Factorio Forums

I don't need more than 1 belt to unload my trains -- they are 1-1 (1-loco, 1-car). If I need more ore delivered, I just use more trains (and more stations sometimes).

With only 1 belt the station can be very compact. This time it's only 4-tiles between rails, with chest-buffer and balanced lane output.
Now I like 1-1 trains even more!
Short YouTube video:
https://youtu.be/BnH9Tq2fAFw
(Same video in Ukrainian: https://youtu.be/Sqylgg_GwX4 )

Some explanations:
* I really like that I could squeeze lane-balancer into the design.
* I fuel my trains on the unloading stations. Usually, I have lots of spare wood -- I use it as a fuel. Just let your drones bring 100 wood per requester chest. If there happens to be no wood available, the coal will be loaded instead (until the wood will come again). There is a simple logic circuit for that. Just connect the coal-inserter with the chest in question with a wire, and put the condition "Work only if everything is below 40".
* I use an underground belt to get both lanes working and decrease belt bandwidth overload. 3 fast inserters will have one lane full, so, for six inserters I have two belts braided.
* There are rail signals, of course. And there are lights.

Any suggestions to the noob?

one suggestion. wire all the unloading from chest to belt inserters together (for each car). set them to read hand contents (hold), enable when * (everything) = 0
that way they will not favor one chest over anotther. chests will be unloaded evenly and make unloading from trains quicker.

while in your layout the belt side balancer works since your only feeding it from 2 single sides, check out this splitter for balancing full belt sides.

i like the small footprint of your setup. well done.

impetus maximus wrote:one suggestion. wire all the unloading from chest to belt inserters together (for each car). set them to read hand contents (hold), enable when * (everything) = 0
that way they will not favor one chest over anotther. chests will be unloaded evenly and make unloading from trains quicker.

Thank you, will definitely try that.

impetus maximus wrote:while in your layout the belt side balancer works since your only feeding it from 2 single sides, check out this splitter for balancing full belt sides.

I suppose it won't work. It assumes we have items on both lanes, while in this case we have items on the right lane only. It will not go into the left underground belt.

BTW, don't understand what for is the second splitter in your design? Aren't lanes already balanced after the first one? With the two splitters there is no room for your balancer design in this station. We'll have to add one vertical rail only for that purpose. And I don't see any problem with my balancer design in this case.

Anyway thank you for your comment.

impetus maximus wrote:while in your layout the belt side balancer works since your only feeding it from 2 single sides, check out this splitter for balancing full belt sides.

In fact in this case we can do it even simpler: And it will work just perfect and will need much less belts.

the purpose of the two splitters is that if one side of the belt after the splitters is being used, it will pull material from both sides coming into the splitters.
its only needed when the splitter is being fed from both or single opposite sides.

the last pic you uploaded works because it's being fed from 2 of the same half/side of a belt. if you pull from one side after the splitter it will pull material from both belts going in.

Do you even really need the splitter at all?

Azraelle wrote:Do you even really need the splitter at all?

It's useful when consumption is not even on the lanes.

impetus maximus wrote:one suggestion. wire all the unloading from chest to belt inserters together (for each car). set them to read hand contents (hold), enable when * (everything) = 0
that way they will not favor one chest over anotther. chests will be unloaded evenly and make unloading from trains quicker.

Not tried it yet, and now I'm not sure it's a good idea.
Sometimes chests are unloaded not evenly. If some chest is full and others are not, the train will not be unloaded with those circuits.
If all chests have some free space, all chests will be loaded evenly even without the circuits.

Chests aren't unloaded evenly. Exactly. That's why the suggestion to wire up the inserters unloading THE CHESTS. It makes them all grab items from the chest and place them on the belt and then wait till all have dropped their items before grabbing the next.

So far the theory. It worked here too at the start. But now I got into a state where one inserter has no items in hand and hovers over the chest. The others hover over the belt dropping items. Once everything has dropped the lone inserter picks up items and turns to the beld while all others turn to the chests. But then they don't pick up items because the lone one has some in the hand. Once the lone one has dropped everything the reverse happens. So I'm constantly alternating between one inserter dropping items and all others.

well that's how it works. i found it doesn't keep all the chest filled exactly the same.
after hours of running i found a couple of chests with a few more stacks than the others.
still a good solution in my opinion.

more than that - circuit unload is slower too. It doesn't fill even one red belt from 6 fast inserters.

LazyLoneLion wrote:more than that - circuit unload is slower too. It doesn't fill even one red belt from 6 fast inserters.

It' not for unloading fast. It's for unloading balanced. With other setups you get for example 4 inserters dropping items on the belt at full speed filling the belt as fast as it clears while 2 inserters are blocked. Then the first 4 buffer chests run dry and the remaining 2 inserters place items on the belt on their own, at half the speed. What you want instead is 3 inserters filling the belt nearly full and then the other 3 inserters filling the belt nearly full. Or all 6 filling the belt but balanced so they all run dry at about the same time.

But here is an improvemt design that might give you both the maximum throughput and balancing. Use the same setup as before but use faster belts for the start. That way items clear faster from the inserters and then get compacted where the belt gets slower.

What might also help is to connect the inserters to the belts after where the inserters drop things. Set the belts to report and hold the item count and the inserters to work when item count == 0. This still relies even more on the faster belts for compacting the gaps away but I found this prevents alternating between one and all others inserters if you connect to the belts in the right distance from the inserters.

mrvn wrote:...What you want instead is 3 inserters filling the belt nearly full

Well, *I* prefer the belt full. Not "nearly".
And I don't really care about balancing unload just for the sake of balancing.
Full belt will unload trains faster than "nearly full belt" anyway.
And I like to have one station enough to fill one red belt. Not "3 nearly full stations for two belts", just one station - one belt.

I really don't see any reason to care that much about balancing all the chests.

LazyLoneLion wrote:I really don't see any reason to care that much about balancing all the chests.

Different use case I think. If you generally have the ore belt backed up all the way to the station, and use the station as a buffer so it is often over 1/3 full, then keeping the chest content balanced helps unloading speed. Since 4 empty 2 chests can happen.

It sounds like this doesn't really apply to you.

Linosaurus wrote:If you generally have the ore belt backed up all the way to the station, and use the station as a buffer so it is often over 1/3 full, then keeping the chest content balanced helps unloading speed.

I suspect that is not the case either.
If I have ore belts backed up for some time (which is often the case, I like it that way), then the chests will be mostly full, and then the time of unloading the train is limited only by the speed of the ore on the belt -- the train will stay and wait until the space in the chest(s) will be freed by slowly moving belts.

It might slow the process a bit for unloading part of the carriage into one chest at start, but the end will be the same moment anyway.

On the other hand, if the belt is never full, and I have enough trains to back up the chests, then unload will be slower again. Because then the belt is the limit, and a full belt unloads chests (and thus trains too) faster then a half-full belt.

I see 2 problems with unbalanced chests.

1) only a few inserters are unloading the train slowing that down.
2) full belts at the start but then some chests run dry and the rest aren't enough to fill the belt

There is a small window of throughput where you consume goods as fast as around 4 inserters can unload where the 2 conditions above combine and make things inefficient. Your production works in bursts , going full speed while a train is there and slowing down when not. The obvious solution is to just throw a second train at it to shorten the times when no train is there. But that's wasteful. I like perfect setups.

mrvn wrote:I see 2 problems with unbalanced chests.

1) only a few inserters are unloading the train slowing that down.
2) full belts at the start but then some chests run dry and the rest aren't enough to fill the belt

1) that's only when the belt is backed up and slow, and chests are almost full.

So, I don't really care about how fast the train unloads if belts and chests are full. Or how balanced. Let them stay there forever. It will not be faster with balanced unload anyway -- the bottleneck is always in consumption.
That's much better than having proportional unloading but having belt not compressed. Then the train will be unloaded slower too, be the belt full or not.

When the belt is backed up my design will fill it (and thus unload the train) as fast as the belt is consuming.
It will matter only in one and only case:
a) train unloads and all the chests are full;
b) train leaves and belt slowly consumes ores from one chest only.
c) that chest has space enough to unload whole train (it will take some time).
d) train comes in and contains ore that can be fit into free chest space.

If on the d)-moment there is not enough free space in the chest(s), than it doesn't matter how balanced unload is -- anyway the free space in the chest(s) will be backed up and train will wait for the belt to consume enough ore.

I usually don't have both backed up belts and not enough trains incoming.

2) see above.

mrvn wrote: I like perfect setups.

Doesn't seem like perfect for me.

LazyLoneLion wrote:That's much better than having proportional unloading but having belt not compressed. Then the train will be unloaded slower too, be the belt full or not.

I rather have proportional unloading and full belts. Easily done if you have a faster belt type at the buffer chests for a few tiles.

mrvn wrote:...full belts. Easily done if you have a faster belt type at the buffer chests for a few tiles.

Nope. Just checked it.

It's NOT because of belts speed. It's because inserters wait for each other, synchronously load their contents and synchronously unload. Thus there is a moment when noone inserter unloads and all other time they compete for space on the belt anyway.
In the case without circuits topmost inserter has a priority, bottom inserters have to wait and thus they load/unload asynchronously. So that there is not a moment when the belt is not full.

It helps a lot (REALLY a lot!) if we disconnect topmost three inserters from the bottom three. Still it doesn't give us fully compressed red belt.

Demo:
http://www.youtube.com/watch?v=ji4lSLD2J8I

P.S.
Helps even more to replace red underground belt with a blue one. Then we really have fully compressed belt at last. Not with all six inserters connected to one circuit.

BTW they don't work synchronously. %)
Somebody should check if they work evenly.

P.S. Yes, seems they are working evenly. Unload will be balanced.

P.P.S Still chests are uloaded not evenly... somehow. Much more evenly than without circuits, but still quite unevenly.