<Scopeuk>
looks like maybe someone hit ctrl+d whilst on a random line, can't see anything in the commit etc sugesting it is deliberate
<InPhase>
I guess rotate_extrude is so fast nobody noticed it was being done twice. :)
<InPhase>
kintel: If you catch the log, did you make a yyp vim error on that one? :)
<InPhase>
Maybe the commit should have been "remove degnerate polygons and add degenerate rotation". ;)
<Scopeuk>
Manifold has introduced too much performance, reign it in a little
<Scopeuk>
:P
crazy_imp has joined #openscad
<InPhase>
Fork it to Justalittlefold
<teepee>
haha
<teepee>
not really manifold though ;-)
<teepee>
is there some sort of guideline for system design or something like that? I think that https://imgur.com/siZLw4Q will work, but the screws are a bit awkward and it would be even nicer to not need support for the middle part
<InPhase>
teepee: Well there's the entire field of industrial design.
<InPhase>
teepee: Why is this 3 parts instead of 2?
<teepee>
yeah, web search finds lots of stuff but it's either software stacks or like really the whole system
<teepee>
the top part is white/transparent, the pilars will host a 8x8 led matrix
<InPhase>
Does the bottom need to be separable?
<teepee>
the middle part basically captures the esp32 which just slots into the rails before
<teepee>
if there's a different solution for the esp32, no
<InPhase>
Can it slide in from the middle buttom instead of the outside bottom?
<teepee>
that part yes, but how to fix it in place after that
hyperair has quit [Remote host closed the connection]
<teepee>
screw in a block stopping it from coming out again maybe
<InPhase>
Well snap fit is one option. Or you insert a hole for a tiny screw behind it.
<InPhase>
It looks like your complexity goes way down with 3 parts going to 2.
<teepee>
snap fit is not ideal, it's exactly against the cable force
<teepee>
indeed, so maybe 2 m3 nuts trapped in the bottom so it's easy to just add a stopper with 2 screws
<teepee>
that makes it more difficult to add the top/front part though, it's supposed to have no screws on the outside
L29Ah has quit [Read error: Connection timed out]
<teepee>
I'll grab some coffee, maybe that helps :-)
<Scopeuk>
hmm I wonder if you would make the outside case a single piece printed say on the narrow end with the electronics an a sled that inserts from the usb connector end
<Scopeuk>
could probably hold it with filament pins at that point if you didn't fancy designing snap fits or similar
L29Ah has joined #openscad
mmu_man has quit [Ping timeout: 264 seconds]
ecraven has quit [Quit: quit]
ecraven has joined #openscad
Guest63 has joined #openscad
Guest63 has quit [Client Quit]
mmu_man has joined #openscad
guso78k has quit [Quit: Client closed]
guso78k has joined #openscad
teepee_ has joined #openscad
teepee has quit [Ping timeout: 260 seconds]
teepee_ is now known as teepee
pca006132 has quit [Remote host closed the connection]
pca006132 has joined #openscad
<InPhase>
teepee: There are tons of variations you can do on the esp32 enclosure to trap it in place. On the outer pieces, another option is making it round and screwing the lid into place with large threads on the outside of the lid. :)
<teepee>
while I love the huge threads solution, that's not going to work here, the thing is supposed to optionally have more than one of the 8x8 panels, so round is not really possible
<InPhase>
Ah. :)
<Scopeuk>
I guess you could do a slide on top, two rails in lower print and two slots in the top cover
<pca006132>
btw, wonder if anyone know how to decide on the thickness of metal plate, for holding a thrust bearing with ~300KG of force... the ID should be 15mm
<pca006132>
if one is using tbb 2021, they should use GCC 13 or clang with libc++
<pca006132>
GCC 12 will leak memory
<InPhase>
pca006132: What are you meaning by "ID" there?
<pca006132>
inner diameter of the bearing
<InPhase>
I thought you were using bearing as a verb. lol
<pca006132>
oh,
<pca006132>
it is a "thrust bearing"
<pca006132>
the flat kind that takes axial load
<InPhase>
The problem is now ill-posed I guess. You have a plate of unspecified dimensions, with a 15mm diameter bearing on top of it, and you're putting, I assume you mean 3000N onto this bearing. Critically missing here are the geometries of the plate with respect to its supports on the outside, and how far these supports are from the bearing.
<pca006132>
indeed
<pca006132>
let me draw some nice figure with comic sans as comments...
<InPhase>
With that included, you're getting into standard civil engineering problem territory.
<pca006132>
hmm, it is actually a piston driven by a lead screw
<pca006132>
for the espresso machine I am planning to make
<pca006132>
because of the area (58mm diameter) and pressure (~9bar), it has to hold around 300KG of force
<pca006132>
maybe I should just order some thick plate and see if the thing fails :P
<InPhase>
Once you get into the range of calculations though, you have to not call 300kg a force. ;)
<pca006132>
true, it is better to consider pressure...
<pca006132>
feel like hydraulic is quite amazing
<pca006132>
btw I am thinking about DIY such a machine because I found that I can get stainless steel tube with inner surface grounded to 0.1mm tolerance really cheap
<pca006132>
so I can probably build something really cheap, and see if it works or how it breaks
<pca006132>
probably 1/5 the price of a good espresso machine, even after including the cost to buy cutters etc to use a very old mini lathe in a lab (that my friend said I can use, as they never use it)
<J24k35>
always amazing if you calculate the pressure within a 3d-printing extruder and realizing that this is in a region of 100bar
mmu_man has quit [Ping timeout: 264 seconds]
mmu_man has joined #openscad
<InPhase>
J24k35: That seems high. I thought more 10bar.
<InPhase>
Hmm. Maybe more like 40bar. Your number isn't so far off from my estimate I guess.
<InPhase>
I was estimating within a factor of 2 of how hard it is to manually press filament through.
<InPhase>
Of course when I went to type the numbers in, I forgot to divide diameter by 2. ;)
<guso78k>
bar=10^5 * N/M2, right ?
<InPhase>
Yeah.
<J24k35>
so we can estimate that an extruder can push with 30N ( torque 50N/cm and direct drive gear has a 4mm lever so that would be more like 100N)
<J24k35>
filament is 2.4mm² so lets say 10N/mm²
<J24k35>
or 10MPa
<InPhase>
J24k35: Yeah. I was estimating that it sometimes takes me something like up to 10N of force when I press the filament through manually. Although when hotter, it will sometimes flow much easier than that.
<J24k35>
on the other hand the capillary pressure is also extremely high
<J24k35>
.. I had my printer dragging itself through the room when the spool was blocked - so they can push hard if needed
<InPhase>
lol
<lf94>
Is OpenSCAD GPU enabled yet :V
<InPhase>
At this point I'm not convinced that would actually be faster. Have you tried the Manifold features in the master branch?
<J24k35>
InPhase a colleagues merlin hot end exploded (PEEK hot end) when the nozzle was clogged.
<J24k35>
You don't need a cooling fan as PEEK is not conducting any heat ..
<InPhase>
J24k35: Hmm. Well that sounds favorable for printing at higher temperatures.
<J24k35>
InPhase well PEEK will also melt above 300
<InPhase>
Yeah, I just noticed you can buy PEEK filament for $600/kg, and then you have to print it at 400C.
<J24k35>
I thought about making a ceramic hotend (already tested one with glas) But Titan has also very low conductivity
<InPhase>
Which ironically is scraping up there toward the melting point of aluminum.
<J24k35>
aluminum blocks are terrible from the expansion .. i used brass as the titanium and steel nozzle broke the thread in aluminum
<InPhase>
So the connector to the block serves two purposes, right. One beging to have a tube to keep the material in, and the other rigidity.
<InPhase>
s/beging/being/
<guso78k>
also aluminium has a incerdibly high surface tension '=D
<InPhase>
And the desired property is thus rigidity while not transmitting much heat.
<J24k35>
yes which is why the "throat" is thin walled
<InPhase>
However, it should be fairly trivial to create a mount that has a thin tube around the filament, followed by thin truss framing for rigidity. This should not transmit much heat at all.
<InPhase>
It only gets thick when you attempt to construct this by screwing one piece onto another.
<J24k35>
my idea would be a ceramic with a metal bushing inside .. using HF induction to get the heat in
<guso78k>
J24k35 shorted resistor wire loop to create the heat ?
<guso78k>
wow have not seem this before. does the light come from the heat or from electrons jumping back from the valence band ?
<J24k35>
in the end a heating cartridge or a solder iron is emitting light in the inside .. you also can solder with a 50W halogen bulb
<J24k35>
every incandescent bulb is emitting 95% IR and 5% visible light
<guso78k>
looks more like an accident :')
<J24k35>
You can use SMD resistor to make a line writer for thermo printer
<InPhase>
J24k35: More traditional are the SERs.
<J24k35>
i mean even a MPU can heat up to over 100°C without damage
<J24k35>
InPhase surface enhanced raman?
<InPhase>
J24k35: Smoke Emitting Resistor.
<J24k35>
a magic smoke .. don't let it come out
<InPhase>
J24k35: Yes. Empirically, the smoke is very important to their function though. If you let the smoke out, they don't work right anymore.
<J24k35>
We had a bass subwoofer tube on a party that started glowing inside .. but seems to be some ceramic coating as it doesn't smell or broke
<J24k35>
well maybe it was something else inside
<J24k35>
but the air flow felt quite hot
<InPhase>
My kid's razor charger mysteriously "stopped working", so for fun I hacksawed it open to find the smoked out part, and was able to forensically determine the output end has been shorted out by some water while it was plugged into a power strip rather than an outlet (from the directionality). Obviously my child had "no idea" what happened to it, even with the mystery demystified. :)
<J24k35>
however heating plastic is quite difficult when you print faster and don't have several inch long heater - i mean professional systems use coaxial oil heating
<J24k35>
capacitors are well known for emitting magic smoke
<J24k35>
but you need some salty water or Chinese safety distances
<InPhase>
Yeah. When I taught introductory electronics in a physics lab, I would just order a new batch of electrolytic capacitors each year, because they very reliably got fried by student error.
<InPhase>
Extensive warnings about how to do it right only slowed the rate of frying them by about half.
<J24k35>
they also make a nice bang when overloaded
<J24k35>
telling them how they explode may also peak the consumtion
<InPhase>
I also entirely gave up on multimeter ammeters, and just used panel ammeters, because of the high rate of guarantee that multimeter ammeter fuses being blown. The panel ammeters recovered better from their short circuits.
<InPhase>
s/that/of/
<J24k35>
use the magnetic field to measure..
<InPhase>
It was a similar rate of frying those to capacitors, but more work to replace the fuses. And hey, youngin's could use some experience with analog gauges anyway.
<J24k35>
or replace the fuse with a PTC
<teepee>
InPhase: this certainly needs more tweaking, maybe also attaching the matrix to the front part - https://imgur.com/a/QU2VlfE
<J24k35>
When i showed some pupils what happens when shortening a 12V cell .. they developed some respect connecting anything to it
<InPhase>
J24k35: I just searched, and apparently there are a lot of instructors putting PTC fuses in multimeters for this problem. :) Clearly I was not the only one. But the panels worked well for the purpose, and were cheap and robust.
guso78k has quit [Ping timeout: 250 seconds]
<InPhase>
J24k35: And accuracy was not too pedagogically important anyway at that level.
<J24k35>
every shunt is changing the accuracy .. i mean use a 200A and it will likely never blow (if they don't short a truck battery)
<J24k35>
but the difference between 20A and 200mA port is already quite a lot
<InPhase>
teepee: Very nice. :) Do you have brightness control? The 0 needs some anti-aliasing.
<J24k35>
teepee that is galaxy black ?
<teepee>
it's esphome, so needs a bit more setup
<teepee>
no, but similar, pearl blue from Das-Filament
<J24k35>
looks sleek
<teepee>
needs anti-anti-aliasing I think :-)
<J24k35>
found flashforge burnt titanium .. that i like to try
<teepee>
sounds dangereous :)
<teepee>
color changing?
<J24k35>
depending on light and view angle .. looks nice on the marketing pictures
<teepee>
ah, 23% off currently at amazon.de, now I'm tempted too
mmu_man has quit [Ping timeout: 255 seconds]
mmu_man has joined #openscad
TheCoffeMaker has quit [Ping timeout: 256 seconds]