math-and-meta

[heady mist]

It’s prob gonna be situational for me personally because I don’t see myself having a lot of mixed belts that go into the same machines (yet) and it seems a little counterintuitive to merge them only to smort them, but I can see myself using it if I’m space restricted or if I start bringing in mixed cargo by train for a production

[fierce ruin]

If you're doing a manufacturer line that takes 4 different lines from previous machines, you can either run 4 different lines all the way or just merge them into 1 and smort it.

🤷‍♂️

[ashen girder]

Greeny.. when's the last time you even played? 😉

[bold crown]

I'm coming back to the game after a hiatus - and can't find the online planning calculator I used to use, my favorite feature of it was that it would track all of the overflow outputs as well as the desired output in a neat little pane on the right. You tell it you want 10 motors/min, it would give you partial building amounts, but I liked being able to round those up and keep track of the excess outputs so that I could allocate them elsewhere. Am I crazy? I can't find it anywhere...sounds familiar?

Maybe wrong channel, sorry

[fierce ruin]

Not sure which one you mean, but I have always found it simpler to design lines that don't have cascading excess outputs.

[frosty owl]

Or don't smort it and just feed the machine
(Add a pog split if there are more machines)

[fierce ruin]

See THIS is what I would consider "not simple". 😂

[bold crown]

Example, I've got a HMF factory that produces what I need, and I'm happy - but there's a little excess coal/iron left over, and I crunch that into excess steel pipe along the way, and branch it off into a smaller factory to make whatever else I'd like - tying up the loose ends. Yes, I could just sink it - but I like neat little boxes lol.

and it grinds my gears rounding OC numbers

[fierce ruin]

That's why I produce in multiples that don't require rounding OC numbers.

For Heavy Encased Frames that multiple is 45.

So I make them in 45/90/135 etc.

[frosty owl]

You can check out the total in-out of your entire world (or single factories by selection) on SCIM if interested ^^

[wind spade]

may 2019

[fierce ruin]

If I need 60 HMFs, I would rather make 67.5 and deal with the extra 7.5 than worry about OC rounding for exactly 60.

[wind spade]

most of the time all 4 lines won't come in the same way. One or two resources will be made 1:1 or 2:1, some resources will overflow to next build, etc.

[fierce ruin]

If you're mixing builds that is valid.
If you're not, the smort manifold doesn't care what ratios things are in, it autosorts and autobalances for you.

[wind spade]

what I mean by 1:1 is that I have e.g. a constructor in front of each manufacturer that makes things like wire or screws

[bold crown]

exactly, same - I just wanted a more convenient way to track how I was allocating those extra 7.5 hmf's, and I had a damn tool that did it but it's dissapeared evidently

[wind spade]

I don't see why would I merge those onto one belt to just split it back again

[fierce ruin]

Just send them to storage?

[wind spade]

not to mention that doing mixed belt leaves me with way less room for expansion

[fierce ruin]

I mean... just build for production target from the beginning and you never have to worry about expansion.
But that is a playstyle choice.

[wind spade]

well it's kinda limiting even for first build

[fierce ruin]

🤷‍♂️ it's perfectly fine to never use a smort manifold.
But I feel many, many people will utilize them once they know they now work properly.

[frosty owl]

Any merging is limiting, it's all about wether its convenient in one place or not
Eg: screws aren't mixed often

[fierce ruin]

Screw and QW builds definitely won't want to do this given the sheer volume they require.

[oblique hollow]

"now work properly" ?

[frosty owl]

Just adding a copy-paste setting would get many's gears turning, imo xD

[fierce ruin]

If you recall OG smart splitters didn't have Overflow?

[oblique hollow]

oh that

[fierce ruin]

That's the function that makes it so thus trivialize mixing belts.

Basically you can assign each splitter to now pull one specific item off the belt and send overflow through.
Then it turns a mixed belt into the same spooling logic as a normal manifold.

If you have exact values it will never stall.
If you're worried about stall put an overflow sink on the very last splitter in the manifold and it becomes incapable of stalling.

So any mixture of items that is within belt capacity can be reduced to a single belt. Which saves space.

@frosty owl I am seeing where poggers could potentially fit into this in a larger scope of things.

But I need a nap, will dm you when I wake up.

[frosty owl]

I find them most convenient to handle "big" sushi lines (with lots of items like for different manufacturing facilities) or for my infamous non-overflow sushi (which just avoids the need of an overflow belt at the end of each line)

[ashen girder]

I'm imagining "just expanding" a factory that ends in one manufacturer. 😂

Like.. that's gonna be spaghetti no matter how you shake a stick at it, unless you're a savant at factory planning.

[bold crown]

Alright, back for round two on my calculator hunt - how about a calculator that is input centric? I tell it what's going into the system, I lay out the buildings and their recipes, and I can adjust the number of buildings to control the desired outputs?

[oblique hollow]

i mean, greeny's calc has that functionality to an extent. you can give it input items and set the available recipes. though you cant tell it to only use "x buildings"
that would instead be done by setting the amount of output you want

[bold crown]

yeah - that's what I've been using in its place

the functionality I'm missing is output as a function of building count - I used to use a tool that did exactly that. Adjust the count, route the extra output to wherever else I happen to want it to go

just like greeny's, but I could tweak the building counts and see the overall change in outputs. It also had the ability to have multliple factories as tabs, and route the excess outputs from one into the inputs for another. Export/import factories, it really was pretty sweet - which is why I feel like I'm taking crazy pills that I can't find it anywhere. Google images, nothing. I literally have a mental image of it and can picture it exactly.

[oblique hollow]

it might be outdated and thus got removed

[bold crown]

I suspect that's the case unfortunately

[wind spade]

The problem with building count is that the tool doesn't know which recipe you will use

5 buildings making stitched plates is way less than 5 buildings making bolted plates

[bold crown]

yeah, in the tool in question, recipe was defined in a right click menu for each machine

so if I had three assemblers each for stitched and iron, they showed up in the tool as two different "buildings" if that makes sense

[wind spade]

Yeah, it was satisgraphtory imo

Not updated since U2 iirc

I just thought you're asking for my tool to support building count production

[bold crown]

oh, gotcha - I mean, I'm using yours at the moment because it's the next best thing for me 🙂

I started playing in Update 4, and the UI for satisgraphtory def isn't what I'm thinking of, but I don't know - I'm chasing a ghost lol

[gusty minnow]

Hi, I'm running into some issues with aluminium.
I have 1200 bauxite, sending it to 10 refinery. Then I send the alumina solution to 5 other refinery. These 5 refinery are producing 120 water each (600 in total), and the 10 other refinery are consuming 180 water each (1800 in total), so I have 10 water extractors.

But often, my production just stops because the water pipes are full, and I have to flush, any idea why ?

[oblique hollow]

water extractor output is not really set. limit the input via a valve

while your scrap refinery is idle, the extractor happily fills everything up

[gusty minnow]

the water extractor is not limited to 120/min ?

[oblique hollow]

and if for some reason it is blocked, water will start to build up inside it

it is, but only if it never is blocked

[ashen girder]

I just dump the water into a bank of coal generators along with my spare coke. 😂

[oblique hollow]

also, are you using mk 2 pipes?

[gusty minnow]

no

[oblique hollow]

thats a problem

possibly

anyway: limit the extractor output via a valve

that should help

[gusty minnow]

I'll try that, thank you

[oblique hollow]

if there is still buildup, just remove the water by feeding it into coal generators

[fierce cypress]

but it shouldnt back up if all the machines are running?

[oblique hollow]

machines sip in intervals, not continuously

the water extractors have enough time to possibly fill the entire pipe and block the last refinery

aluminum cycles are a complicated thing

[fierce cypress]

yea ig

you should also probably put a valve coming from the last refinary too

[gusty minnow]

Is there a way to send only the water overflow to the coal generator ?

[oblique hollow]

yes: build a pipe upwards

[gusty minnow]

I already have this one ^^

[oblique hollow]

pipes at the bottom are filled first

[versed violet]

OR, you can set apart some bauxite refineries and feed them only the recycled water.

[gusty minnow]

Oh, yes, it should be easier

[versed violet]

You dream xd

[ashen girder]

Gravity!

[heady mist]

I am very much looking forward to wrangling with pipes when I start my big fuel plant (hopefully tonight) 😂

My first true permanent build of this save XD

[gusty minnow]

The valves after the water extractors seems to do the trick

[oblique hollow]

tadah

[sullen cloud]

only trust a fluid based setup after it has been running for several hours 😬

[gusty minnow]

Yep, I'll see at the end of the day ^^

[sullen cloud]

Does this mean that your VIP junction also works without pumps?

[oblique hollow]

not always funny enough

the pumps really do some work

[sullen cloud]

is there a specific reason?

Btw where can I find all of your pipe schemes?

[oblique hollow]

in my manual of course

linked in the pins

[ashen girder]

Only trust a fluid based setup that never fills up, and has been running for several hours. 😄

[thorn bane]

this destroys the whole appeal of sushi for me
feeding a manufacturer with 1 line
like how cool do those look

you keep talking about prog. splitters and i still have no idea what you mean
do you have a screenshot/example for me?

[oblique hollow]

never seen them?

[fierce ruin]

Basically it turns sushi into the manifold vs. balancer debate.

Both work, it's about aesthetic.

[thorn bane]

i know
but how cooool is that like come on

[fierce ruin]

Poggers are the second splitter you unlock under caterium.
They cost supercomputers to research and to make.

[thorn bane]

no i know what prog. splitters are xD
i wanna know how @frosty owl is using them to improve sushi

[ashen girder]

...wants to know how prog. splitters are used, shows screenshot that can only be done correctly with prog. splitter...

[thorn bane]

?

[ashen girder]

?

[oblique hollow]

settings not shown

[thorn bane]

that screeenshot doeesnt use prog. splitters?

[ashen girder]

That splitter's outputting more than 1 thing to the right.

The whole point of programmable splitters is a smart splitter outputting more than 1 thing to an output.

Oh, that's a merger isn't it.

[thorn bane]

#907847228188655676 message
whole image

and even then for splitters im using normal splitters and it works

[oblique hollow]

i dont see progs they all look like normal splitters

[ashen girder]

I misunderstood the screenshot he shared.

[thorn bane]

i know
i made it
theres no prog. splitters

[ashen girder]

You could pretty much get rid of.. all of that with a programmable splitter.

Like.. all of it. 😂

[thorn bane]

how?

[ashen girder]

Push those lines into a programmable splitter that outputs those things to the side and overflows?

I guess if your input ratios are wildly off you'd have to adjust those.

[oblique hollow]

these are ratio splitter assemblies

[thorn bane]

but theen you need 4 inputs for the manufacturer that was my point

[ashen girder]

Not if the inputs aren't wildly off.

[thorn bane]

more examplees im working on currently
using normal splitters

[ashen girder]

Yeah, if you're driving manufacturers from boxes, that's different.

Even then, you could do it from 2 smart splitters just as easily. 🤷‍♂️

And.... still replace all of that mess.

[thorn bane]

i dont get that
if the ratio is right you dont need overflow
it the ratio is wrong it will clogg even with an overflow splitter (if you only use 1 input for the manf.)

[ashen girder]

Correct. 2 smart splitters means you can service all 4 inputs separately and overflow back to the cans.

Instead of.. that.

(1st smart splitter filters input 1, input 2, overflow to 2nd smart. 2nd smart filters input 3, input 4, overflows back to can.)

And will literally never clog.

And doesn't require.. 13? 14 splitters and mergers? Just two mergers and 2 smart splitters.

2 more smart splitters if you want to sort it back into the cans afterwards.

[thorn bane]

thats my point i want to use only 1 input instead of 4
otherwise you can just not use mixed belts

[ashen girder]

And my point is instead of having one belt at the bottom, I'd rather have that huge pile of junk in the middle replaced with two belts.

[thorn bane]

like you have A and B merge it to AB and split to A and B
why not just use A and B that makes no sense

[ashen girder]

Because you can merge onto your one line in multiple places.

Because you can move the parts across the factory to the manufacturer instead of running 3 separate finished goods lines.

Because it takes up WAY less space.

Because it works better for adding more manufacturers afterwards.

[thorn bane]

ok i guess this is just preference
if its using more than 1 input per machine its not sushi imo

[ashen girder]

It is 100% just preference. I won't argue that.

And that's just arbitrarily changing the definition of things. 😂

[thorn bane]

oh btw
priority mergers can be build form simple splitter/merger arrays
ive used those in this screenshot from 2 months ago to backfeed overflow sushi
but you need a second input for that so i stopped using that

if you do 2 smart splitters into the foundry you dont even need the right ratio on the belt right?
it can literally be what ever if you sink the overflow
sorry but thats not sushi for me xD

[fierce ruin]

Meh.
I personally don't plan to just saturate with a failsafe Sink, so all mine will be exact ratios.

🤷‍♂️

Also the amount of keystrokes you and Tug could have saved if you just called them poggers.... 😭

[thorn bane]

nope

[ashen girder]

I played pogs. I refuse to use that word.

[thorn bane]

oh btw people have been doing this for ages from mixed trains
comes off as 1 line
use smart splitters to divide

[fierce ruin]

Yes.
They didn't work originally though because the overflow option didn't exist.

Also see my comment about "if this is known".

[fierce ruin]

Yes.

But with all 3 outputs just that face.

[silver tundra]

Thx, but need more info how. Hard to see on picture

[thorn bane]

https://www.youtube.com/watch?v=q4WCHx1W0KY

[silver tundra]

Cool Thanks 👍

[wind spade]

"progs" saves more

[fierce ruin]

It isn't poggers though. 😏

[wind spade]

even better

[muted crypt]

debating if this is ""too much"" for a t3/t4 factory. I'm tossing quickwire into it because it's close enough to where I'd be processing steel

[thorn bane]

seems fine
lots of awesome points

[muted crypt]

overclocking the one pure copper node with a mk2 miner will get me my 270 copper, and just putting the mk2 miner on the pure caterium will cover that

[thorn bane]

kinda hard to bank on having fused quickwire at that point though

[muted crypt]

for reference, I plan on building here on the map - those three coal nodes are the ones in the crater lakes. Some nodes are already being used here because I needed to get some basic steel

I only enabled recipes I already have

I lucked out and got steel rotor, encased industrial pipe, solid steel ingot AND fused quickwire

I have these alts so far:

• Biocoal
• Cast Screw
• Charcoal
• Cheap Silica
• Compacted Coal
• Copper Rotor
• Encased Industrial Pipe
• Fused Quickwire
• Iron Wire
• Pure Caterium Ingot
• Pure Copper Ingot
• Pure Iron Ingot
• Rigour Motor
• Seismic Nobelisk
• Solid Steel Ingot
• Steel Rod
• Steel Rotor
• Steel Screws
• Steeled Frame
• Wet Concrete

I've gotten super lucky with my alts so far.

[thorn bane]

ye ive waited (as usual) until jetpack and rifle in my u5 playthrough 🙂

[muted crypt]

fair

hmm... I may actually need to cut down on the quickwire/limiters

I don't really want to diverge iron from steel to make iron wire for automated wiring

[thorn bane]

iron wire 🤮

[fading wasp]

my game is crashing when selecting inverted ramps

[muted crypt]

known bug

[fading wasp]

ok

[muted crypt]

unfortunately just have to not use 'em until they're fixed

thoughts, @thorn bane?

I wouldn't really like to have the 2.5/min automated wiring, but... it's better than diverging more off

[thorn bane]

id use a second copper node

[muted crypt]

though, realistically... how much do I use AI Limiters right now?

[thorn bane]

dont you have 3?

[fierce ruin]

@wind spade nothing is better than poggers.

[thorn bane]

its mostly for awesome sink points
they are insane for early points to get all the new cosmetics

[muted crypt]

3 pure iron, 1 pure copper, 1 pure caterium, 3 pure coal, 3 normal limestone; can pull two more normal limestones from further away if needed

[thorn bane]

oh thats iron not copper i cant see sorry

[muted crypt]

lol no worries

okay so at this point in the game, realistically, I don't think I need even 10 AI Limiters

[thorn bane]

ye i guess then oyu have no choise but to make less ai limiters 😦

[muted crypt]

mhm

They'd only be used for smart splitters and power switches anyway

[thorn bane]

but think of all the
13800/min

[muted crypt]

10 AI lims in the sink per minute is 9200 points

are you

ah

[thorn bane]

thats with 15/min sorry

[muted crypt]

dw I wasn't following

am dumb, dunno why this is my favorite channel

[thorn bane]

❤️

[muted crypt]

so there's actually a net loss of 1000 points per minute if I choose 2.5 automated wiring over 5 AI limiters

(920 * 5) - (1440 * 2.5) = 1000

so if I want to maximize points earlier on, I'd want to take the slower speed of the wiring

oh, wait, 2.5 should be plenty for progression

[thorn bane]

i mean it all depends on how much stuff you want from the awesome shop
i still have only bought double ramps and ladders and im at t8 right now

[muted crypt]

I only need 100 for the second elevator payload, and I only need 750 total for the third elevator payload (100 adaptive control units)

and I'm sure it will take me more than 5 hours from second payload to third payload

so yeah 2.5 is fine

[thorn bane]

ye true

[muted crypt]

back to math

I'm now trying to decide if 1215 steel/min is excessive or not

would be 4.5 mk3 belts lol

[thorn bane]

thats alot xD
i think i prefer 2*270
i only did 1 this time and it was a bit rough

[muted crypt]

the alt for HMFs without screws is heavy encased frame right?

[thorn bane]

yep

[muted crypt]

let me factor that in for preparation later on

[thorn bane]

(which i still dont have grrrrrrrrrrr)

[muted crypt]

because I am willing to go hard drive hunting pre-jetpack/rifle

my preferred method is runners/basher anyway

[thorn bane]

ive found like 50 hard drives or something
no heavy encased frames

[muted crypt]

yikes

yeah I've gotten 20

oh dear that's a lot needed for 5.625 HMFs/min

though oddly enough the restricting factor was limestone availability

however, that assumes I only want maybe 5.5 EIBs

let's just... worry about HMFs later

I know once I unlock everything up to t6 and get a nice sized fuel plant going I will be rebuilding everything anyway, so it's up to date

[thorn bane]

HMFs are a good contendere for your first outpost
if thats sorta your thing

[muted crypt]

so I'll worry about it then

See normally I'd just build everything in one place, but I want to try to break myself of that habit

[ashen girder]

I'm addicted to doing HMFs and Computers together, I think.

[muted crypt]

if there was another copper node this'd be fairly simple

[fierce ruin]

The fact these numbers aren't even double-digits irks me.

[muted crypt]

it's a t3/t4 progression factory

ultimately it's temporary, but it will last me long-term

[fierce ruin]

You don't just build the whole thing for what it will be and connect what it can be in the moment?

[muted crypt]

no

[thorn bane]

i made 2 HMF/min and skipped ai limiters until way later xD

[fierce ruin]

¯_(ツ)_/¯

[muted crypt]

My progression factories usually don't look good

because I know they're temporary

[fierce ruin]

You do plural progression factories. That's our principle difference it seems.

I have a single spot that constantly gets remodeled for what is needed in the moment.
All others spots are built for what their mk3 miner/mk5 belt potential is.

[muted crypt]

I do one for each set of tiers - a t1/t2 factory, a t3/t4 factory, a t5/t6 factory. Normally I'd just do it all in one place and rebuild for t1-t6 once I hit t6 but I want to avoid that this time around.

[thorn bane]

imagine not having a main bus that never changes from t1 until t8 (only belt upgrades)

[muted crypt]

the horror

[fierce ruin]

Ew. Main busses.

Might as well just mod the game at that point. 😏

[thorn bane]

main bus is life main bus is love

[muted crypt]

you must've originated in factorio

[thorn bane]

nope never played it

[muted crypt]

that's the typical progression strategy there

[fierce ruin]

Main bus is the reason I want them to make a new truck station vehicle called a bus just to confuse the shit out of people 🙂

[thorn bane]

xD

[thorn bane]

i hate when people say that
they go and build a factorio style bus in satisfactory and then say busses suck even though they just didnt use SF ways of building it (like using smart splitters etc.)

[muted crypt]

I know more than one person who says the game sucks because it doesn't have blueprints

even though 1) there's a mod for that and 2) that kinda defeats the purpose of building a factory that's unique to the player

I see now why the calculator is wanting me to use the regular rotor recipe

it's using steel rod here and turning some of those into screws for the rotors

[thorn bane]

ye steel rotor is pretty bad resource efficiency wise

[cedar mica]

For rotors, its just a question of how much iron/steel, you want to use

[muted crypt]

with my current recipes, miners and belt tiers I can support up to 1215 steel per minute

[thorn bane]

its just a trade of of complexity vs resource efficiency
using steel rod -> def. screw and pure copper -> steamed copper sheets is just insanely big and complex but you are rewarded by it because it is the most resource efficient

[muted crypt]

so I don't really think that's an issue

https://www.satisfactorytools.com/production?share=QtQxKUQSENt4FpMnaiye

gimme some thoughts here

if I wasn't restricted by my copper I would 100% add more automated wiring

wait.... lol I just realized versatile framework isn't even on here

back to the drawing board

oh this works out fairly well

https://www.satisfactorytools.com/production?share=KAiYKuCTNEd1HQEkpcWb

I'd much rather have a higher count of versatile frameworks because of how many are needed

500 for second payload, 2500 for third payload

[thorn bane]

nvm im stupid

[muted crypt]

those are third payload, Zyr lol

[cedar mica]

Dont really need high production. Just some time

[muted crypt]

and normally I permit too much time

I want to get away from that, try to produce more

[thorn bane]

god you have so much steel its insane

[cedar mica]

Default production, 8.3 hours to make 2500...

[thorn bane]

which is fine

[muted crypt]

and I could be making so much more, that's the thing

I'm not going to, but I could

[thorn bane]

thats third elevator right? only need 500 for second

[cedar mica]

Just saying, 1 machine is good enough, if you have something else to do. Like hunt hard drives

[thorn bane]

just progressign from t4 unitl t6 should be atleast 8hours

[fierce ruin]

Or like... going to work. Sleeping also helps "speed up" perceived production time. 😛

[thorn bane]

nah thats cheating (and bad for the enviroment)

[muted crypt]

so are NFTs but you see people mess with them anyway

[fierce ruin]

Sleeping to let your factory run actually accomplishes something though.

NFTs do not.

[thorn bane]

also if your modl. frames. are running for the whole time you probably have a big supply for addition versatile frameworks
Just make sure to store the extra in a container
and then if you are done faster you can just use the extra 30 steel beams to make more versatile frameworks

[fierce ruin]

So you know how the magic oil number was 27 for plastic/rubber.

I found the number for and all-iron stitches plates line.
Unfortunately it makes the line beautiful, but it results in the miners overproducing 1 ore every 25,000 minutes....
And idk of that is going to bother me or not.

That means the miner would switch to yellow after 4.75 years of playtime...

UNACCEPTABLE 😉

[ashen girder]

I'd say just remove one every year, but then you're under. 😂

[fierce ruin]

Do it once every leap year. Should be fine. 😂

[ashen girder]

Add a sign: "Remove 1 ore on 2/29."

[fierce ruin]

Remove 84 Ore on 2/29

But yeah, lol.
I will solve for Frames and CBs tomorrow then make a thread about all the complex items you can get from a single resource.

And what the ratios are.

[thorn bane]

yo interesting math fact
you start with
then you calculate the continued fraction of 8703.69/1000 https://www.wolframalpha.com/input/?i=continued+fraction+8703.69%2F1000
that has a 99 so thats probably where the rounding is so you make the continued fraction 8+(1/(1+(1/(2+(1/(2+1/(1+1/(2)))))))) which is 235/27 the end result

[fierce ruin]

It also would only be yellow for 1 cycle, then resume and not turn yellow again until every 17 days. Lol

[fierce ruin]

I literally have no idea what you just said, but cool?

😂

[ashen girder]

Why do you need a continued fraction for 8703.69/1000?

[thorn bane]

because that is rounded to decimals so you want to find the closest fraction to it

[ashen girder]

Ooooooh. Neat.

[thorn bane]

basically youre aproximating 8703.69/1000 with 235/27 and some rest
and because you know that 8703.69... is an aproximation anyway you cut of the rest

[fierce ruin]

Huh. So that link will speed solve other 2 for me is what you're saying?

[thorn bane]

yes

[ashen girder]

Hmph.

[fierce ruin]

Saves me time. Thank you.

[ashen girder]

Wikipedia for Continued Fraction: Best_rational_within_an_interval

[fierce ruin]

Continued Fracture is the story of my right arm during childhood.

[ashen girder]

In case you want the math behind it. 🤷‍♂️

[fierce ruin]

Oh wait...

[pliant finch]

Hey guys, I'm wondering if my math is wrong
I have 15 refineries producing each 40m^3 fuel so a total of 600m^3 per minute
And a fuel generator has a fuel intake of 12m^3
So the math comes out that I have capacity for 50 fuel generators
But the problem is the power fluctuates and the feed seems to be lagging behind of the rate of the generator's speed of consumption even tho it should be just enough?
I left them run on their own hoping it would stabilise out eventually but I keep having power fluctuations

[fierce ruin]

Doing exact fluid math for generators will never have the result you want because fluids are fucky.

If you have exactly enough fluid for 50 gens.
Build 45 gens and call it a day so they never run out.

[pliant finch]

Alright, thanks

I'll build below capacity next time

[fierce ruin]

Just have an overflow hook to package and store/sink the excess.

[twilit axle]

I built slightly below on mine plus some buffers to store the slight excess.

I’m using a single oc pure node to do 20 heavy oil residue refineries to 16 blenders (diluted fuel) for 1600 m3 split 4 times with 33 generators running per line…. I think

Technically I had room for 1.333 more but it works out better

[pliant finch]

Should I use valves to limit fuel flow to exact imput rates or is that also fucky?

[twilit axle]

I didnt bother because of the buffer.

[pliant finch]

Oh right

In what situation should valves be used then?

[twilit axle]

There was a good post on reddit for fluids, let me see if i can find it

[fierce ruin]

When you don't want backfill.
Like for overflow hooks and after fluid buffers.

[twilit axle]

https://www.reddit.com/r/SatisfactoryGame/comments/jmqw9d/pipeline_infographs_fluid_update_version/

[pliant finch]

Thanks

[thorn bane]

thats pinned in this channel btw

wait thats like super outdated
use this one https://www.reddit.com/r/SatisfactoryGame/comments/obqjrd/ficsit_pipeline_plumbing_manual_first_edition/ (the one pinned)

[twilit axle]

I’ll remember the pin, on my tablet at the moment so a bit harder to check them

[sour wyvern]

5x refinery polmyer resin, 5x refinery HOR, 6x blender diluted fuel, 18.75 refineries rubber from polymer alt, 12.5 refineries recycled plastic, 7.5 refineries recycled rubber
Output:
0 fuel.
525 plastic
450 rubber.
awww yeah.

[thorn bane]

#907688977778212936

[sour wyvern]

yeah that's a better setup - I was starting from the old and completely wrong picture in the wiki

[twilit axle]

I need to figure out what i’m with the output from my HOR refineries.

[vapid gorge]

So I'm going back and forth between Turbo Electric and Turbo Pressure - Seems like the trade off is spending about 33 qwire and 15 Al ing or I could spend 50 Ngas per with Pressure motors.

Pressure motors sound pricey in N ... but then again you don't use N for a ton?

[thorn bane]

ye this is like the best way of spending your nitrogen

[vapid gorge]

That's the way I was leaning. Even with 3k on nuclear that leaves about 9k in the world... and everything else doesn't need a ton. Well Fused Frames 32 which isn't NOTHING...

[frosty owl]

Sorry, I missed the pings somehow
The poggers garantuee ratios on a sushi belt are maintained during a split

[thorn bane]

ye im gonna need a screenshot and detailed explanation
i have no idea what youre talking about

[frosty owl]

The left/right split for iron ingots is a balanced split. Though this is with only 2 items, any additional item would only mean adding 2 more ruels to the split (left and right) for the new item

Eg: imagine concrete and steel pipes for encased pipes instead of the ingots

Even better example with quartz and limestone for cheap silica @thorn bane
https://media.discordapp.net/attachments/558721941410807812/899438167453745183/Screenshot20211018-01101100000.png

[thorn bane]

what are those prog. splitters set to?

[frosty owl]

Quartz and limestone on each connected output

The cumbersome thing is needing to have higher tier output belts than input belt (from each pogger)

[thorn bane]

ah so the individual items (for example silica) go left right if you use a prog. splitter?

*go

[frosty owl]

Yeah, they maintain their ratio. You just need to merge the correct amount on the belt (via production or balancing)

[thorn bane]

does this work for all ratios? not just 1:1 1:2

[frosty owl]

Whatever the ratio is, the split guarantees that items are sorted evenly

[vapid gorge]

You have a sink at the end of the line for safety though right?

[thorn bane]

why does it need to be a higher tier belt?
wouldnt this work?

[vapid gorge]

Braver man than I

[frosty owl]

Only for silica

[vapid gorge]

I'm considering doing some line mixing for late tier products just to save on logistic space but if I do I'm having a sink handy XD

[ashen girder]

Careful, if you just do the sane thing and use smart splitters, they might call you a phony. 😉

[frosty owl]

Game mechanics details.. . If the outputs are the same or lower tier than then input, the split might be "random" or you might even get pure belts 🤦‍♂️

[vapid gorge]

Wait, line mixing without smart splitters?

[ashen girder]

Oh, don't mind me. I'm just a petty bitch. 😂

[thorn bane]

so youre saying this should work?
because that would be insane

[frosty owl]

Just be careful of throughput
If you chain too many mixed lines and then some production halts for whatever reason, it might all "clog" due to too much input with little consumption overwhelming the sink

[ashen girder]

And that's why I'll always do single input smart splitters instead, no matter how much line mixing I do.

[vapid gorge]

Line mixing w/o splitters is my idea of self flagellation to purify me of my sins.

[frosty owl]

Input MK4, output mk5? Yeah, no issues there

[thorn bane]

no just the setup in general
that i can split the 15:10:2:2 sushi line in 2

[frosty owl]

Just list each item both front and right

[vapid gorge]

As long as the sink line doesn't have more than a possible 780 u pm it's fine though right?

[ashen girder]

casually hides his 4 sink set up in his central storage 😅

[thorn bane]

adding a sink defeats the whole purpose
feeding a manufacturer with 1 line
like idk about you but this just looks awesome

[ashen girder]

I get that you like it, but if you turn around, I bet it doesn't look anywhere near as awesome. 🤷‍♂️ I like the whole thing looking clean.

[vapid gorge]

Oh it looks amazing. And figuring it out sounds like torture to me.

[thorn bane]

just hide it in a box xD

[ashen girder]

Instead you can just have one line run across all of those, and use smart splitters, and cycle it back to feed other things. 🤷‍♂️

[vapid gorge]

Well part of my solution is like that. I'm just hiding the whole back end of the manufacturers in the floor or wall XD

[thorn bane]

instead you can just hook up the 4 inputs normaly

[ashen girder]

Yeah.. that's running dedicated lines to every input.

[thorn bane]

can also take the sushi line for a walk like this xD

[vapid gorge]

And people thought my vertical refineries looked painful

[thorn bane]

also you only need this merger/splitter zoo if you do balance merging like i do
if you just make the ratio by adjusting how much you put on the belt like @frosty owl you dont need that many mergers/splitters

[vapid gorge]

Doesn't it require a lot of precise timing and linking belts in a specific order to do so?

[thorn bane]

just linking belts in a specific order
you create the right ratio by merging in specific ways

for example if you want 1:2 you just have a merger thats connected to 1xA and 2xB

[vapid gorge]

Pretty and painful ❤️

[thorn bane]

oh ye this is 100% unsuitable for actualy gameplay
like it probably takes 10x as long to build xD

[vapid gorge]

Are you doing fairly large production or just small factories?

[thorn bane]

this
could be replaced by 4 belts from the containers to the 4 inputs

[frosty owl]

Yeah, but that all depends on how much you drop on the same belt
Eg: you place iron ore for modular frames production on the belt, having it 3/4 full. The smelting takes most of the ore away and you can reuse the belt in the factory until it comes out with the frames and the overflow ore (1/4 of the belt full)
Then you repeat for another factory and end up with the belt half full
Repeat for yet another factory, your belt is full at 75% and "ready for storage"
Issue: if you rework your modular frames factory and the smelters stop functioning, your belt might be clogged by ore before reaching the other 2 factories, blocking new items from coming in, easily everwhelming the sink...

Tldr: just be mindful not to merger too much stuff, especially if far away before reaching a sink

[vapid gorge]

Oh yeah yeah. I'd only being doing merges for specific modular factories and end of lines. Like manufacturer recipes that total up far less than 780 pm.

[thorn bane]

hm the hiding in a box is actually a decent idea
maybe i make a factory where all the splitter/merger zoo is on a second floor underneath and you only see thesushi line come up

[vapid gorge]

Most of mine lines I run under the machines and belt up to them so all covered by floor. I have central logistic lines that lead in and out of factories under the floor with glass foundations above them

[thorn bane]

anyway thank you @frosty owl for explaining the prog. splitter i finally understand it now ❤️

[frosty owl]

You're most welcome~

[frosty owl]

Manifacturers love being fed by mixed belts
Most of those stacks of 4 belts can get reduced to a single mixed one

[ashen girder]

I definitely think I'm gonna use smart splitters for my next foundry/assembler builds, too.

[vapid gorge]

Oh yeah. I'm going to do my best to make my Ur Rod manufacturer line be as simple as pos

[thorn bane]

im never gonna use stuff like this in an actual playthrough xD
i go full time>looks/inefficient factories/main bus

[vapid gorge]

You can make buses pretty 🙂

[thorn bane]

but the whole point of a bus imo is to be time efficient and not to make it look good

[frosty owl]

"Clean" is surely subjective, but I think the reduction in beltwork can make for some clean floorplans (when you merge from machines instead of containers)
||Sorry for WIP||

[wind spade]

I'm still on #teamNoBus

[thorn bane]

🙂

[heady mist]

Quick question, MK2 pipes are "fine" as long as you're not working around exactly 600 flow rate, right?

[vapid gorge]

Is there a reasonable way to do big builds w/o buses?

[thorn bane]

you do a small build 100x

[ashen girder]

To be clear, I agree with that. I just disagree with trying to balance lines onto a manufacturer with a splitter/merger abomination and saying it looks nicer. 😂

[vapid gorge]

Ah so forget about space reduction. Gotcha

[wind spade]

Depends on what you mean by big builds. I usually have dedicated factories for each product, is that a big build?

[vapid gorge]

Hmmm ok for example a factory hub that does all 2100 ur on the map? I don't know how I'd do it w/o buses unless I wanted it to spread out worse than sub urbs

[frosty owl]

If you have lots of production inputting/outputting mixed belts along a small place, having a bus might make sense, but logistically speaking it is usually more optimal to have different dedicated lines rather than a central bus of lines

[ashen girder]

The concept of a bus just doesn't make a huge amount of sense in satisfactory: a line or lines where you split off material to build a factory that continues on to feed other things.

[frosty owl]

Well, if one really likes prime splitting and balancing...

[ashen girder]

All of my builds aggregate items onto single lines that then feed the next stage of production and overflow into storage/sink.

[wind spade]

Most of intermediate products are only used in next step, so just put that next to where the product is made. No need to carry 50 different resources around whole base on belts

[vapid gorge]

Ah yeah yeah . I think I'm planning on a mix of both? I'm making hubs with few or 1 dedicated item with multiple 'arteries' of resources

[vapid gorge]

Ah ok. Maybe my idea of a bus is slightly different. I just make veins of belts running through the factory going in and out as needed.

[thorn bane]

right now i have dedicated factories for steelpipes/plubber/copper sheets/automated wiring
those get onto the bus via a train

[frosty owl]

I think one would really start feeling the appeal if he were to make lots of buildings of only high-tier items. Then a single belt can manage multiple buildings' in/out and centralizing the main belts can become quite convenient

[wind spade]

Bus (reusing definition from factorio) is set of belts going in one direction, which you split ingredients from and merge products to.

[thorn bane]

my (double sided) bus

[ashen girder]

Yeah, that's fair. But I think most people think about busses and think about moving low tier stuff which is just not a great idea in Satis.

Er, no offense to the others here doing just that. 😂

More power to you. I colocate all of my full-belt intermediates.

[thorn bane]

ye buses get better the more lategame items youre making since the logistic stays const. while for dedicated factories the logistics scale with complexity of the item

imo its roughly around computers/hmfs where it evens out
but you dont see the true benefit until like nuclear and stuff

[vapid gorge]

Slightly dif. My raw resources are all like that. I think I have 2 buses in the SW corner of the map each with 22 mk5 belts incoming, but from there I'm planning to direct them so that there is as little back and forth in the body of factories as possible

[wind spade]

We had discussions about this and it doesn't work like this, but to each their own I guess 🤷

[thorn bane]

multiple belts =/= bus

[ashen girder]

I feel like that comes from factorio where you want to build the processing separate from the extraction.

You're just eating frames doing that in Satisfactory.

[vapid gorge]

Yeah I think my 'buses' stop being buses once they get to a hub. Then it's just a logistics core

[frosty owl]

Ores (and low tier-item) buses are overall "belt-inefficient" and hardly give many benefits (though they do give a certain cool vibe)
@vapid gorge This is why I suggest having "dedicated" lines for ores and manage as much of the low tier products locally as possible

[ashen girder]

I can't throw too many stones. I ship quickwire and aluminum scrap by train. 😂

[frosty owl]

I once had a main bus with quickwire on it

[ashen girder]

At least I put it in stack-based logistics instead of belt-based. 😮

[vapid gorge]

Yeah I see that, I think I derive a lot of pleasure from just taking everything within a 'reasonable' distance and planning it out that way? It's like 'great, what can I do with these 12 iron lines'

[thorn bane]

i have screws on my main bus
actually used it alot for the time where i made HMFs and Computers usign the def. recipe

[vapid gorge]

Youthful foolish days gone by?

[ashen girder]

1 stack of rods turn into 1.6 stacks of screws. Bad juju.

[frosty owl]

The sexyness of 1:1 quickwire feeds shall reign supreme

[vapid gorge]

I've got a 6300 qw pm factory planned out. But the belts are just going straight next door XD

Or ... depending on looks... straight up

[thorn bane]

ye my plan is 4532 qw on my main bus

[ashen girder]

Holes are the best thing since conveyor lifts.

[vapid gorge]

Nuclear?

[thorn bane]

just for everything including nuclear

[frosty owl]

Speaking of nuclear... I'm surprised I took so long to notice you need just as much silica to make X Uranium Fuel Rods as you need to make the Plutonium Rods needed to sink the waste the former would produce

[frosty owl]

Coincidence? Think not

[vapid gorge]

Damn lizard people. With their monopoly on silica

[thorn bane]

10 per nuke

[frosty owl]

Yes, 5 per side of the production

[thorn bane]

thats a suspiciously even number

[frosty owl]

Dunno why anyone would complain about weird ratios

[vapid gorge]

5 being a famously even number yes

All this bus talk has reminded me to back up my spread sheet

[frosty owl]

Or make the floor 4m high and carve tunnels for beltwork between the machines 👀

The 3 manif on the top left have their belting completed btw. UFR, with the cells being made behind

The belts going up are purely for FactoryCart™️ purposes

[thorn bane]

xD

[frosty owl]

@ashen girder Compact enough for you?

[ashen girder]

Looks good to me. Did you just, like.. deconstruct all the producers? 😮

[frosty owl]

A mod that disable any "static" factory part from rendering

[ashen girder]

PURGE IT ALL

What's up with the diamonds? Are those assemblers?

[frosty owl]

Kind of a bug, but makes for fun screens ^^

[frosty owl]

The moving parts of assemblers, yes 😆

[ashen girder]

I couldn't remember what they looked like in action. 😂

[frosty owl]

Dunno why it renders cables too though...

[ashen girder]

They render dynamically when they first hang, so probably because of the mesh type.

[frosty owl]

A on/off comparison can be found #screenshots message

[ashen girder]

Rails are on there too which is weird.

[clever moss]

don't they kinda stretch when being built

[frosty owl]

True that

[ashen girder]

Yeah, I guess they do. 🤷‍♂️

[loud dune]

I feel like that comes from factorio where you want to build the processing separate from the extraction.
You're just eating frames doing that in Satisfactory.
Ores (and low tier-item) buses are overall "belt-inefficient" and hardly give many benefits (though they do give a certain cool vibe)
Okay so... I buss like the game is still factorio, putting on things like iron ingots, steel ingots, coal, concrete. Is it more efficient to use a main buss style of building, or building everything as close to each mining node that you can, and only import major items to the center base?

[frosty owl]

Making as many items as possible close to the nodes can cut simplify the logistics needed by a lot
But there are many ways to maximize logistic efficiencies once you have access to enough alts and transportation options

[icy sun]

if i use SatisfactoryTools to plan a factory, but then want to have more constructors, etc to underclock for better power, how can I calculate what the new power consumption will be?

[thorn bane]

if you build your machines at 50% you will use 66% of the power
math formula is power = OC^0.6

[icy sun]

I think i'm confused. 66% of the power at 50% clock speed, wouldn't that mean two machines at 50% consume 132%?

I thought underclocking yielded less power usage

[sand epoch]

The per rod is not energy required :/

[thorn bane]

total power 66% instead of 100%
so you save 33% power

[icy sun]

but at the cost of producing half as much?
1 Iron Plate Constructor at 100% makes 20/min and consumes 8MW
So 2 Iron Plate Constructors at 50% make 20/min and Consumes 10.56MW?

[sour wyvern]

..... it is PER ROD

[thorn bane]

no if you do the same amount of items
(hence the energy per iron rod)

[sand epoch]

He's fixated on the wrong stat

[thorn bane]

so 2x iron plat const. at 50% would take 8 * 0.66 = 5.28MW power

[icy sun]

i am a man of small brain

[sand epoch]

2 @ 50% is 2.6mw..

[icy sun]

gotcha, 66% of total power. ty for the clarification

[thorn bane]

oooh ye sorry 4MW * 0.66 = 2.64

[icy sun]

alrighty, could I get some help with the equation? Let's say I know my underclock percentage and how much power in MW 100% would draw. How exactly do I calculate to find the total power at that underclock percentage?

4MW at 100% and I want to underclock to, say, 32%

[thorn bane]

math formula is power = OC^0.6

[icy sun]

and then 4MW*.5?

[thorn bane]

yes

[icy sun]

which means that the total power to produce the equivalent amount, if all machines were at 32%, would be 2MW?

[thorn bane]

yes

[icy sun]

cool

[thorn bane]

not really
9/10 times building an additional coal generator is faster :P

[sand epoch]

Wouldnt just building it and reading the display be easier?

[icy sun]

I mostly just like to know how things work one layer past what I really need. Plus this way I can fiddle with things in a spreadsheet or paper when i'm away from my PC. I'm still early game and wanna know ahead of time whether or not I'll need to get more power built

[sand epoch]

Better off accepting that you need to pre build big power systems that dont need to be managed and expanded later.. lol.

[icy sun]

no doubt! 😛

[wind spade]

2.78 constructors = 278 total %, so e.g. 2 at 100% and one at 78%. Or 8 at 34.75%

[vast jungle]

+or just 3 constructors at 100%... one of them will switch off from time to time 😉

[icy sun]

that's a faster way than I was calculating the underclock speed, ty

[wheat pasture]

can someone explain the underclocking again, I don't really get it. Because I don't know what to look at, if you don't have to look at the items per minute

[vast jungle]

lets say you have a machine producing X amounts of items per minute... but you need X/2 for your next production step

instead of the machine switching on and off all the time because of blocking, you set the over-/underclock value from 100% to 50%... and now the machine produce only X/2... (and use less than half the power it needs for 100% output rate)

[thorn bane]

is there a way to limit a belt to 333.33333?

[vast jungle]

I would say its possible but difficult... because we cannot put 1000 on the belt and divide by 3 😉

[thorn bane]

hm this should work
split the line into 3 x 120
then backmerge the third 120 with 26.666 so only 93.333 comes out
120/26.666 = 9/2 so you 3-split 2 times and then merge 2

@frosty owl i wonder if this is usefull for sushi in anyway

[fierce ruin]

@frosty owlHi,
Have you attempted to feed more than one machine with a sushi belt without smart splitters?
It sounds like it could be an interesting challenge.

[vapid gorge]

Nightmare fuel

[fierce ruin]

You are truly an artist.

[frosty owl]

Not for what I normally do 😅
I usually prefer fiddling with clock rather than balancing to cut down on the belting required xD

[frosty owl]

There's actually a scenario where that is a thing...
3-way splitting a 1:1 sushi belt! Ahah
Eg: you have a belt with caterium ans copper ore (Ct-Cp-Ct-Cp... on the belt). If you split this 3 ways with a normale splitter you get:
Ct left; Cp middle; CT right
Cp left; Ct middle; Cp right
So the split is "self balancing"

Note: this requires the order of items on the belt to always be the same repetition of items, one after the other (ie: not Cp-Cp-Ct-Ct)

[crystal charm]

when doing math without meth

do you decide how much per minute of an end item you want and backwards calculate what you need?

or figure out how much per minute you can make based on a few local sources of nodes

[frosty owl]

I usually do the former since I prefer having excess ore rather than other products

[river night]

i'll usually have an idea of how much i want to make, then check locations and available resources and perhaps adjust that requirement to match

so.. both?

that usually also contains including or excluding a variety of alts to shift resource use between different types of resources, as needed

[frosty owl]

Right, don't fixate yourself on one way over the other, be flexible ^^

[thorn bane]

i just follow this

[crystal charm]

my nuclear pasta is well on track, only about 50 of those left

the assembly directors sitll have like 3900 to go and that's slow as hell at the moment, i'm building a rail line to a new spot to start a system for thermal rockets

[fierce ruin]

Yeah, I figured that out after thinking on it for a while.
Also, I'm thinking that merging 2 1:1 belts will give you ...AABBAABB..., and splitting that would get you back to 2 1:1 belts.
And of course, splitting a 1:1 belt would give you two separated belts, you just wouldn't know in advance which would be which.

[frosty owl]

Exactly 😆
Except... If you merge 2 identical belts, you do get ABAB, not AABB ^^

[fierce ruin]

I think the only way to get ..ABABAB.. from a 2-1 merge is to merge ..AAAA... with ...BBBB...

[cedar mica]

Why does it matter how a belt merges things?

[fierce ruin]

Depends on what you want to do.

[crystal charm]

cause if you're gonna make sushi belts, how they merge things can become important

although, i'm of the opinion that the simple use of smart or program splitters + overflow fixes that

[frosty owl]

If you plan to split with normal splitters, that is

[cedar mica]

Ah, 1 belt for multiple things... In that case, just put a overflow to sink at the end, so it keep moving

[crystal charm]

overflow to sink is not always great

[fierce ruin]

That only works if you use smart splitters to feed the machines.

[crystal charm]

depends on what you are belting, i.e i'm belting all the components for the assembly director system, i sure as hell don't want to waste circuit boards or computers into a sink, so i created a loop of overflow smart splitters, the items just keep circling around, only entering the machines if required

[frosty owl]

Oh yeah, I misunderstood
Merging 2 sushi belts can get weird results (for a normal splitter's perspective)

[fierce ruin]

Otherwise you need to make sure that the machines input doesn't back up.

[cedar mica]

You cant really do normal splitters for mixed belt, as even perfect ration, will get screwed, if they stack in different sizes

[frosty owl]

There are specific cases where you can @cedar mica

Btw, this sort of "perfect sushi" is only useful if you want to feed your machines without having to smart split the items for each input again (and no overflow)

[cedar mica]

Just saying, if you are mixing wire, with iron plate, you can end up with 1 iron plate blocking, becase internal storage is full, as there stack size is different

[frosty owl]

If they're not consumed and produced in the right ratio, that's bound to happen, yes

Probably even if they had the same stack size

[cedar mica]

Even if produced in right ration, there is still a different between machines, in the ration they want

There is a difference between making 10 a minute and 1 every 6 sec

[frosty owl]

That is very adjustable though, especially once you consider bigger numbers of machines
The example from before refers to some very specific (and limiting) scenarios, but once you bring poggers into the mix, that's when you can start sushing for real (still without overflow)

As long as all the machines involved are in the same grid too, the system won't clog even if you cut the power

[cedar mica]

If you can pull it off, then please make a video about it, so the rest of us can easier follow

[frosty owl]

That'd be quite the hefty upload 😅 😆
If interested, I have posted quite a few pics of the "Uranium Sushi Rods" factory I made, can be found via search

Spoiler: I like machines to be as load balanced as possible, so the factory (~350 machines) reaches 100% in ~10 minutes

[thorn bane]

god i feel stupid
im making an outpost for recycled rubber/plastic, so i looked at my planner how much recycled refineries the website gave and calculated demand from that, so 1150 rubber 650 plastic.
Now thats 1800 together so its perfect for a 600 oil tripling build so i made the whole plan build the whole factory.
Only to realize i didnt account for residual rubber.
so now im short 200 rubber
and have to make a new oil extractor just for that
god fucking dammit

[silver tundra]

Have you a link? EDIT: Found it, but a Picture or video would be better if possible.

[frosty owl]

Urgh, upload speed go brrr...
I don't even clearly get what should I record, wouldn't be easier to just pass around the Heavy Sushi Frames save?

#screenshots message

[void owl]

has anyone ever tried doing a fractal system for smelters/constructors?

[wind spade]

wdym by that?

[void owl]

do you know what a fractal is

[wind spade]

yes, but I have no idea what do you mean by fractal system for smelters

[oblique hollow]

you cant really do infinite repeating patterns?

[void owl]

so, imagine you have a splitter that goes to two smelters on either side. Then you take that, rotate it and take a copy of that mirrored with a splitter in the middle. Rinse and repeat

[oblique hollow]

aaaand what about the outputs?

[void owl]

you could do the inputs higher than the outputs

[thorn bane]

oh imagine
240 alumina solution -> alu scrap making 140 water
-> 140 alumina soltuion -> alu scrap making 81.6 water
-> 81.6 alumina soltuion -> alu scrap making 47.6 water

[late holly]

seems like more of a tree than a fractal

[void owl]

i say fractal because it is self-referential

but tree might be more accurate

[oblique hollow]

is it truly a fractal if its the same size everytime.
or is it just an regular irregular tessellation

[void owl]

i made this to demonstrate

[oblique hollow]

seems more just like geometric tiling

[wind spade]

looks cool, but definitely isn't a fractal

[void owl]

hmm

regardless of definitons, my point was has anyone tried anything like this

[wind spade]

no idea. Seems like it's a lot of work for no real gain

[oblique hollow]

funny pattern but thats it really

if anything people prefer to just do large elaborate balancers over entire machine patterns

[void owl]

my concern is that if you have a ton of the same machine in a row, it can take up too much horizontal space whilst leaving a lot of forward space unused, and getting to the other side can take a lot longer than if it was compacted like this

[wind spade]

make two/more rows

[void owl]

yeah, which is a kind of tesselation

[wind spade]

even one row is that 🤷‍♂️

[void owl]

my image was not to show an exact format but an example

[ashen girder]

I just do long rows. 😂

[wind spade]

  X  X  X  X  X  X
  |  |  |  |  |  |
--S--S--S--S--S--S
  |  |  |  |  |  |
  X  X  X  X  X  X

easy two rows

no need to do weird shapes

[ashen girder]

Maybe do two sets of those if you want something not quite so long. I'll stack'em sometimes too.

[thorn bane]

nothing wrong with long lines

[late holly]

i think if you want to make perfectly square builds this would work well

[void owl]

in that you could instead have

X  X  X     X  X  X
|  |  |     |  |  |
S--S--S--S--S--S--S
|  |  |     |  |  |
X  X  X     X  X  X

[late holly]

but would be complex to work out

might be nice to do a big borg cube of it

[void owl]

good idea

[icy sun]

a world where all the factories are big floating ominous black cubes would be sick

[wind spade]

how does that help? it's more space and worse logistics due to input being in the middle

[void owl]

i'm not one where i like to rely on overflow

[eager solar]

Just use greeny's model, with a main input and output lines perpendicular to the double row of machines

[void owl]

balances faster, and you'd have it be a widthweize build rather than lengthweise building

[wind spade]

it's not that much faster though

and width/length is irrelevant, just turn my setup 90 degrees and it's the same lol

[wind spade]

e.g. for 20 machines @ 30 items/min, it's 17 minutes to fill instead of 20 in my case

[void owl]

depends on the belt speed, the ends on the first have 1/3^n chance of getting an item, wheres in the other they have a 1/2*1/3^(n/2) chance of getting an item, if you have tier 2/3 belts thats definitely a massive increase

[wind spade]

I've had this discussion a few days back, belt speed doesn't really matter when talking about manifold fill times

except for very special cases when belt speed = machine input requirement

[void owl]

if you have a splitter with a tier 3 input belt and a tier 1 output and a tier 3 output, doesn't the tier 3 get more items out?

[wind spade]

that depends on how many items are going on the belt, but yes, the mk3 will get more items out. But that's pretty much irrelevant for manifold fill time

using slower mks for side belts has even proven to slow down the filling

[void owl]

sometimes you want the end machine to get items as fast as the first machine gets them

[wind spade]

well if you manage to do it, then it's a balancer and not a manifold 😉 manifold is where you have to rely on the overflow

[void owl]

you can mix the two

[wind spade]

and yeah, it depends what you want from your setup.
As much machines working as possible at start? Use mk1 sides, but you'll have longer time before the whole manifold starts working at 100%.
Fastest fill (working at 100%)? Use mk5 sides.

yes, you can mix balancers and manifolds, but in all the cases you've shown, it's always a manifold (even the middle-fed one - it's just two manifolds)

[void owl]

so you're talking about end game

[wind spade]

I'm talking about anytime

you can replace "mk5" with "highest tier available at the moment"

[ashen girder]

I don't think there's a meaningful difference between balancers and manifolds with mk1 belts. 😂

[void owl]

a smelter taking in 30 items per minute for its production, i'd like every machine to be doing that immediately

[wind spade]

then make a balancer

[void owl]

kinda what my tesselated machines example are

[wind spade]

sure, but you can also just do this

instead of placing them in that weird shape

you can then "manifold" the output easily as well

[void owl]

i am aware

[ashen girder]

People always talk about balancing inputs. Why nobody talk about balancing outputs?

[thorn bane]

oh boy do i have news for you

[void owl]

but i need to use 16 smelters for my overclocked mk2 miner, and if i were to do them in a line that would be a hella long line

[ashen girder]

Anyway. My trick, @void owl, is to just connect them to the miner and power before connecting the outputs. Usually by the time I'm ready to actually feed the next step they're all full already. 🤷‍♂️

[ashen girder]

It.. would be 8 machines long. 😛

[void owl]

8?

[ashen girder]

..yes?

[ashen girder]

I really appreciate that you literally just flipped it.

Or rotated it. Whichever.

[thorn bane]

tessereacted it

[wind spade]

flipped. twice. Once horizontally before I realised that did nothing

[void owl]

480/30 = 16

[ashen girder]

Well, that's the same as rotiating 180 degrees.

[wind spade]

also if you think 8/16 machines is a long line, you won't be able to play lategame lol

[ashen girder]

16 / 2 = 8

[void owl]

ok you're talking about how long it would be if you added another side to greeny's original image

[ashen girder]

Why would you run it one-sided like that?

[void owl]

some people do, and its natural to continuously expand the line in the same direction

[ashen girder]

I firmly disagree in it being "natural". Especially if you care about it being balanced. You also didn't actually answer the question.. if you care about it being 16 smelters long, you can just fold it in half and make something 8 smelters long, and it lends itself to being balanced just as easily that way. Probably more easily, honestly.

[wind spade]

it's pretty hard to "expand" the line in any way other than doubling it if you want it to be balanced

[silver tundra]

Thanks for that

[wind spade]

hence why people like manifolds - they can just add one or two machines at the end and it just works

[ashen girder]

Also because trying to balance a 780 line sounds futile.

[thorn bane]

speaking of balancing
is there a nice way to split a line into 16 vertically?
im building nucleear plans right now and i have rows of 8 facing each other and i dont want radiation
and if i split it horizontally i dont have enough space in my row

[wind spade]

I like how 90% of issues with balancing are due to "I've built the buildings without thinking about belts and now I want to connect it but don't want to redo the buildings"

[ashen girder]

"How do?" "..string splitters along the inputs like a sane person." 😉

[wind spade]

"I have 4 rows of buildings, I want to split my 3 belts into them properly"
"well maybe next time build one row of buildings for each belt instead?"

[hollow meteor]

Might be hard to see but this splits 12 vertically and then x3. the total array is 1:252

[thorn bane]

hm since i only need 16 i think ill stick with something more vertical like this
but thank you anyway 🙂

[frosty owl]

Plenty where that came from~
I feel sorry I don't have any photo collection or tag to ease the search 😅

[frosty owl]

I totally didn't just finish figure out how to feed 2 sets of AI limiters assemblers with copper sheets without making 3 sets of refineries instead of two 👀
Turns out I could have just 2 refineries with the same clock and split the output of one to 1/6 to feed the smaller set of assemblers (a single one)

That's acceptable balancing imo

[ashen girder]

Y'all are crazy with your ratios and stuff. More power to you though.

[frosty owl]

I like to rebuild stuff a lot before finalinzg a design, so quick boot times for my factories are important to not take forever in making them ^^
The factory should reach 100% in 15min or less if possible

[ashen girder]

Just keep an isc full to flood it. 😁

[frosty owl]

Sure. One ISC for Crystal oscillators, one for beacons, one for ECRs, and then some more for steel pipes, quickwire, silica, iron pipes and plates, wire, ingots...

[ashen girder]

Well, yeah. Exactly. 😛

If you're rebuilding, drain the whole line into ISCs.

[frosty owl]

Eh, too much work to set up a single boot. I like it when I just connect the input belts and everything goes brrr without needing to fill ISCs with tons of items (which is kinda hard for me atm due to lack of production anyway)

Ngl though, the appeal in seeing the machines boot up in order with the belts not clogged is strong on me

"clogged" as in "full an moving in stop-and-goes"

[ashen girder]

Yeah, I think that's what drives most people. 😂

(That care about balancers)

[frosty owl]

You can be like (POV: You're watching uranium fuel rods manifacturers):
Here come oscillator 1... And 2... Aaaand 3! machine starts producing

[frosty owl]

Either that or just not wanting spooling times
On high tier items they become quite big. Nuclear gives another reason to balance too

[ashen girder]

Yeah, that's probably the best argument I've ever seen for balancing.

Someone else pointed out that you can go the opposite way and use smart splitters to force them to fill sequentially.

Which also helps speed up the ramp up.

[frosty owl]

Yeah, but honestly, without copy-paste function for the smorts I don't think it's worth the effort outside of special cases :/

[ashen girder]

Eh. I'd rather put that effort in than the balancer mess people seem to like doing.

[frosty owl]

Mhh... As for me, I think I would rather use the time to change the clock of the machines instead, so they can be belted more nicely (if the ratios allow for it without going mad)

[fierce ruin]

Single Resource Items

[hollow meteor]

i'm kind of all over the place. some techniques seem to fit certain lines better than others. but I do tune almost all my clocks, generally under

[unreal echo]

I’ve only recently started under and overclocking; I discovered that pure copper ingots can be factored nicely, 12 ore in to 30 ingots out.

[vapid gorge]

Stop creating eldritch horrors

So in running rough numbers for the big hubs I'm making... it sounds like I have to resurrect Screws into my world. Damn you huge quantities of wire creating different bottle necks

[crystal charm]

how did you get screws out? Are there enough alt recipes to do that?

[vapid gorge]

Yup! There are enough alt recipes that you can completely remove screws from your world if you like!

The downside is that it often requires copper/wire instead and for big projects copper and caterium to extend the wire become limiting resources for things

[vapid gorge]

What I'm ending up doing is designing factories for parts that need screws to have Steel Beam Screw constructors directly feeding each machine above that needs 700 screws pm XD

[fierce ruin]

Completely eliminating screws from all production is one of the highest priorities when selecting alt recipes to unlock.

[vapid gorge]

It's REAL good early/med game yeah

[fierce ruin]

Given the game doesn't start until you unlock mk5 belts I would say it's good for the entirety of the game 😂

[vapid gorge]

Well I'm doing some largish plans atm and looking at recipes. And there are some naturally high volume products that are just SO much faster than other alt recipes. But requires screws. And take up less iron doing so.

So steel beam screws for me I guess?

Make the logistic belt of 700 screws pm be 5m long

[fierce ruin]

"Faster" is really a factor of "how many buildings" so not sure what specifically you're building..

Also iirc screws is cheaper if you make steel rods then use the base screw recipe. (Compared to the steel screw recipe proper)

[vapid gorge]

I like neat and compact factories. Just looking at bolted RI plate and Bolt Frames

[fierce ruin]

Base RIP has the same production rate as Bolted and is cheaper.

Stitched is cheaper than both though.

Even if you use iron wire and do stitched from nothing but iron, it is cheaper.

[vapid gorge]

Base RIP is 5 pm Bolted is 15

[crystal charm]

I built a screw factory not so long ago, currently deals out 600pm

there's room to expand it too, though it would need a secondary output belt of course

[fierce ruin]

3 buildings of base will use less iron than Bolted and make 15 in the same amount of time.

[vapid gorge]

Painful with regular recipes right?

[crystal charm]

well, it was painful until i built that output, now it's nothing

[fierce ruin]

But I've had that debate numerous times and very few people seem to comprehend that base is the same production rate as bolted.
🤷‍♂️

[vapid gorge]

I'll have to make a volume comparison.

[crystal charm]

of course, if i jump back on and try to build the new output for the space elevator, i'll either need to make another screw factory (not so hard) or figure out which alt recipes i have to avoid them

[vapid gorge]

Gear up and go HDD hunting!

[fierce ruin]

Either way it is better to just use Stitched.

[crystal charm]

i am about due for another hunt, although i just built a train line to a new location where there is oil, pure iron, pure copper all within a reasonable distance of each other, i haven't checked yet for caterium

i dont' need to worry about more bauxite, my current aluminium plant is outputting so little at the moment compared to the intake of bauxite that i'm thinking of building a new plant somewhere else with a bit more room, and doing the math much better, as opposed to the current version of just winging it

[vapid gorge]

I'm making large enough factories that I've needed to use caterium to make wire just to get the volume but you can avoid that pretty easily if you don't make giant set ups

[crystal charm]

i kind of want the spelevator done, but the math in getting it done within like 100 hours of it just sitting there is abominable

[thorn bane]

just dont go overboard and eliminate all screws
copper rotors is still the most resource efficient
even when you only use steel screws

[fringe pawn]

Of course, the full resource efficiency rabbit hole for copper rotors also requires steel rod -> screws which gets kinda tedious. Until you're making a megafactory, steel rotor and default stators pair nicely.

[thorn bane]

imo steel screw -> copper rotor is already good enough for me
the issue is just the metric shit ton of water you need for the sheets since they are 76% water after all

[fringe pawn]

Pure copper and steamed sheets. Such refineries. Much energy.

[thorn bane]

MUCH water

[muted crypt]

It's the same rate if you factor in needing three machines running the recipe, yes...

Unless you're counting recipe cycles per minute as production rate, in which case both are 12s, or 5 cycles/min, sure

[topaz hedge]

Now this is going to be exciting.. https://u4.satisfactorytools.com/production?share=o8ngrENDTiARNIb3jMkX

[golden stag]

Probably been asked plenty of times before, but hasa anyone got a good way of distributing 5 even supply lines down to 4 receivers?

[wind spade]

easiest is to merge all to one and then manifold

[golden stag]

I would but the quantities per line are 240, so they max out at 480 at my current research level

[wind spade]

if you can't do that, then just split one of them to 4 and merge to each.
or underclock the receivers to 80% and build 5 instead for a nice 1:1

[golden stag]

oooh, good tip, thanks

[thorn bane]

highest you have is mk4?

[lethal wing]

look up factorio balancers and convert and build them in SF.

[thorn bane]

no dont do this
belt balancers are bad
use belt compressors they do the same but just better (they use smart splitters)

[eager solar]

it's also possible to do a manifold with mergers added between the machines when it is needed

then feed the merger the other belts

[golden stag]

yeah mk4 for now, went with Greeny's suggestion and it's worked a treat

[thorn bane]

heres what i would use (a bit overkill but it works)

[eager solar]

looks complicated

[golden stag]

Feel like that kind of setup is detrimental just because of the space it takes up, I do appreciate the tip though

[thorn bane]

if you stack them vertically (which is what im usually doing) it looks like this

[golden stag]

Okay, that's kinda groovy

[thorn bane]

the way it works is it merges two 240 lines and then sends the overflow to the top line
so since only 300 is used 160 is send to the next line
that then gets merged with 240 with overflow going top etc.

[eager solar]

I'd do something like that. Let's say you have 2 belts, one able to feed 3.5 of the 5 machines and the other 1.5, putting a merger between the 3rd and 4th machine will allow you to keep the excess from the main input and complete it with the second belt

it also works with fluids

you can tweak/expand it as needed, keeping it manifold-like

[thorn bane]

how would that work with 5 lines?

[golden stag]

ahh, okay, that's an interesting approach. Doesn't that mean that it will fill the first machine up much faster than any of the others down the line?

[eager solar]

just do the math to know where you need to add mergers, according the amount required by the machines and what you have on each belt

basically, when the first belt ran short of being able to completely fill the first machine, add in a merger with the second bely. when that new amount fall short, add in a second merger and the 3rd belt and so on

[thorn bane]

i dont really see how that would be possible with 5 lines 4 outputs

[eager solar]

I'm assuming you have the right amount of resources, so you could either have a section with two mergers between two machines, or just merge your belts until you have less than 5, even if they are not balanced

[thorn bane]

could you make that picture with 5 lines and 4 X?

[eager solar]

something like that: s---m---m---s

or you just fuse two of the five belts together (or more why not) to get it down to four belts or less

the point is: feed as many machines as allowed with one belt then take the leftover and complete it with the next belt

[thorn bane]

but how do you make sure that the second belt isnt just blocking the input from the first

[eager solar]

*bit lat but mostly yes, it's essentially a manifold after all

[eager solar]

math. as I said, you need to know how many machines your belt can feed

in my exemple, I know that I will only get back enough to feed 0.5 machines from the first belt

[golden stag]

gotcha, I've always tried to keep each machines per-minute input the same, at all times, so not just filling one up while the others get half feed rate. But your solution does work

[eager solar]

and manifolds being what they are, everything will be used as long as they have a place to go

which is guaranteed since you will have one main belt running through all machines

[eager solar]

if you mean load balancer I highly advise against doing them (or at least not only those) as setups will get extremely complicated and brain-bursting later on

while manifold is always the same thing

the one downside of a manifold is the time for the prod line to fully start, since you need to first machine inner buffer to be full before the second one start to really get the items

and so on

but well, it's a one-time thing, and you can add that touch of load balancer to reduce it (which means doing the same thing as in my pic above)

[thorn bane]

it would be like this right?
(i think you need smart splitters but not 100% sure)

[eager solar]

no need for smart splitters and yes it would be like this (though it would also depend on the actual amount of items per belt)

if your reason for doing that is belt size, you're most likely able to at least reduce the number of belt you're working with making it less bulky

like in your pic, I see that you merged the first two belts before the first splitter

[thorn bane]

i dont think this would work without smart splitters

[eager solar]

I don't see why you want smart splitters, we're talking about only one kind of item right ?

[thorn bane]

yes i mean with "any" set to left overflow to middle

[eager solar]

no point if you place the mergers in the right place

think of it like that: merging the .5 leftover and 1.5 from the belt just means you're making a 'new' belt able to feed 2 machines

[thorn bane]

ok so while starting the setup would merge 2x 240 for 480 then split into 240 etc. until the last line
that line would then fill up (like a manifold)
once its full it would only allow 300 throughput so it would limit the last belt to 150
so instead of 240 input youre now only getting 150 bottlenecking the whole system

[eager solar]

if your machines are working, the belting will never stop because the machines will keep using a set amount of items per min

it would clog tho if you were to place the mergers to early

[thorn bane]

so the machine will use 300/min
and youre merging 2x 240
so the belt will back up

[eager solar]

back to my example, it would clog if you place the merger between the first and second or second and third machine

[thorn bane]

but theres no other way to place a merger in the example with 5->4
so unless youre using smart splitters in that example that would not work

[eager solar]

that's what I mean with math, also we're taking into account both belt size and items consumption

[thorn bane]

but theres no other way of placing the mergers
no matter what math you do

[eager solar]

I mean that as merging two belts into one prior to injecting them into the system (granted, my second schema was inaccurate)

[thorn bane]

so how else would you do it?

[eager solar]

the initial problem is that you can't merge all five belts into one because of your current belt limit, but even then it should be possible to merge at least two of the belts together and reduce the amount of belts your working with

so you no longer work with five belts but less

then you just setup a manifold as usual, simply completing it with the next belt where needed

[thorn bane]

well lets say you have 3 belts with 480/480/240
how would the system look?

[eager solar]

how many ipm would each machine consume ?

[thorn bane]

300

[eager solar]

what's your max available belt size ?

[thorn bane]

mk4

[eager solar]

gimme a sec to make the pi

oh, I get what you're asking now, in the case with several full belts you could actually split them further to avoid the clog

[eager solar]

yeah, that work too in this case

[thorn bane]

thats feeding the manifold from the left and the right? not just injecting
hm

[eager solar]

that being said, which machine do you have in mind?

[thorn bane]

idk im just going with the example from the guy

[heady mist]

Could also split the last 480 and inject in batches if you want to stick to injecting

[thorn bane]

how?

[eager solar]

because I doubt you're feeding only 1 machine with 300ipm

[thorn bane]

well its a factory that takes 300/min

[eager solar]

that's important

[thorn bane]

for example 1 factory with 10x steel pipes
another with 5x steel beeams

[fierce cypress]

480 -> [-300] -> (inject 240)- > [-300] -> (inject 240) -> [-300] -> (inject 240) -> [-300]

[eager solar]

because depending on the ipm required by the machines on the second factory, you could actually use the 180 leftover before having to merge

as I said earlier, merging too early cause a clog

[thorn bane]

again wouldnt this require smart splitteres?

[heady mist]

What @fierce cypress said basically

[eager solar]

nope

[thorn bane]

but while filling the last 240 will clog

[eager solar]

yeah, that pic refer to what I said about splitting further if needed

[eager solar]

how so ?

[heady mist]

That’s 240 output though, not 300

Oh I see

[thorn bane]

no its 240+240 and the factory is full and only takes 300

[heady mist]

But in your pic your first three machines take 240

[eager solar]

if I understand your pic, it shows that you have more ipm than actually needed

[wind spade]

just underclock 300->240 and split 480 into two

[thorn bane]

theey dont "takee" 240 its just 480 split into 2

[thorn bane]

just smart split with any top overflow right

[wind spade]

underclocking ftw, don't need smart splits

[thorn bane]

300 * 4=1200 = 240 * 5

(and now if you overflow split you basically have my design xd)

[heady mist]

Yeah I see what you mean. Personally I just let it clog till they fill (usually I manually fill everything lmao)
Depending on numbers it’s possible to use belts to limit some outputs so the machine gets the full amount immediately, but if you get to that point then might as well as smart split XD

[fierce cypress]

but this will work if you let it back up

i always let my machines back up when overflowing

[thorn bane]

i really dont think so....

[fierce cypress]

why wont it?

[thorn bane]

because full machines doesnt solve the 240+240->300 issue

[fierce cypress]

but if you let everything back up

you just need to feed the system enough to keep up with demand

[thorn bane]

yo i think im just gonna build it gimme a sec

its the same with 24+24+24+24+24 -> 30+30+30+30 right?

thats easiere to build

[fierce cypress]

yea

[thorn bane]

actually then the belts wont clogg...

[fierce cypress]

o yea right

lol

[thorn bane]

what about 5x 30 into 4x 37.5 using mk1 belts

[fierce cypress]

i think that would work

might be some funny ratios tho

[thorn bane]

nah i just oc an iron smelter to 37.5

[fierce cypress]

what abt the input tho?

[thorn bane]

take a mk1 beelt split into 2

[wind spade]

if you need 5x30 into 4x37.5, just UC the 37.5 to 30 and have it 1:1 lol

[heady mist]

Basically what will happen is:
The last machine gets fed 480 but can only take 300.
When that machine clogs, the second to last machine gets to take the full 480 rather than the 240 from a free moving belt. This overflow fills its buffer and will keep feeding the machine when it’s only getting 240, making it possible for it to get a continuous 300. Also because the average of 480 and 240 is 360, this machine slowly clogs as well and the same will happen to the upstream machine.

[fierce cypress]

its for the 1200 --> 4*300 with mk.4 belt problem

you just wait untill the entire system is full

[wind spade]

which belts carry the 1200?

[fierce cypress]

2.5*480 belts

[wind spade]

hence my suggestion - underclock 300 to 240, have 5 belts, solution obvious

[fierce cypress]

yea but im saying that an overflow that has been backed up will also work

[thorn bane]

but the second to last is still splitting 480 into 2 right? its just that the right going 240 is clogged to 150
so its 240 + 150
or am i missing something

[wind spade]

oh yeah it will work. I'm just trying to suggest easier approach

[fierce cypress]

would it be easier tho? its the same amount of work, just less machines

it'll take like 1 min to fill up

[wind spade]

it's nice 1:1 instead of having to think about how to make injected manifold

[fierce cypress]

so you wouldnt manifold it?

[wind spade]

I would do it 1:1 - split 480 belts to 240 -> have 5x240 belts, each goes into one machine eating 240

so underclocked from 300

[heady mist]

What do you mean by 150?
The second to last machine takes 240 only if the belt going to the right is not clogged. If that belt doesn’t move, all the input goes to the machine until the belt moves again

[thorn bane]

but its not all its just whats not getting used be the right most splitter

[heady mist]

Yeah and that plus the 240 averages higher than 300, so that machine would fill and eventually clog too

[fierce cypress]

probably simpler, but the whole reason this is even still being debated is because of a manifold ¯_(ツ)_/¯

[heady mist]

If machines have zero buffer then it wouldn’t work

[fierce cypress]

but machines have a stack of items as a buffer so it does work

[heady mist]

This is how I run all of my factories XD like I’d make a ton of steel pipes then run it through all my pipe-using factories and have them take what they need lmao

[fierce cypress]

i mean it works as long as your supply keeps up with your demand

[wind spade]

manifolds work even underfed 🙂

[heady mist]

Yeah and I can just send whatever leftovers to storage

[fierce cypress]

?

their efficiency will decrease

wont it

[heady mist]

Although I technically should be using smart splitters and send forward only the overflow 😂

[wind spade]

yeah but that's not a property of manifolds, but property of any underfed build 😉

[thorn bane]

smart splitters are insane
they make factorio logistics seem like shit

alright this is the setup

[fierce cypress]

well, yes, but im assuming 100% lol

[fierce cypress]

it wont work if your sinking the items

[thorn bane]

as you can see the right line is starting to fill up (its the 300 480 thing)

[fierce cypress]

disconnect it from the sink

let the belt back up

[thorn bane]

im sijnking the outgoing items not the incoming ones

[fierce cypress]

and then reconnect it

[thorn bane]

it is

[fierce cypress]

it'll be quicker

[thorn bane]

oh you mean as start up to prefill everything?

[fierce cypress]

u dont have to prefill

just let everything fill up

before you start production

[thorn bane]

then why would i disconnect the output

[fierce cypress]

wait lol my bad

sorry

[thorn bane]

ok i prefilled everything
the first 2 smelters are losing items (down to 80 from 100)
the last 2 work fine
the bottom 480 (60) line is clogging

i dont se how the first 2 would ever get 300

since the splitter after the second will always go left/right/left/right it will always take atleast 240 from the left

[fierce cypress]

try disconnecting the output

and letting it go untill the machines stop producing

[thorn bane]

like this?

[fierce cypress]

(or prefill)

yea

[heady mist]

I see the issue and I think I have an idea on how to solve it

Let me think

[thorn bane]

smart splitters 🙂

[fierce cypress]

if all the belts and machines are fully saturated

it should work

[heady mist]

What happens if you swap around the first two 30 lines?

[fierce cypress]

ohh wait i see

[heady mist]

Yeah this is super cool

[thorn bane]

ok system is completly full

[fierce cypress]

yea yea

it wont work i see

[thorn bane]

ye the first splitter is already going left/right/left/right
so its only providing half so 240

[fierce cypress]

it might still work because of the saturation tho

[fierce cypress]

but if the right is full it just goes left

[thorn bane]

i mean it will just lose 37.5-30 items /min
an i am NOT gonna wait until 100 is empty xD

thats like 13 min

[fierce cypress]

lol

[thorn bane]

but ye you just do this
that way the merger will never clogg

once its full it will go 300 top 180 right

[fierce cypress]

yea

that works

[heady mist]

I think for it to work with this setup you need to put the split lines from the 480 input upstream so it can feed into the first machines when the downstream clogs

But that’s too much thinking XD

[thorn bane]

you mean feed from both sides?

oh btw i think this would work if you had a priority merger instead that would always prioritize the fresh line

[heady mist]

I’m thinking more like changing the order of the injections, like if you swapped your first two 30 lines the second machine prob would have worked fine too. Personally though I would prob take one 480 and feed from the other side, but smart splitters def works XD

[fierce cypress]

what if you reverse overflowed it with the extra going in the centre?

[thorn bane]

i think this works xD

[fierce cypress]

like this:
480 -> (-300) -> (-300) <- 240 -> (-300) <- (-300) <- 480

[thorn bane]

ye you just split it into 2
2.5 * 240 = 600
but at that point thats balancing and not a manifold :D

[fierce cypress]

technically thatsbalancing

but if u extended it out it would manifold

[thorn bane]

ok the image i posted actually works since it makes the mergers take 75% from the fresh line and 25% from the previous one
so it will fill up since its blocking the previous line (only 120 goes in 480-120 > 300)

[fierce cypress]

something like this:
780 -> (-325) -> (-325) -> (-325) <- 390 -> (-325) <- (-325) <- (-325) <- 780

idk if thats still balancing or not

[fierce cypress]

@thorn bane

[heady mist]

Actually I’d be interested if this works:
(480) <machine> (240) <machine> <machine> (480) <machine>

[fierce cypress]

whats the belt direction?

[heady mist]

Left to right, numbers are injections

[fierce cypress]

300 ipm?

for machines

[heady mist]

Yeah

[fierce cypress]

its got the same problem that the first one had

[thorn bane]

i mean sure if you wanna do math
i just do 780/780/overflow -> (-325) - 455/780/overflow -> [belt compressor] - 780/780/less overflow -> (-325) [belt compressor] 780/overflow (-325) etc.

[heady mist]

Oh nvm

It’s too late I need to sleep 😂

[fierce cypress]

lol

[fierce ruin]

Did I miss anything?

[wind spade]

hard to say, which recipes are you using?

[thorn bane]

18.75 ai limiters is 3.75 assemblers not 4
150 rotors is 37.5 assemblers not 30

[wind spade]

150 heat sinks is 20 assemblers and they require alclad sheeds and copper sheets

[thorn bane]

im assuming he has the heat exchanger alt

same with insulated osc.

[wind spade]

tools don't like that one

[thorn bane]

which one?

[wind spade]

heat exchanger

maybe with more alts

🤔

[thorn bane]

wiki is saying its less WR aswell
but max points is using it
i guess the copper just has more weight than bauxite in practice

hm use 240 copper to save on 60 bauxite
per 240 copper there is 81 bauxite on the map

i guess youre just more capped by bauxite than by copper (since you can just use iron wire instead of copper)

[brave jetty]

Hi guys i need help getting a 12/min belt from a 30/min belt

the satisfactory calculators dont help xdddd

[wind spade]

just put down a splitter and it'll eventually balance itself

[brave jetty]

aight thank you

[fierce ruin]

I don't care if nobody wants the answer anymore.
Add a merger to the 30/min belt where 6/min will be added to make 36/min
Split that into 2 X 18/min, one of those is finished, the other gets split into 3 X 6/min, 2 of those are merged into 12/min and the third goes back to the first merger.

[tepid gull]

what's the best way to balance pure, normal and inpure nodes @ mk3 OC to 2 lines??

[wind spade]

Ideal is to build different amount of lines so that you won't have to balance anything

[tepid gull]

this is 3 lines going to 2 freight platforms

so its an even load to both freight platforms

[wind spade]

Merge impure and normal to one, you still have known amounts of items in both

And use what you get

[tepid gull]

...........................

[wind spade]

There's no need to balance in this game

It usually just adds extra work for no added value

[frosty owl]

Except taking away waiting times for machines, reducing radiation to near zero...~

[wind spade]

In this case it's neither though

[frosty owl]

Right, balancing would "only" mean one wouldn't have to wait for the stations to fill up and overflow with ores before the each Staton s gets all it needs... Half an hour at the very least 🤔

[wind spade]

Why

Each station gets one belt without overflow

At the other end, you use what's on the belt

No wait time, no balancing, simple setup

[frosty owl]

If the setup at the unloading station is simple enough to need such inputs, that's convenient