
Fuck these retards who upload this garbage.
>>
File: 2024-05-10--21-37-51.png (213 KB, 826x524)
213 KB
213 KB PNG
What is the white goop that bambu use to seal around their nozzle?
Becasue these fucking things leak like sieves unless I tighten them so hard it's got me worried i'm going to snap something. I't hard to grip the heatsink part without endangering the heater pad.
I noticed that the pre-assembled hotend had white sealant around the nozzle tip when i got it, but i've changed tips since then. I stopped it leaking the first time by sheer force, but if theres a goop i can use like plumbers tape I'd like to try it out.
It might actually be the white ?silicone" cement often used to pot PCBs or attach heatsinks. Hardens flexible, but not much.
Anyone know?
lol, i'd try actual plumbers tape if it wasnt teflon (toxic offgassing ~250C)
>>
>>2797817
sometimes the models are grouped parts which you can see and explode in meshmixer. it's not planned it's just another artifact of bad design.
>>
File: 20231210_192752.jpg (765 KB, 2287x1960)
765 KB
765 KB JPG
guys how do I get motivated. I've been staring at fusion360 for two days and of the two things I need to make I've drawn two circles for the hinge tube and gotten no further. I've got the printer I've got the software I've got the skills I just don't want to make anything suddenly; I don't want to make decisions. what the fuck is wrong with me.
>>
>>2797873
Yeah I can see what it is; probably a hud overlay or some other hud/visor layer, but the way these cunts post this shit on a 3dprinting specific site is just obnoxious.
>>
>>2797679
Yes, I use the addon exporter sometimes. Mostly when I'm in a hurry and my part doesn't have any modifiers I need to apply before export. And yeah, I've noticed too that the errors don't seem to matter much at all in orca.
>>
File: lex prime.jpg (155 KB, 999x800)
155 KB
155 KB JPG
>>2797875
when i get like that i usually go print something big and useless, like a prop from a movie or video game, first was the space marine helmet, now its a lex prime from warframe.
gives me a day or 2 of playing around with filament swaps and almost always leads to me designing a stand or something to go with the prop, plus i usually give them to friends and family since all i can see is the faults and what i could do better, while they just see this magical object made real.
>>
>>2797633
orca has built in tools for splitting
>>
>>2797840
That's no sealant, just boron nitride paste. A hotend seals itself mechanically by heat. If yours doesn't, it's either defect or wrong assembled.

>>2797777
Not heard, watched footage on youtube. There's not a single clip out there without blurfest.
>>2797776
Dunno what translates to. They're like 30€ here.
>>2797775
>CSI cam
Yeah, which makes my other concern: pi getting overwhelmed from processing. Not fond of blobs on my print.
>>2797770
Looks shitty enough, i'll look for some footage.
>>
>>2797679
>Sounds like you expected blender to have a button that neatly separates and arranges parts
No I didnt, just the way I've been doing it for my past couple projects took way to long and stuff wouldnt line up perfectly after the split
I try to model these helmets with a way of splitting it but with this one I really couldnt figure it out and have it look right.
I remember a while back using mesh mixer but thats super cumbersome
>>
File: Capture.png (371 KB, 1259x769)
371 KB
371 KB PNG
I made a model of my living room in Inventor to help with visualizing furniture and its placement. My goal is to print out a .dwg drawing of the empty room scaled down to a standard piece of paper, and that much I have done just fine.
I also want grid lines included in the drawing though, to help arrange stuff. Specifically, I want grid lines that are scaled in the same way as the model of the room is scaled, and preferably have them aligned to a useful reference point (say the far left wall) instead of just being arbitrarily slapped onto the sheet. I want these to be actually printed out, so that we can use the grid lines to be able to move a scale paper cut-out and say "oh if we go over 2 lines thats the same as 2 feet" or whatever.

Any ideas? Should I just put a grid in the original part file soit shows up and gets scaled correctly in the derived drawing file, or is there an easy and fast way to do this?

pic related. The top down view is what I am going to print out and what I want grid lines on, the weird diagonal view is just there to show the jist of the 3d model I derived it from and wont be included.

>TLDR: How do I add scale grid-lines to a inventor .dwg drawing file when I print it out?
>>
>>2797633
I've used meshmixer and it's worked pretty well for me. It could be an easier process but it works well enough as long as you are careful with where you make your plane cuts and not changing them after you start. There are probably other options that are better however.
>>
>>2797633
I've used freecad but it's awful. The mesh toolbox isn't parametric like partdesign. Next time I'll figure build123d out.
>>
>>2797875
Sounds like it could be an executive dysfunction issue. (basically means that the part of your brain that does thinking, planning, risk/reward, I have the same kind of thing sometimes. Hate making decisions, committing to anything, have difficulty getting started and especially "getting around to" doing stuff, even stuff I enjoy doing and have wanted to do for a long time. If it's an ongoing thing in your life you might want to get checked out by a therapist or tested or something. Like specifically for executive dysfunction. It can be its own thing or be co-morbid with many other things like ADHD, autism, anxiety, depression, etc. Trust me, if you are seeing similar patterns in other parts of your life, you REALLY want to address it NOW and not 8 years later like I did. Because it CAN be treated.
>>
>>2798052
I have similar problems, but instead of doing nothing I waste a bunch of time on projects that don't matter. Shallow vs deep work, I've heard it referred as.

Also please close your parentheses.
>>
>>2798111
Yeah I notice I do that too a lot. Like even playing minecraft, I avoid the cool but complicated stuff and instead tend to default to just grinding in mob farms to get absurd quantities of materials.

And yeah, I noticed I forgot to close the parentheses right after it posted lol.
>>
>>2798052
what was your treatment?
>>
>>2797777
Only time I use autofocus is when taking pictures with my smartphone.
>>
My friend got a 3d printer and has been using it to make a bunch of cool shit like model tanks and car parts and now I want in. I've spent 3 days watching how-tos and guides but I still can't decide on what to buy for a first printer. All I really care about is print quality with PLA, ant not costing over $1000. Preferably not locked to proprietary software like bamboo, but if it's best in class I can cope.
>>
>cleaning out garage
>find two unopened rolls of PETG, still in their package
>put them aside while I brush the bench down
>look back at the rolls
>one is covered in ants
>move them both outside
>shake the spools out of their boxes
>ants everywhere
>ant eggs stuck to one of the spools
>carry them down to the outdoor tap
>rinse ants and eggs off the spool that had the nest
>rinse the few spare ants that got on the other spool too
>vacuum seal doesn't feel as strong
>look closer, the mark where the spools fell out onto concrete broke the seal
>on both spools
>water actually got inside one of them
>buy some plastic containers the right size for a roll of filament upright
>take spools out of their bags, wipe them down if needed, and shove them in the containers
>add some bags of silica gel i'd been saving for a special occasion
>leave in sun
shit, will the spool that got wet be ok with just desiccant and a bit of warmth, or will i need to really heat it up? i'll probably buy a dht11 module to measure it.
might as well modify one of the boxes to work as a heated dry box that can be directly printed out of anyhow.
>>
>>2798188
Sovol SV08 is pretty good if you don't try to print giga-fast with it, and you're willing to get your hands a little dirty with klipper. I'd also consider the Creality K1 series, and the Bambu series.

FYI PLA isn't very good for car parts, the hot temperatures inside parked cars in summer can really soften and deform them. PETG might be good enough. Personally I'd aim for the ability to reliably print ABS, since it's stronger and cheaper than PLA, but needs a heated enclosure for large prints without warping. But that's my desire as someone who already owns two unenclosed bedslinger printers, so take it with a grain of salt.
>>
>>2798191
>Sovol SV08 is pretty good
That's one I've been looking at a lot, especially with it having a decent print bed size. I can get the SV07 for a lot cheaper though.

The other main ones I were looking at are the AnkerMake M5C for (apparently) being noob friendly and retard proof, and the Ender 3 V3 KE for having a large install base for troubleshooting.
>FYI PLA isn't very good for car parts,
I'm aware, just something he makes. I'd probably just make nerd shit like models for DnD and other crafts, nothing that's really heat or tension sensitive. I'd like to do ABS eventually but it seems like a lot of hassle, I'd like to get some hands on experience first.
>>
>>2798188
A1 mini. if your friend prints look good imagine them with color.
>>
>>2798188
Neptune 4 Max if you want cheap and a big print bed.
>>
>>2797981
>>>/3/
>>
File: file.png (26 KB, 680x363)
26 KB
26 KB PNG
>do orcaslicer flow rate calibration
>Save profile
>Can't find it in drop down
Why?
>>
>>2798201
recommending a neptune 4 max as babies first printer is insane
>>
>>2798190
petg is kinda hygroscopic but a couple minutes of exposure to water won't do shit. it can stand on ready for months in a humid room and still be perfectly printable.
>>
File: 2024-05-11--15-13-13.png (284 KB, 550x542)
284 KB
284 KB PNG
>>2797894
>That's no sealant, just boron nitride paste.
nah. It's in the outermost couple of threads between the nozzle and the heatblock, and its silicon-ish when set hard.
It's almost certainly what i said:
https://www.aliexpress.com/item/1005006079288411.html

>A hotend seals itself mechanically by heat.
Should. But the white silicon-goop it's supplied with would imply otherwise.
The original sealed nozzle worked ok, but i switched it for a 0.6 and that leaked. Then that nozzle got a hard clog i didnt feel like removing in-situ, so i put a new 0.6 in, and strongly tightened it. It leaked.
I just now strongly tightened it again, at 240C. It leaks. Less, but it still leaks.
Admittedly I'm not disassembling it to do this, the silver paste under the ceramic heater and thermistor makes the idea of that extremely undesirable because its messy shit.
But I do have little spanners - they may heve beed supplied with the hotend, i forget - that fit both the nozzle and the end of the heatblock, so I'm tightening it pretty effectively.
The hotend was supplied as an assembled whole, but given that they supply it with sealant in its nozzle thread I'm going to assume they tend to leak without it, and that the bambu nozzle design is a bit *more* shit than the original, and somewhat shit reprap way.
Frankly the original concept of screwing a nozzle into press fit against a threaded heatbreak that is also screwed in from the opposite end of the heatblock... was a slightly shit idea. But the way bambu just hope the nozzle will seal against a ...what is it? steel? titanium? flange. Well its an even more shit idea.
It really needs some kind of 300C-tolerant ring to seal it.
...hmm, or a tiny copper washer...
...or some white goop, which I'm ordering now.
>>
I've always had best results tightening nozzles at high temps. I'll try tightening it at lower temp next time, but it needs to be hot or the traces of petg thats in there already will just interfere.
>>
>>2798248
nozzles are supposed to be tightened hot
>>
>>2798250
I know, but maybe I should tighten them a bit below what I use them at so the thermal expansion is at least working in the right direction.
I tend to tighten them with temps that are above what I usually print at, and that may be an issue with this particular setup.

Haven't done it yet because right now i dont really care if boogers drop every so often; they arent bouncing the hotend and im only doing 200% speed rough prints to get a base design so I can start refining an orthotic.
>>
File: 1715424243721.png (521 KB, 1271x966)
521 KB
521 KB PNG
voron 2.4
ge5c modded and w/tap
why the fuck, x axis looks perfect
resonance graphs are joke, it feels like i am getting raped by it
>>
File: 1715427562236.png (509 KB, 1302x1002)
509 KB
509 KB PNG
>>2798289
here's an x axis if anyone's interested.
mind that the slight bump on 250hz is very likely due to tap (rc8, printed, z1 preload)
>>
>>2798298
i mean 125 what was i thinking lol
>>
>>2797633
prusa has built in tools for that
>>
File: Capture.jpg (320 KB, 2484x1067)
320 KB
320 KB JPG
How can I get cura to add supports at this specific spot?

Also how to get better quality in the areas that are supported so they don't end up looking like ramen noodle cake?

Has to be printed on it's side for strength reasons, it's an arrow nock and will be taking massive impulse forces.
>>
>>2798328
that shit looks small, even miniscule
i'd rather flatten top and bottom to make it printable supportless
>>
>>2798052
>executive dysfunction issue
The immediate goal was too hard or too easy. The top of the Todo list was defective in that it would produce skills and artefacts that aren't worth the effort, so the committee rejected it.
>>
Do .f3d files include any metadata I should be concerned about? Uploading some replacement parts for tabletop games to BGG and figured I might as well include everything if there's no reason not to.
>>
File: stepfiles.jpg (765 KB, 2140x1073)
765 KB
765 KB JPG
>>2798363
Not exactly. They don't contain identify info, and they can't be linked directly back to your Autodesk account. They do have some distinct info, like references to local files, which includes paths that exist on your computer. Someone could potentially glean your Windows username from this info, not exactly a big risk.

It's worth noting that STEP files often include extra unnecessary bullshit, including the possibility of identifying information.
>>
File: PXL_20240511_153540393.jpg (409 KB, 4032x3024)
409 KB
409 KB JPG
>>2798371
Right on. Names and paths are fine by me, my Windows username is just my first name, and that's on my Steam profile anyways, as long as it's not leaking email and last name and stuff like that. Thanks!
>>
File: IMG_20240511_142725.jpg (1.43 MB, 1362x2017)
1.43 MB
1.43 MB JPG
what do you guys think about my robo sluts?
>>
>>2798289
>>2798298
I don't understand what I'm looking at here.
>>
>>2798387
They're lovely whores, anon.
>>
>>2798328
Increase overhang support angle or something.
But that's a Bad model to print. Maybe change those angles up? Don't go above 40° to make it printable with no supports.
Your biggest issue is that you put your z-seams on those overhangs though. You can't start your print in air big guy idk why cura let's you do that anyway...
>>
>>2798328
>How can I get cura to add supports at this specific spot?
LERN2CURA
Use paint-on-supports or whatever cura calls it.

>don't end up looking like ramen noodle cake?

Do an overhang angle test print, and then design your overhangs to be within your printers tolerance.
https://www.thingiverse.com/thing:5194357
Straight angles are going to work better than curves.

>Has to be printed on it's side
>>2798340
>i'd rather flatten top and bottom
this
>>
>>2796977
Don't do it, it will make mustang gas
>>
File: 2024-05-12--12-09-23.png (683 KB, 884x634)
683 KB
683 KB PNG
is it just me or does PLA always favour the low end of its labeled temperature range?
captcha: SSS0Y
>>
>>2798588
what the fuck, are you printing with decade old filament again?
>>
>>2798395
klipper resonance test graphs, you can use it to diagnose your machines mechanical state or tune input shaper parameters
basically i want to get a single peak, not multiple peaks
>>
>>2798631
nope, well yes, but not this particular filament. It's <1 month old kingroon PLA.
Note that it printed the flat section nicely, but lost its shit on the steep slope...mind you its printing 0.4mm TALL layers too, so theres that....as soon as i lowered the temp to 190C s it behaved normally again.
Every PLA ive ever seen has some 190-220C advisory, and yet i found myself using 180C with the end of the last batch, i guess this new stuff is similar.
Starting to lose interest in PLA though, it lacks plasticity::breaks too easily.
...now I'll just have to use the remaining 9Kg on 'other junk'... idk what though.
>>
File: hq720.jpg (46 KB, 686x386)
46 KB
46 KB JPG
is there a program that can recognize a shitty polycam scan of a few different people's heads, and then provide an easier way to modify them?

tl;dr for a project I needed to scan and print a few people's heads and faces but even under the best conditions, the stls I have of the heads are all pretty shit like pic related. I normally use blender for basic geometric shit but for more organic shapes like this I'm clueless other than painstakingly trying to individually move the x,y,z of each individual vertex. I'm hoping to just import an stl and
>import/recognize face
>"alright the eyes look like shit with no definition"
>clean up eyes
>"great ok lets make the smile a bit bigger"
>drag smile slider a little bit to the right

etc
>>
>>2796787
Left gets quick end. Right however not because he looks like a redditor and his YT personality is annoying and gay
>>
>>2798654
reduce the polygon count or remove all the superfluously small triangles... sure. Thats pretty easy to do with one of several modifiers.

Automatically apply faggoty subjective adjustment to aspects of the mesh that only something like modern AI can even recognise? Don't be so naive.
There may be an AI that deals with mesh, and it may even be able to take input like the garbage you specified above, and it may even be interfaceable through blender.... but theres no way you will get it with the zero effort you want.

To do anyhting with technical *with no effort*, you need something approaching a doctorate level knowledgebase in some other related fields.

Dont be a lazy fuck. Look up blenders decimate modifier. Or all of the similar pointcloud reduction tools in any other program.

And if you then still need to mash peoples faces away from reality, look into 'proportional editing' in blender.
>>
>>2798658
>you need doctoral level knowledge to edit meshes

kid just stop
>>
>>2798659
lol, 'kid'... only in spirit.
also: your reading comprehension sucks

If you only know how to manipulate vertices you obviously need to learn the related blender tools. Or another applications tools.

Cleaning up a mesh that is too low res to even capture the details you are talking about is extremely tedious, and you need skills you dont have to even do it becasue it involves *making the new geometry yourself*

Try just putting the texture onto the mesh, that at least visually enhances a shit mesh.
...you did capture the texture at the same time as the geometry...right?
>>
File: 2024-05-12--17-03-44.png (243 KB, 1920x984)
243 KB
243 KB PNG
also: try opening a new blender session with the sculpting default
play with that, remeber the shortcut keys and then import the stl to try your hand at it.
fgt
>>
File: 20240512_160448.jpg (762 KB, 2099x1967)
762 KB
762 KB JPG
>>2798652
ntc thermistors aren't really known for accuracy/precision
there's a high chance that your hotend is hotter that what it reports
it is advisable to check the actual temp
>>2798387
looks very nice, which multicolor solution did you use?
>>2797889
that eerily sounds similar to how i feel about my designs

also check out my asa-cf benchy
carbide nozzle very good
>>
>>2798665
>ntc thermistor
its the stock bambu thermistor, with paste rather deftly applied..even if i do say so myself.
i guess maybe its an ntc component in the glass bulb, but idk. I wouldnt think they could be orders of 10-20C off being correct either.

I've seen this same behaviour across numerous boards/thermistors/hotends too.
I'm starting to wonder if maybe its the matte light grey pigment used in PLA that makes it like that.
I have some vague, decade old, memory of white PLA being weird like that too.

>>2798665
Nice. also: Pshaw. Perfect benchies are so 20th century.
What else have you printed with that machine? Surely you have something equally sexy in the 'not a benchie/calibration cube' category?
>>
>>2798246
>nah
>silicia gel
I'm not even gonna bother.
>>
>>2798671
your reading comprehension sucks too
its funny and horrifying at the same time
>>
>>2798670
oh, matte or white ones
i have no idea what actually is going on but it seems additive/pigment really do some funny stuff.
all "white" filaments i've used were kinda off
they were kinda pain in the ass to deal with and i just stopped using them alltogehter at some point. but i am building ercf now and might have to deal with them again..

the machine is still being tuned and assembled, it was a first print being other than its own parts lol
>>
So I'm printing some functionals part and am getting weird inaccuracies with any circular hole. Overall the parts seem to fit just fine, but for instance a hole that was meant to be 15,1mm in diameter ends up being 14,75, 6mm one is 5,7 or so etc. Is this a mechanical issue or slicer issue(dunno, "wall" builds half inside and half outside of the model or something)?
>>
File: eee.png (377 KB, 698x333)
377 KB
377 KB PNG
https://www.youtube.com/watch?v=wpKCfZ4k4uA
it's up
i wonder if a thickness sensor can be incorporated into this to make extrusions consistent regardless of what bottle you cut up and accomodate for welded joints.
also seems like that papilio belt extruder should be able to print tape in the same way if not better than 2d printer rubber rollers.
>>
File: IMG_20240512_105131224.jpg (366 KB, 4096x1856)
366 KB
366 KB JPG
>>2798588
Print temp towers before attempting prints like this with a new filaments.

The rating on the box is wrong because 1. your printer may be off by few degrees 2. they want you to turn the printer on, load filament in and work no matter what it is - if your printer runs too hot it will still get the basic shape right, if it runs too cold it has the "reserve" to still work, had it been super duper accurate you'd have situations where the client is complaining because he set up the temperature right where it should be and the extruder couldn't make it leave the nozzle.

Generally the lower the temp the better, less overextrusion problems and less stringing, especially if you're doing something decorative it matters.

t. tried to push that side of my cheap printer to the limit
>>
File: 2024-05-12--19-18-16.png (30 KB, 559x350)
30 KB
30 KB PNG
>>2798681
its a common and well known issue
iirc nobody fixes it becasue everyone already works around it by adding a specific amount ot the hole diameter
orca has it as an option, probably any slicer based on the same source does too (prusaslicer etc)

I have premade objects i use in blender for all of the metric bolts etc that i own, their boolean shell component is 0.4mm larger than the bolt mesh component
eg: a 3mm dia bolt has a 3.4mm dia boolean
but that was for a 0.4mm nozzle, im using a 0.6 now so its probably not quite the same, but iirc it still works fairly well (havent printed a bolthole for a while). That said its almost always a good idea to run a drill bit into a hole by hand just to make sure its smooth. If you happen to need to cut out a little extra material to get it in there its still pretty much the same effort. (none)
>>
>>2798686
>Print temp towers before attempting prints like this with a new filaments.
i hear you... but meh; normally i just notice it immediately and adjust.
I was also just using a petg profile instead of actually thinking about it, the only real diff is temperature, and I guess I know now that hot is no good for this grey pla batch, just like the prior one.

>1. your printer may be off by few degrees
It's not off by 10-20 degrees.

It is a little unusual though; when the minimum suggested is 190C and 180C is actually better.
It's possible that this is limited to grey PLA though: All my pla is grey.
I've only just started printing PETG again, and I'm noticing its suggested temps are actually useful. But cyan, green, and dark grey are also different pigments..in a different filament medium.
It's also possible ChAiNa just puts a generic text on all labels and only changes the material type to suit. (Which seems to be like what you were saying)

>Generally the lower the temp the better,
For aesthetic-only printing; sure. But for people wanting part strength or a slicker surface/smoother layer integration; hot is almost always better, at least as far as ive seen with PETG/PLA.
>>
>>2798665
>looks very nice, which multicolor solution did you use?
thanks! Printed entirely in silver PLA and then painted using a fine brush and acrylic paint :3
>>
>>2798686
thats the most beautiful print I've ever see
wish mine cam out that pretty
>>
>>2798673
>I just got this laser that can burn stuff
>I'm sure it's not dangerous to my eyes
>>
>>2798686
what nozzle diameter?

also: just switched to 0.8mm nozzle, becasue now in want-fast-prototypes mode
>>
>>2798694
wait... so you *painted* a print and were passing it off as good print quality?
YOU SLUT!
>>
>>2798694
i... kneel..
thought it was a gold/silver silk filament with red handpainted bits
>>
>>2798673
Too bad none of the reflections went into his eyes
>>
>>2798673
what a moron
i guess those cheap ass handheld laser thingy having that funny lightsabre shape doesn't really help too
>>
>>2798705
For these precision things I use 0.2mm.

I've tried 0.1mm ones but they just don't work, layer adhesion at 0.08mm layer is piss poor(iirc the textbook rule is that 80% of nozzle diameter is the limit of where the print will at least keep its shape), the support turn into these stringy things you can't get off the model(and you have to add more of them because the maximum overhangs you can do are smaller) and you can't go down with layer size because from what I understand all printers do Z axis movements in 0.04mm increments and at 0.04mm the heat radiated by the nozzle deforms the print underneath to the point where a small amount of overextrusion will start blobbing up until it will inevitably make the nozzle break it off the build plate.
There are some solutions to this, but I couldn't be arsed to check them.
>>
File: 2024-05-12--22-10-28.png (1.83 MB, 1920x1080)
1.83 MB
1.83 MB PNG
>>2798719
>all printers do Z axis movements in 0.04mm increments
Well the thermal aspect is part of thin layers, yes.
But the Z resolution is a function of stepper steps, and leadscrew thread pitch (assuming its a leadscrew and not belt or something else)
Normal nema =200steps per 360 rotation
Typical 8mm diameter leadscrew pitch is 8mm, but sometimes 2mm iirc
So for a 360deg rotation of the nema you get 8mm of bed/nozzle elevation.
8mm/200=0.04mm of elevation per step
...which isnt necessarily the smallest you can do with that nema/leadscrew combination, but it is the smallest you can get.. i guess... reliably/accurately.
But yeah, ive used 0.2mm nozzles to print interlocking gears with crisp dimensions, and it sucked.
It did however help to print the first layer as thick as possible, and then use thinner layers after that.
Also PET tape was pretty good, much better than a textured plate. Of course you can really only use PET tape with PLA prints.

picrel: first time printing 0.8mm nozzle at 0.6mm layers... for speed
I think I like it.
>>
ive got two ender 3's with the marlin 2.1.2.2. one has a silent board 4.7 and "jolts" the steppers every time its powered on (did it before the board swap too) the other is 4.2 board and says media init failed when i take out the sd. they both crash occasionally when inserting SD.
is there a more stable version or marlin or should i just deal with the bugs. also want to know if any upgraded stepper driver will change the sound of the steppers or just the ones that advertise "silent". i like the sounds but still want improved performance.
>>
>>2798737
ew, marlin
ew, SD cards
Sorry, cant help with marlin any more now that I'm a klipperguy
SD cards are something you should try to move up from. Get a wifi dongle or something and use a print manager like octo/fluidd/mainsail etc.
But yeah, maybe marlin cringes so hard when it registers an SD mount/unmount that it dies. Sounds a little like an electrical/firmware interaction bug though.

The noise your machine makes is a function of the frequencies the motor is being driven at. Sometimes the signal is within the audible range, and/or things wiggle within the audible range as harmonics.
32bit boards and modern drivers tend to be quieter, and the whole TMC 'silent' thing really is almost totally silent.
Not sure what board the ender 3 has, but if its not 32bit then you may not get a lot of silence from silent drivers. But I dont actually know.

It is however very nice to have no sound coming from your motors. Mines all fan and bearings now. When i get around to switching out the Y bearings for polymer bearings it'll just be swishing and fan sounds.
>>
>>2798684
i got a bit more ideas for this
instead of drilling the entire nozzle to 3 mm, a highly conical shape would be preferred. not sure how to make at home. an alternative can be stepped 2-2.5-3 shape.
the thickness measurement and adjustment can be done within a single unit with a cheap simple microcontroller that hijacks the extruder step input and then correlates it with ribbon thickness and adjusts the steps accordingly on the fly. baisically, as far as the rest of the printer is concerned, you're printing regular 1.75mm filament, while the extruder does the thickness adjustment magic. it's a simple integer operation with a bit of memory to account for the delay between sensor and nozzle, should be possible to do on something as simple as attiny 85 for maximum cheapo. handling retracts should be easy, record backwards steps and ignore as many forwards steps before resuming normal operation.
for even maximum-er cheapo, maybe incorporate this as a thickness gauge?
https://jfs-agri.com/index.php/bubsbuilds-projects/just-playin-and-concept-demo-projects/flexure-fun/68-displacement-sensor-double-compound-flexure-based
probably incompatible with the thickness measurement idea above, but the ribbon will be a lot easier to push, both by the extruder and into a circular hole, if it's creased down the middle, which can be done with cheap pipe cutter roller against either a rubber roller or a solid v-groove roller on the other side. tested, both seem to work, but require quite a bit of force and can cut the would-be filament so it's a step to add to the "production" step when you're spooling the bottle up. that way thin filament can be pushed harder.
>>
>>2798706
ahhh sowwy ;_;
>>
>>2798762
nah, my bad, i thought they meant that the benchy was painted
..which would be even gayer than printing figurines ;)
(says the guy with printed tits stuck to the front of his printer)
>>
>>2798673
Go on, what didn't i comprehend?
>>
Is there any way to permanently set the flow rate to 110% on the K1C? With the FlowTech hotend I need to set it to that or I get bad bridging. For now I'm just changing it to 110% after the print starts, but it's annoying to have to watch it do it's little dance every time before being able to change it.
>>
>>2798852
Your slicer should have that setting
>>
>>2795242
"white" "petg"
are you fucking serious?
10KG?
100% sure that you'll regret it, hard
>>
>>2798956
Why?
>>
>>2798956
i need(want) to do some functional prints that should be color-matched closer to a beige-ish interior paint and exterior coated aluminium railing, neither of which are exactly white, but are closer to white than grey.
Most of what I print are prototypes and get tossed in the trash anyway.
I decided when i first started 3d printing; to recognise filament as a consumable, and that being precious about it only inhibits my ability to explore printing. So now I buy cheap filament in bulk, and intentionally 'consume' it as a means to enjoy designing and then making all sorts of random stuff that usually fails interestingly, and sometimes becomes a life-improving part of my existence. I love the process and the way its generally edifying in tangents to electronics, mechanical design and engineering in general. I wish I had one of these things when i had all my potentially genius ideas in my 20s.

TL;DR: After already using 60% of the 10Kg of this disgusting 'green' >>2798732 I bought not long ago, I regret nothing. So 10Kg of white can't possibly be worse.... right? lol
>>
File: shaper_calibrate_y.png (144 KB, 800x600)
144 KB
144 KB PNG
>>2799004
petg is imo harder to deal with than abs and white pigment is a cherry on top
2 things i avoid with at all cost are:
white petg and silk pla
those 2 such a great headache inducer

also i got my y axis problem fixed
squaring and de-racking the gantry really made some difference
>>
>>2799004
Why not get clear then? It never fits perfect, but also never bad.

>>2799009
>petg is imo harder to deal with than abs
The fuck what now?
>>
>>2799021
>clear
ehhh, i dont think it would be as subtle as i want it to be. In one case I also need it to be mostly opaque, to block the sunlight entering around the edges of a roller blind is it's purpose.
But mostly because my most convenient and cost effective supplier wants nearly twice as much for any 'popular' color. Which completely negates the whole "its cheap so its fine to just burn through it" approach.
https://www.aliexpress.com/item/1005006064018961.html
>>
>>2799021
>The fuck what now?
ikr
even if it didnt stink in a way thats dangerous to health and airborne-particle filthy, ive seen the contortions people have to suffer for ABS prints.

I do have a roll however, for when PETG fails me structurally.
>>
>>2799033
I really don't what you're even trying to hint at. For me PETG is just higher temp PLA in every printing behavior.
>>
File: IDC-4201C.png (101 KB, 800x800)
101 KB
101 KB PNG
>everyone prints on pei-covered steel plates nowadays
>every printer uses permanent magnets that hate being heated up to secure the plate and prevent usage of glass base, requiring probes and shit to compensate for warping
>every printer uses resistive heating to keep the bed warm, heating a lot of things that aren't the bed and requiring extra insulation
Now, hear me out. Inductively heated bed. A single-stop, elegant solution for all your troubles- glass's flatness with PEI sheet's convenience, and inductive plate's efficient heating and magnetism that magically vanishes once the power is out. It's not an off the shelf solution yet, inductive plates we have for cooking are too good at being heaters and not good enough at being magnets, and don't have precise enough control to keep a particular temperature while operating continiously to keep the magnet going, but I can see them being a big thing if they are adapted to printing needs.
>>
>>2798739
Marlin is supposed to support a thickness sensor, so if you have spare pins you might not need a separate microcontroller. Maybe the optical sensor can get thickness directly.
>>
>>2799080
The problem with that is that regular filament is round and ribbon-ified bottle is flat with fixed width, so former cross-section increases by square while latter is linear.
>>
>>2798794
>Go on, what didn't i comprehend?

>>2799044
>I really don't what you're even trying to hint at
>>
I want to 3D print a structure that will be immersed in ferric chloride etchant for weeks (it's a spacer for a pair of electrodes). Will PETG work for this? Or should I try to make the structure out of PVC bits from the hardware store?
>>
>>2799052
I suggested this a while ago here, got no responses. I think it would be a bit of a pain to construct, and expensive to make the driving electronics. Cheap induction hobs (like mine) use bang-bang control, but for such a thin piece of metal you'd want that either to be rather fast or to use continuous power variations. The heat will also be concentrated above the middle of the coil, so getting the temperature even may not be easy at all.

You'd probably be better with a radiative heating method, using infrared lamps/LEDs to shine through the glass and on the underside of the bed.
Or just use a thin-film kapton+copper heating element between the glass and the bed itself. Don't know why we ever put the heated bit below the glass in the first place.
>>
>>2799100
>ferric chloride
I'm guessing, but a quick wiki consult makes me think it looks like petg would be ok.
>>
>>2799093
Still no argument..
Here's one for you: your silicia gel dries up to coarse powder at continuous 200C. Now let's check upper PLA, lower PETG temperatures..
But sure, keep acting like a faggot with superiority complex.
>>
>>2799116
>silicia gel
where did he say silica gel?
>>
File: 2024-05-13--21-26-40.jpg (225 KB, 1622x1080)
225 KB
225 KB JPG
Are you silly girls still fighting over me?
squeee XD uwu rawr

I use far to many words so you buttclenching angrycunts wont """"ReAd"""" my posts. Stop it, now.

>>2797840
>It might actually be the white ?silicone" cement often used to pot PCBs or attach heatsinks.

image related: what it looks like
link related: (again) what i think it is (or high-temperature similar) https://www.aliexpress.com/item/1005006079288411.html
It sure as fuck isnt silver paste or (lol) silica gel
>>
>>2799052
>>2799102
yeh. For all the reasons already mentioned I wouldn't think this is a practical idea either.
But also mostly because a printbed *surface* is a CONSUMABLE item. You are going to need to replace it some time, if not fairly often.
Plus of course thinking you can get a magnetic locking force from an inductor thats being used to generate heat... seems sort of inefficient, or maybe just very impractical... i'm not sure which. Maybe it can do both simultaneously, but that feels illogical for some reason.
>>
>>2799102
>Don't know why we ever put the heated bit below the glass in the first place.
For flatness
>>
>>2799093
Yeah, yeah. Phone ate my words.
>>
>>2799100
if only you had some ferric chloride you could put petg in and test
>>
>>2795679
Yes.
>>
>>2799123
Your link is silica gel, faggot.
>>
File: adtm.jpg (204 KB, 910x1012)
204 KB
204 KB JPG
>>2799081
I see the appeal of doing it with a small microcontroller, but the extruder stepper driver is almost certainly on the motherboard and at least for me it's easier to flash a new Marlin that does the right math, than to intercept the TMC2208 STEP pin to add/remove pulses at the right time.
>>
>>2798657
>CNC kitchen
>Not annoying and gay YT personality
Hi Stefan.
>>
>>2799100
It's compatible, PETG is unaffected by ferric chloride. However, big caveat, most 3D printing filaments have other shit in them and PETG is one of the most inconsistent from one brand to the next. There's no guaranteeing the PETG you've got won't be affected. Only way to know is to test it, but odds are good that it'll be fine.
>>
File: IMG_20240513_101137_043.jpg (955 KB, 4032x2268)
955 KB
955 KB JPG
Filament grinding problem was really an extruder grinding problem. Shimmed it up to fix.
>>
File: IMG_20240513_114938_953.jpg (859 KB, 4032x2268)
859 KB
859 KB JPG
>>2799211
And installed a dual-z kit with a second stepper motor
>>
>>2799211
Nope, still not fixed. Fuck this shit. 3d printing is ass, what a waste of time
>>
>>2799211
get rid of that thing and mount a direct drive instead
>>
>>2799273
oh and make sure to tune the tension on the extruder spring
>>
>>2799130
A thin film heater with uniform thickness will inherit the flatness of the glass below it.
>>
>>2797875
>>2798052

You bit off more than you could chew.
Stop making it seem like it's not something directly due to your actions.
It's not inherent to you.
But making stupid decisions are.
Executively stupid decisions.
>>
File: 1702348355266030.webm (725 KB, 480x640)
725 KB
725 KB WEBM
>BambuStudio updates
>Lots of good new features
>Fans now max out every time a print starts and have to be manually lowered
>>
File: 2024-05-14--09-25-45.png (265 KB, 1145x613)
265 KB
265 KB PNG
>>2799162
the link
>>
>>2799211
I recently solved my bowden issue, you can too.
-those gears suck; need rounded contour for the filament to sit in
-anyhting that doesnt have two gears driving the filament; sucks
-filament doesnt like being curved. The way it comes off the roll (coil, with ~150mm diameter) is the path of least resistance, if you put it through a bowden that bends it in the opposite direction, or rotates it axially, smaller diameter curves, etc, or pretty much anything else aside from its free-hanging spiral.... it will resist.
-- There is a small sharp angle where the filament comes down ftrom the roll and has to turn into the extruder: major resistance.
-- it then gets pushed through that convoluted, small radius curves, bowden tube which also has zip ties adding small points of increased resistance

Ditch the bowden bullshit and find a way to get a direct drive extruder onto that rig and you will be instantly much happier.
Trust me.
I had a bowden setup for years, every day the filament would snap at the extruder... just snap due to temperature change as the sun came up every day... (old pla). Often I would find that the entire lenght of the (smoothly arching curve) bowden tube was filled with 1-2" lenghts of shattered filament.. thats how bad the brittleness was.
Every fucking day this shit pissed me off.
Finally frankensteined a direct extruder to it. Glorious success.
>>
File: 1699874315789194.jpg (823 KB, 1183x1192)
823 KB
823 KB JPG
>>2799362
>the link
Imagine if one could simply google "st922".
>>
>>2799370
imagine not being able to even conceive that what is clearly not silica gel is only being referred to that way because chinglish/translation.

Its a silicone based thermally conductive adhesive paste.
Not a crystalline silica moisture absorbent.
OBVIOUSLY.
Look at the product photos, read how its used.

What are you? babbys first A.I.? Or sematically-fixated sperglord?
>>
>>2799371
Be a faggot and reap what you sow. I tried.
>>
>>2799372
>reaps a simple fix for a few nozzles that leak

fuckyeah boiiiiiiiiiiiiiiii
>>
File: 20240513_210347.jpg (3.39 MB, 2992x2992)
3.39 MB
3.39 MB JPG
I hope there is no adverse consequences from this
I refuse to spend another $50 for a box that gets hot
>>
>>2799385
Did you just escape >>2792179?
>>
>>2799385
nice
as long as the temperature management is accurate it should be perfect. Do you know how often the controller polls/adjusts temperature?
If it was near my 3dprinter I would be tempted to use a spare thermistor to graph its actual temperature.
Just maybe put the filament in something to protect it from any possibility of dust/debris falling/floating onto it.
>>
>>2799401
one potential catch i can see is that the thermistor in the kiln may not be very accurate at the low temps you will be using.
Is 60C *ever* used in a kiln?
>>
>>2799401
>>2799403
It's an electric smoker, meant to control temps as low as 100f.
IDK how exactly accurate the temp sensor is but it's PID tuning is decent enough to keep within 5f of target, spool is reasonably close to the sensor.

I mostly worry about residue from grease burning off but I haven't seen any smoke since I pre heated it so it's probably fine
>>
>>2799410
>smoker
oh... ok
yeah grease is likely to be an issue
isnt there some kind of spray-on caustic cleaner you could use to get the big goopy bits off?
would the smoker tolerate being hosed out? pressure washed? otherwise its all by hand, so probably wont happen.
Still, should work ok as-is. Some kind of ventilated box, or fiberless heat resistant fabric wrap for the filament might be a good idea though.
>>
>>2799410
probably very inefficient for small batches of rolls though
a power monitor attached to it might be insightful
at some point the maintenance of filament becomes untenable
>>
File: download (9).jpg (167 KB, 1200x1200)
167 KB
167 KB JPG
I have an Ender 3 V3 SE and I’ve been using these 2-sided beds, one holographic and the other PEI. Do I need to calibrate every time I switch sides? Because I did an automatic calibration for both sides and the results were slightly different. I wanted to know if it’s normal and not necessary to calibrate for use
>>
>>2799612
Does taking the bed off and putting it back on again the same way make the calibration any less different?
>>
>>2799612
calibrate... what?
lets assume you mean bed mesh/nozzle height:
-it doesnt magically change thickness by flipping it
-the surface texture should have no significant impact on any calibration
-it SHOULD be perfectly flat from corner to corner in all directions, but its possible its a smidge thicker on one edge/corner, so putting it in flipped may result in a sliiiiiiightly different calibration. Even if true; this should not significantly effect your nozzle adjustment

It's possible you are failing to see that he difference is in microns, and not very many of them.
Or there is some grains of grit under the plate.
But most likely is that springsteel sheets are almost always slightly concave/convex depending on which side is up. The magnet on the hotplate usually pulls this into equilibrium with the edges, but is never going to be utterly perfect.
...so again; you are probably seeing a difference that is less than a very thin layer - like 0.1mm or less. Which is essentially irrelevant to your ability to print things.

TL;DR: flipping it shouldnt matter enough to make a difference. So you shouldnt need to recalibrate nozzle height every time. But its trivial to do a bed mesh at the start of each print, so if your OCD is really bad just do a bed mesh calibrate at every print start.
>>
>>2799751
oh ok thanks, it was the option leveling.
I thought that the difference between the numbers had a greater impact.
>>
>>2798738
>I'm a klippergay
ok bud
>>
>>2799781
well its hard to know, because the terms you are using to describe what 'leveling' you are talking about are completely vague and could be a number of things... and you havent shown us any of the actual numbers.
y'know... the basic information that is your responsibility to supply when asking for remote help.

>>2799802
UNlike homosexuality; you should try klipper before you assume the outdated thing you already use is better and therefore worth fanboying over.
>>
File: IMG_20240515_120955_065.jpg (2.69 MB, 4160x3120)
2.69 MB
2.69 MB JPG
What should I do about this stuff?
I've tried sanding paper
>>
>>2799873
man, its like you guys are just trolling us.
Stop trying to make us guess wtf is even happening in the photos and supply some useful information, or more context photos... ffs

>inb4 'try lern2 3dprint'
>>
>>2799876
I took off the support from the print and it left that residue and I don't know how to go about cleaning it.
>>
>>2799879
is that part of the support? Or is it part of the print?

Generally you try to get a small, sharp screwdriver in between the print and the support, and lever it gently in a bunch of places so the support pops off.
If that is support material you may be able to do that still.
Or get a large diamond file and try to remove it like that.
When I print something like that I usually start doing test prints of just that section... like boolean it off the rest of the print and isolate it... then do test prints to find out what support parameters work best. Or in this case; least bad.

Are you sure you cant print the model in some other orientation where the support isnt stupid gnarly?

(Show us what you are printing ffs)
>>
>>2799879
For this one just sand it. For the next time you print with supports, fix your slicer settings.
>>
>>2799893
I think I'll try sanding it but with the drill and that little disk adapter maybe that'll work.
I just used the default creality slicer support.
>>
>>2799873
Scrape it with a craft knife, or maybe use files. Fine stringing can be removed with a lighter, but I think your residue is too thick for that to work. Sanding is only really viable for small amounts of material removal on a 3D print since the plastic gums the sandpaper up, YMMV. Consider filler.
>>
>>2799913
I'm going to print it again with tree supports this time. I chipped a part of the piece by accident anyways.
>>
>>2799922
can you show us the model?
ive been leveling up on my support mastery a bit lately
>>
>>2799612
Probably because PEI is more bumpy so if it sits face down it will elevate the top facing smooth side. Bed mesh probably stays pretty much the same just ajust z-offset.
That being said doesn't the Ender V3 do load cell z-offset anyway?
>>
Intending to get the cheapest chinkshit printer possible that will actually work (with a couple of tweaks on my end if needed), I'll only be printing silly little things every month or so rather than making it a hobby.
I've seen the EasyThreed ones reviewed as "surprisingly functional for the price" (I think the K9 is the popular one), any more reliable suggestions?
>>
File: print.jpg (1.08 MB, 3072x1948)
1.08 MB
1.08 MB JPG
>>2799898
It looks like the supports aren't really supporting the print properly. Either they are somehow too far from the print or for some reason your extruded filament isn't getting laid down properly. Might also be not enough cooling so the filament is getting dragged along the nozzle instead of staying in place. Pic related is how it should look, the lines are pretty far from eachother but it's clean and nothing is hanging. Try using different slicer, like orca and check cooling.
>>
My room hovers around 45% humidity. How quickly do I need to use my PLA if I'm just leaving it out?
>>
>>2799279
Thank you very much, I overtightened the tension when reassembling after cleaning the extruder. Appears fixed 30min into a print.
>>
>>2799369
Appreciate the experience, thanks. Will add a dual gear direct drive onto the list. Hopefully the plug and play creality one is ok.
>$300 modded Ender 3 FML
>>
Question, I am rtunning V6 hotend on my ender 3, and other printers, however my ender 3 appears to like to piss me off

>filament started leaking on top of heater block
>stopped using printer for m onths because of this
>have 4 other printers
>decided to fix ender 3 and get it running
>do 3-4 prints
>clogs and leaks filament on top of heater block

I am confused, my 2 other printers using V6 hotends are perfectly fine for the past 2 years

so my question is, is there v6 style hotend or what hotend would I need to stop fucking around with heatbreak and nozzle pressed against each other tightening it, heating it up to tighten it again, and just have it fail again.
>>
>>2800044
NOte that even with direct drive, it is useful to feed the filament through ptfe tubes. This helps keeping the bends in nice large radius curves, and less grit gets attracted to the filament.
Just make sure the feed tube you get has a 3 or 4mm hole in it, not the tight 2mm that are normally used for bowden setups.
Bigger hole, less friction, but all the same curve softening properties.
>>
File: shelf.jpg (228 KB, 1152x2048)
228 KB
228 KB JPG
Well, I did it. I printed a shitty face frame for this shitty little bookshelf.
18g brad nailer works fine through PLA. You just have to make sure there are enough layers on the surface you're nailing through. 2.4mm of top layer was enough to prevent shattering.
I know it looks like shit. And don't make fun of my books.
>>
>>2800178
loctite
>>
>>2800207
This is fucking bizarre. Good job, anon.
>>
>>2800207
what's the point of that, just to hide the shitty edge of the plywood? i'd probably have gotten some timber strips instead of green plastic, but sure it does the job.
your books look unironically good, if you've actually read them you're probably pretty based
not like i'd know, i don't read.
>>
>>2800207
Where is "12 rules for life" by Dr. Jordan B Peterson, clinical psychologist?
>>
>>2800212
Yeah, it's to hide the end grain -- this is normally done with hardwood, as you say, but I didn't feel like buying any and I wanted to see if I could make this work. I won't be doing it again.

>>2800213
It's on the larger shelf below it, right next to "The SAS Fighting Techniques Handbook".
>>
Anyone know of any good joining methods for 3d prints? What im making is similar to one of those dog cones. Printed flat, and joins together at both ends. Want it to be pretty user friendly and want to make sure it wont come undone. Almost thinking of doing a pause print to add bra strap hooks. Maybe some type of snap button. The cheaper and easier the better. The product is for women so I dont think i can do dovetail type slide joints. Adhesives are also not ruled out.

This is the first product I have thought about designing for women and they love cute shit, but you do it wrong and they fucking hate it.
>>
>>2800207
I don't get the negativity. Looks interesting and could certainly work for a full shelf wall.
>>
>>2800234
May i ask what you're trying to sell? Seems weird to me jumping straight to product without ironing hows first.
That said, superglue or acetone are the usual ones. Use stubs for mor surface area.
>>
>>2800261
Im making sleeves for cups. Theres a little more too it, but thats basically it. Needs to look good, and feel good. My first prototype turned out looking like absolute shit so maybe its not doable.
>>
>>2800285
>sleeves for cups
So you're building a knitting machine? Jokes aside i really don't get this one. Are you trying to connect TPU to other stuff? In that case just heat might be proportional better than acetone.
>>
File: catcup.jpg (108 KB, 810x1080)
108 KB
108 KB JPG
>>2800287
So the plan was to take those paper sleeves they put on hot coffee and sell "premium" cute sleeves that are reusable. I think its mostly a dead end so far as there isn't a lot of mechanical adhesion. Could maybe try TPU cats or dogs, or hearts etc. Havent even asked a woman yet if it would be cute or not. Its probably out of my wheelhouse. Leave it up to a woman to make the design, and me to do the manufacturing. I cant do both so scrapping this idea I think.
>>
>>2800296
These would be laser cut too and have a wider band than shown with maybe some better spacing and design, but I dont think you can get that Tulle material to look good easily. Women would not be happy. First pause print though so that was cool. New 3d printing skill unlocked.
>>
>1kg of PETG net weight
>reel is 250g of PS
Why do they do it?
>>
>>2800332
I've started documenting my spool weights recently. Now that its so trivial to switch filaments on this direct extruder its really useful to know exactly how much filament is left.
My latest orders of pla and petg both have 193g spools.
I havent finished a cardboard spool yet, but that should be extremely light.
>>
>>2800332
You could just order refills.
>>
File: collage20240516110412.png (3.63 MB, 1900x1754)
3.63 MB
3.63 MB PNG
400/95/9
Time to call it.

>2800453
>>2800453
>>>2800453

You know what it is.


>2800453
>>2800453
>>>2800453


>2800453
>>2800453
>>>2800453

New bread, got going.


>2800453
>>2800453
>>>2800453



[Advertise on 4chan]

Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.