Let's see your clever builds

[Aru]

@DerivePi, what program are you using to make those schematics?

I direct you toward this thread:
viewtopic.php?f=8&t=24320&p=154493&hili ... pi#p153906

[FunMaker]

I Always thought of an oil train that was not limited by reserved full/empty barrels to maximize transport capacity of full(one way) and empty barrels(other way).

But if there were no reserved stacks most of the time one or another raw oil producing outpost would clog up because of too many empty barrels (because it was not easy possible to ensure that empty barrels stay on their route if there is only one drop of for full barrels.

So i came up with this solution:

Oil.jpg (313.53 KiB) Viewed 15291 times

The goal here is carrying only full stacks of empty/full barrels. So there will be always the same amount of barrel at an outpost. At the drop of they system is similar to keep the train always full with barrels. There is a small chance, that the train will depart in the moment the inserter is half the way while replacing a stack (empty -> full) but this did not happen in 60 hours of game time. This is also ensured by the equal distripution of full barrels in the loading chests.

The only problem with this is, that you have to keep some stacks of barrels at the station to fill them while the train is on rail and fill up the loading chests with full barrels because the unloading of one stack is only triggered if a complete stack of full barrels is available.

Hopefully my confusing explanation shows what i wanted to achieve and how it is achieved.

Maybe my setup is obsolete with 0.13 - but that's another story

[Aladrius]

But if there were no reserved stacks most of the time one or another raw oil producing outpost would clog up because of too many empty barrels (because it was not easy possible to ensure that empty barrels stay on their route if there is only one drop of for full barrels.

I use the arithmetic combinators to calculate the number of empty barrels a pump station needs for the current oil on hand.

I use 10 full barrel /10 empty barrel/10 unrestricted slots on the trains, but these are arbitrary and could probably be more like 5/5/20 without making much impact. Restricting at least a few slots ensures that the early stations in the line can unload some full barrels onto the train even if they currently don't need more empties or need less than a full stack of them.

Longer explanation here:
viewtopic.php?f=8&t=25931#p167285

[siggboy]

These barrel handling schemes look overly complicated to me.

In my current game I did not use Rail Tanker and came up with the following barrel handling:

1. There are two trains, one runs with full barrels, the other one with empty barrels (let's call them "black" and "white" trains).
2. The white train collects empty barrels from the oil unload and takes a tour of all the oil wells. At each oil well, it unloads enough empty barrels so that the total amount of barrels (black and white) at the oil well is a set amount (the amount should be lower for depleted wells, but that's an optimization that doesn't matter much).
3. The black train just picks up all the full barrels it can get from the oil wells.

This way, there can never be more than X barrels at a given oil well, and the amount of white barrels is regularly replenished everywhere.

I'm using SmartTrains to make the tour a bit faster, but it's not necessary. All you need is smart inserters and a few combinators that add white and black barrels and stop the unloading when the limit is reached.

[siggboy]

• A smart chest holds the fuel for each reactor, the smart inserter only inserts if there is 0 fuel in the chest, this guarantees each smart chest has at most 1 fuel.
• The smart chest is linked to a decider combinator which sends a Red = 1 signal down the shared green wire if the smart chest has 0 fuel. Red means "I need fuel!"
• The requester chest to the left is set to request 1 fuel assembly - bots will deliver up to 4, but that can't be helped.
• The smart inserter at the requester chest is set to insert only when Red > 0, so if any of the smart chests run out it starts sending fuel assemblies down the belt. Any excess loop around and get put back in the requester chest.

You should connect all the chests directly, and input the total count to an arithmetic combinator that divides by the number of chests. If the result is 0, then at least one chest is missing fuel, so that is the condition for the unloader smart inserter to put fuel on the belt.

That way you only need one combinator total and not one for each reactor.

[FunMaker]

These barrel handling schemes look overly complicated to me.

In my current game I did not use Rail Tanker and came up with the following barrel handling:

1. There are two trains, one runs with full barrels, the other one with empty barrels (let's call them "black" and "white" trains).
2. The white train collects empty barrels from the oil unload and takes a tour of all the oil wells. At each oil well, it unloads enough empty barrels so that the total amount of barrels (black and white) at the oil well is a set amount (the amount should be lower for depleted wells, but that's an optimization that doesn't matter much).
3. The black train just picks up all the full barrels it can get from the oil wells.

This way, there can never be more than X barrels at a given oil well, and the amount of white barrels is regularly replenished everywhere.

I'm using SmartTrains to make the tour a bit faster, but it's not necessary. All you need is smart inserters and a few combinators that add white and black barrels and stop the unloading when the limit is reached.

If it's okay to use 2 trains for that purpose - that's fine. But i don't like reserved stacks for oil transport.

[MaexxDesign]

My solution for green circuits.

https://picload.org/image/rgwwwwcp/fact ... ircuit.jpg

[PiggyWhiskey]

Nice one!

Aren't those ratios off though? Green for efficiency doesn't touch productivity. So you'd need 3 cable and 2 circuit assemblers?

[AutoMcD]

Nice one!

Aren't those ratios off though? Green for efficiency doesn't touch productivity. So you'd need 3 cable and 2 circuit assemblers?

Don't think it matters when there are so many of them!

Personally I think green circuits are one of the best places for productivity modules since they are so widely used, over half the resources go into circuits. Efficiency not to be underestimated though for reducing power and pollution.

[MaexxDesign]

The production is at full capacity.
I need them to produce red circuits.
And I need red circuits (and green ones) to produce blue circuits.
I produce 11.000 red circuits per minute.
I think that's enough to produce blue ones.
My goal is to start one rocket every minute.

That's the beginning:

https://www.youtube.com/watch?v=KjYAqp4YBPM

[PiggyWhiskey]

I tried to optimize the sorter design for one belt of full density and with proper output side-balancing.
The first version without backlog protection has full throughput and is probably about as compact as it gets (fits within 9x4).
The second version with backlog protection is quite a bit larger (16x6) handles the full belt but only allows for about 75% throughput in the long term, there is probably room for improvement here.
I'm interested to see, if there is a way to do backlog protection without sacrificing half of the output.

PS:
Don't forget to set the signal for both sorting units correctly, otherwise these designs just work as huge splitters/balancers.
By setting the signal, you can choose the output side for every item type. This allows for sorting and rebalancing of more than just two types.
By default the signal is set so that every item type gets split equally.

Image 1 Without Backlog Protection

Image 1 With Backlog Protection

So I found this post a quite a few pages back as a filter without Smart Inserters
I'm trying to use it with Bobs Mods Galena (Lead and Nickel Ore output) but I can't seem to figure out which splitter sides needs the first item as a Control Switch.

Anyone have any ideas?

[Neotix]

Side is not important (i tested it). Important is to place control item before second splitter to switch output for that item.
I explained hot it works here viewtopic.php?f=8&t=6008&start=300#p135617

P.S. In v0.13 it won't work anymore because of changes in splitter behawior.

[pieppiep]

P.S. In v0.13 it won't work anymore because of changes in splitter behawior.

What is changing in its behavior?

[Neotix]

Splitter will split items equally on all lines, not on belts.

[PiggyWhiskey]

Side is not important (i tested it). Important is to place control item before second splitter to switch output for that item.
I explained hot it works here viewtopic.php?f=8&t=6008&start=300#p135617

P.S. In v0.13 it won't work anymore because of changes in splitter behawior.

I think I'll be on 0.12 for a bit until mods get updated.

I saw that post and read it. I struggled to understand it.
The problem I'm having now, it's I can't get it to work reliably.
If I do a simple test it works fine, but the moment I use the larger design, it goes pear-shaped.

Like This

The design on the right worked fine the first time through. The one on the left does the Lead (Black stuff) but mixes on the other side.
And I'm 99% sure I did the same thing.
Because they're 2 sets mirrored, I put 1 item on each half where the Medium Power poles are lined up.

[Neotix]

I didn't try that setup so i don't know where is the problem. Try to remove all splitters and place them again (it reset internal switch). Set filtered items again and try if it work.

[mooklepticon]

Side is not important (i tested it). Important is to place control item before second splitter to switch output for that item.
I explained hot it works here viewtopic.php?f=8&t=6008&start=300#p135617

P.S. In v0.13 it won't work anymore because of changes in splitter behawior.

I don't think this changes in 0.13: viewtopic.php?f=5&t=27021

[PiggyWhiskey]

I didn't try that setup so i don't know where is the problem. Try to remove all splitters and place them again (it reset internal switch). Set filtered items again and try if it work.

So it seems it is side dependant.
Putting the Lead on the outside lanes seems to have fixed it.

[Neotix]

I'm sure that first 2 splitters are not sensitive to on which lane you put lead item. I tested it on my design and just now I tested it again (test setup https://i.gyazo.com/995114ef98e0fbe0fe4 ... af88fb.png). In all cases (testes all 4 input lines, even that not used) i get sorted items on internal lines and unsorted on external ones. Maybe rest of design is sensitive but I didn't tested it.

EDIT: Ok I tested it.
Test setup https://i.gyazo.com/d95fcf050f0cc3863c2 ... 8d49f5.png
When I put lead item (iron plate) in both splitters i get mixed output. I tired all lines and still get mixed items.
But when I put copper plate in left splitter and iron plate in right splitter I get sorted items (IP on left and CP on right)

When I added third item I get that item on both outputs but IP and CP were still sorted.
https://i.gyazo.com/65e7590d3127b6b468e ... c781bc.png

[Hammerchief]

Simple automatic buffer
I got annoyed by my buffer inserters always stealing items even when i needed belts full capacity. So I made this simple buffer which will take only surplus items:

factorio_buffer_1.jpg (308.29 KiB) Viewed 12483 times

It's using standart behavior of splitters - that they will send all items on one lane when the other is full. This allows to lock or open one lane comming from second splitter and either of that will not hinder belts throughput.
Reading items on standard line will not tell me, if items flows or not. So we will make some room by spliting a lane first. Standart count of items moving through lane behind splitter is 3-4 when on full capacity. If lane gets clogged later on, item count in checked half-lane will increase and way to the buffer will open.

Advantages:

• It operates only when its needed - when consumption slows or stops and items starts stacking up on the belt.
• The surplus is very well recognized even when consumption is just somewhat slower.
• It doesn't change throughput of a belt, its full capacity is sent to factory when needed.
• It will never take items from belt when it isn't fully compressed.
• It's not dependent on technology, works the same with normal, fast or express transport belts.
• It's easily blueprintable, settings is the same for any item.
• Its capacity is easily scaleable. Just take or add more chest rows or columns. It takes about two minutes to fill one steel chest worth of items from express belt, three minutes from fast belt and six minutes from normal belt.

Disadvantages:

• Chests are not filled equally. Buffer will take items slower when at almost full capacity, as only last few inserters will work. Use MadZuri's smart loading if you want to change that. (Google for MadZuri's smart loading train station Guide, I cannot post URL.)