Factorio Forums

I have found a unique use for the cargo car. As a buffer. Because of the way I play, I needed to have a chest that I could pull from more than just three sides.
The Cargo Car has that ability. I just realized if I only used one or two inserters to fill the car I could get even more outputs.

The reason for my need is I build my factory completely different than any I have seen(qualified). Every factory I've seen so far (and I have seen a lot), they all seem to use the full belt feed method. It seemed to me that a full belt of 150 red science to do a 20 red science tech, was overkill. Early game I would rather use those resources to build up my base. Also it seemed to make me constantly find more and more Iron and Copper to keep the belts full. So the method I use, I call the Double Buffer and Count method or (DBC). In this method for every item that you assemble, you build a Primary Buffer, a Local Buffer and a Input loop counter.

Here is a snapshot of my Green Circuit Distribution Hub using a Cargo Car as a Distribution Buffer.

http://images.akamai.steamusercontent.c ... 8CD32702F/

The Cargo Car is in the center and from it I have several belts with smart inserters and OR Combinators controlling them to the various locations. Since I share belts for multiple destinations I have to make sure that the Green Circuits are available to each inserter when ever they need it. The problem I had earlier using smart chests was that they would get imbalanced and trying to feed six different smart chests with six different amounts was giving me a headache. I have gotten to this point in my factory without the need for a train. As I was researching how to build a train network, and read about the cargo car and how it can take many inputs and outputs I realized it was a solution to my problem. With it I only needed to keep one chest full and I could pull out of it from many different locations and could actually use it to distribute to over 6-10 needs and with using OR Combinator circuits to control them I could double or triple the amount of locations I could distribute to out of this one long chest. I wish someone would create a mod for a chest that is 1 segment wide and 10 segments long. Anyway as you can see in my snapshot it works quite elegantly!

Here is a snapshot of my new Iron Plate Production using the Cargo Car Buffer. Because of the economic benefits of the DBC Method I just have this fill a 4K buffer and stop. It really saves me on Electricity, once the buffer is filled the plant will only run for maybe 10 sec every minute or so to keep up. So by using smart inserters to stop the output, the furnaces have filled their internal buffer and will output a massive amount the second they are told to, which is usually enough to backfill the main buffer, then shut down, using no electricity. If you looked at my Power Stats you would see that Inserters and Assemblers use the most energy not the Electric Furnaces. Depending on when you look, the furnace power requirement is often zero and I am never starved for Iron or Copper Plates.

http://images.akamai.steamusercontent.c ... 0440BE051/

I'm sure this has been done before, but I have yet to find an example using this method of Factory Design, so I will go into a little more Detail how it works.

PRIMARY BUFFER:
The Primary Buffer is a large buffer. My rule of thumb is the Primary Buffer should be a minimum of 10 times the Production Cycle requirement and if its a long distance from the local buffer I add another 5-10 PC's worth. A production cycle in this context is the amount of parts to make a complete assembly (ex. for a Green Circuit uses 4 copper wire so if I had only one assembler making green circuits the Primary buffer would be 40 copper wires. I have twelve GC assemblers and a 500 copper wire primary buffer) Generally larger is better for the primary buffer.

Local Buffer:
This is a buffer right next to the assembly area and feeds the Input Loop with a measured amount of parts. This buffer can be as small as 2X the PC requirement. You can make this larger for a safety factor, it will just mean that your main buffer won't have to work as hard. For a single Assembly there will be one Local Buffer for each part the assembly needs. For Green Circuits there will be two local Buffers, one for Copper Wires and One Iron Plates. As long as the local buffer has enough parts your assembly machines will run at full speed, if you feed the input loop properly.

Input Loop with Counter
The Input Loop feeds the assembly machines with parts. The parts on the assembly loop do not have to be sorted you can put them on any way you want as long as they go by assemblers and can be picked up by the counter. I try to make the loop a compact as possible. After the parts have passed the assemblers, if there are extras they will be picked up by the smart inserter which is filtered to only pick up one type of part. It goes into a smart chest and then a regular inserter places it back into the loop. Since the regular inserter is slower than a smart inserter, it will cause a backup in the smart chest. I hook the smart chest up to a smart inserter at the local buffer. The smart inserter will activate when the counter smart chest registers less than a constant I pick based on how many are needed. In most cases you can start with 1. Since there is a delay from the time the local buffer releases till a part will actually get counted you will often inject 10 to 15 parts into the loop. Enough to satisfy a couple of production cycles. The counter will pick them up and start to build up its count shutting off the local buffer. By changing the constant the counter is compared against you can adjust how many parts you inject into the loop and affect the timing of the input.

After all that the output of the assembly machine gets put into another smart chest which has become the primary buffer for the next assembly that that part is used in. If you do this for every assembly you get this:

http://images.akamai.steamusercontent.c ... 03A35F2A5/

At this point I have only about 5 techs left to research. I am going to redesign the factory bigger to make my space ship which I understand takes a lot of resources. Look at my mini map and notice how little pollution I am emitting for this stage of the game. I've gotten rid of most of the biter nests near by and the pollution doesn't reach the other so they don't attack. I may have an issue since there was so little pollution the biter nests did not get super big. I only have about 750 artifacts so I may have to range far to find more biters.

Anyway, just wanted to show something a little different.

Nice idea The reason why everybody is using full belts is that resources ingame are generally cheap and you produce everything in mass quantities. And becouse of the big buffer, if you suddenly empty eg. chest with blue belts the whole factory won't be iron-starved in the meantime.

By the way, you may be interested in another usage of cargo wagons:
viewtopic.php?f=8&t=6008&start=70#p69814

psorek wrote:Nice idea The reason why everybody is using full belts is that resources ingame are generally cheap and you produce everything in mass quantities. And becouse of the big buffer, if you suddenly empty eg. chest with blue belts the whole factory won't be iron-starved in the meantime.

By the way, you may be interested in another usage of cargo wagons:
viewtopic.php?f=8&t=6008&start=70#p69814

I understand that there is no resource limitation. My factory is still running (almost out) on the original Iron ore deposit. The primary reason for doing it this way is that it is a much more interesting factory to watch and manage. I can empty my Blue Circuits and not starve my factory, in fact I often do that just to watch everything kick on. The local buffers take care of 5 or 6 production cycles and that is plenty of time for replenishment from the main buffer. And I have found that I don't need 1000 destroyer bots. With 20 or so I can take out all the biters in my area, because they are not huge pollution fueled colonies. The reason for that is I don't pollute, most biter colonies near me don't get any. And since I cleaned out the close one, the others only grow at the lowest rate. I only have about 15 gun turrets and a few laser guarding my factory.

Your factory looks cool.. Although I prefer the sight of belts full of stuff
But in your factory it probably feels more like every produced item has a purpose instead of just being dumped on a belt, that is awesome.

Not sure if you found this already but Equalizer chests.
These chests share their contents with the equalizer chest next to it, like this you can create a 1x10 chest by just placing 10 in a row.
Might add that the sharing between chests become less effective over distance so basically you still should add at least 1 INPUT inserter for every 2-3 chests. And there is a minimum amount before it starts to share to avoid a giving back and forth loop.

Using wagon as buffer have one advantage. You can set filters and prevent from overflow. Chest can be easily filled with one type od item and entire process will stuck.
The disadvantage of that solution is wagon placement. You have to use 2x2 grid, wagon have different vertical and horizontal length, it's not so easy to organize production chains with many wagons etc.
I used that while ago when somebody firs announced that idea and after time I resigned from using it.

Neotix wrote:Using wagon as buffer have one advantage. You can set filters and prevent from overflow. Chest can be easily filled with one type od item and entire process will stuck.

You can set filters on a chest, can't you?

nope, you can only limit the avaible stacks in chests, a cargowagon can have the slots set to a specific item
but a wagon have much less space per tile than a chest

Neotix wrote:Using wagon as buffer have one advantage. You can set filters and prevent from overflow.

Actually it has another big advantage, similar to warehouses (in the warehousing mod): you have way more places to add inserters, allowing it to function as a natural balancer.

On station (un)loading, an easy solution for the 'unbalanced chests' problem is that you replace the row of chests with another wagon, and place a double row of red inserters in between. Then, you can simply have 6 blue inserters (un)loading the 'buffer wagon', and you won't have the problem that the first chest is full while the last chest is empty.

Of course, 6 blue inserters can't remotely keep up with 12 red inserters with stack size bonus, but if there is enough time between trains this is no problem.

To balance buffer in station and in many other places I'm using simple logic circuit and it works perfectly. Using wagon as unloading buffer is waste of space and resources.

About blue and red inserters. If i remember right, blue insertes is 2x faster then red one. Long ago double line of red inserters was little faster because of arm stretching but that issue was solved and now double row have no advantages.

Using wagons make sense to work with assembly machines because we can avoid belts spaghetti, and because of filters in wagons, entire setup is safe. Chest don't have filters (warehouse is just big chest) so entire setup can be overflowed. You need some kind of big smart chest to set condition in smart inserters to avoid overflowing.

Neotix wrote: To balance buffer in station and in many other places I'm using simple logic circuit and it works perfectly. Using wagon as unloading buffer is waste of space and resources.

Well sure there are other ways to balance, logic circuit or splitter cascades all work fine as well. But that's not the point

Neotix wrote: About blue and red inserters. If i remember right, blue insertes is 2x faster then red one. Long ago double line of red inserters was little faster because of arm stretching but that issue was solved and now double row have no advantages.

Yes, blue and 2x red are almost or exactly the same speed. But wagon->wagon gets inserter bonus, and wagon -> belt doesn't, hence the inserter to the belt needs 5 times as much time as the wagon->wagon/chest inserter.

I'm speaking about this setup: it also use inserter stack bonus.

The only difference (speed wise) is, that red inserters must move a longer way. But when it's 10 blue inserter vs 30 red inserters, the red should win.

Neok wrote:Your factory looks cool.. Although I prefer the sight of belts full of stuff
But in your factory it probably feels more like every produced item has a purpose instead of just being dumped on a belt, that is awesome.

Not sure if you found this already but Equalizer chests.
These chests share their contents with the equalizer chest next to it, like this you can create a 1x10 chest by just placing 10 in a row.
Might add that the sharing between chests become less effective over distance so basically you still should add at least 1 INPUT inserter for every 2-3 chests. And there is a minimum amount before it starts to share to avoid a giving back and forth loop.

Thanks, for the comment on my Factory. I just completed my first game using this method. It took 49 hours. I don't know what the average is but I think that this was probably a long game, as I have seen speed runs done in a couple of hours. It is fun to watch the factory turn on but its hard work to get balance in trying to get just the right amount of product rotating on the belts. I wish there was a smart belt mod I could add. One that would provide information on the contents of a belt section.