176 to 176 belt balancer
-
- Burner Inserter
- Posts: 14
- Joined: Sun Dec 22, 2024 5:38 pm
- Contact:
176 to 176 belt balancer
i have a rather large build that spits about 2million green circuit per minute spread into 176 belts, and now i am having difficulties designing a belt balancer for this monstruosity, i'm trying a dividing approach of 8x22, but a 22 to 22 belt balance doesn't exist and i'm not good enought yet to build costum belt balancer, if anyone is an expert on this matter, i would love some help, (PS : i cannot upscale or downscale to fit into a better option due to width issue with upscaling and throughput issue with downscaling
- Attachments
-
- 12-23-2024, 13-02-49.png (1.22 MiB) Viewed 818 times
-
- Burner Inserter
- Posts: 14
- Joined: Sun Dec 22, 2024 5:38 pm
- Contact:
Re: 176 to 176 belt balancer
i might have found a cheaty way to do it, instead of doing a single number division, i did a multiple number division
first i splitted 176 into 24x2 and 128, then 64x2 and 24x2, then back to 24x2 and 128, then another 64x2 and 24x2 then back to 176. every lanes communicate with ever lanes, some lanes communicate more than once, but i think it shouldn't be a problem since it'll balance them anyway, any though ? it's bulky and can probably be shorten by a belt balancer pro but i don't have the skills to do that
first i splitted 176 into 24x2 and 128, then 64x2 and 24x2, then back to 24x2 and 128, then another 64x2 and 24x2 then back to 176. every lanes communicate with ever lanes, some lanes communicate more than once, but i think it shouldn't be a problem since it'll balance them anyway, any though ? it's bulky and can probably be shorten by a belt balancer pro but i don't have the skills to do that
- Attachments
-
- 12-23-2024, 14-12-02.png (17.73 KiB) Viewed 779 times
Re: 176 to 176 belt balancer
I never belt balance past about 8-10 and very occasionally 12 or 16 and instead rely upon the inherent balancing of railway connected production blocks which I find works well, scales easily etc. I automate via circuit train stop limits and priorities which seem to ensure even enough distribution.
I know this is not the answer you are looking for, but your post reminded me why after going large scale with belts once in the distant past, I thought never again.
These days I just do smallish simple production blocks that are basically up to the capacity of a 8 wagon trains in/out for each of the ingredients/products etc. This becomes especially easy to ad-hoc grow as needed when you combine with city block/rail grid type arrangements. Need more of something? just plaster another copy of a block down, connect it and job done and move onto next problem.
There used to be some quite useful threads here about large scale balancer theory/design from many years ago - might be worth searching a bit if you really wanted a large scale belted solution.
I know this is not the answer you are looking for, but your post reminded me why after going large scale with belts once in the distant past, I thought never again.
These days I just do smallish simple production blocks that are basically up to the capacity of a 8 wagon trains in/out for each of the ingredients/products etc. This becomes especially easy to ad-hoc grow as needed when you combine with city block/rail grid type arrangements. Need more of something? just plaster another copy of a block down, connect it and job done and move onto next problem.
There used to be some quite useful threads here about large scale balancer theory/design from many years ago - might be worth searching a bit if you really wanted a large scale belted solution.
-
- Burner Inserter
- Posts: 14
- Joined: Sun Dec 22, 2024 5:38 pm
- Contact:
Re: 176 to 176 belt balancer
it would be a good idea as train network cost way less to build than this massive river of green belts, however the sheer amount of green circuit produced makes it impossible, i'd have to make a train go in the station, get filled up and leave the station, every 12,3 second for each units (using 1 wagon trains), which is very probably undoable with the space available, you can't do mono directional due to the limit space between each unit recquire to turn around. now if you do more wagon per train it might be more possible, but the sheer amount of train needed will most likely overflow any rail highway, and the longer the trains the longer it wil get blocked, disabling other trains to exit in time, drastically reducing throughput, but if you think it can be managable i'd love to see what you come up with, might try something on my own aswell
-
- Burner Inserter
- Posts: 14
- Joined: Sun Dec 22, 2024 5:38 pm
- Contact:
Re: 176 to 176 belt balancer
here is what i designed so far, but i'm having a lot of trouble with the exit, i need to make it atleast 2 rail wide, because if my math is correct i'm gonna need 44 trains every 48 second. wich is roughly 1 train per second, i'm sorry but idk if train network is doable
- Attachments
-
- 12-23-2024, 17-46-29.png (1.42 MiB) Viewed 684 times
Re: 176 to 176 belt balancer
Hard to see, but looks like 4 carriage trains, but really not sure about the unloader setup - seems a bit skinny to handle a lot of throughput, but hard to see.Globetrotteur wrote: ↑Mon Dec 23, 2024 4:46 pm here is what i designed so far, but i'm having a lot of trouble with the exit, i need to make it atleast 2 rail wide, because if my math is correct i'm gonna need 44 trains every 48 second. wich is roughly 1 train per second, i'm sorry but idk if train network is doable
I havnt actually done much since space age with high thoughput stations, but my starting points per carriage would be stack inserters for bulk stuff, and I would guess 8 of them (double sided) would be needed to stack and compress a green belt, or 12 rare stack inserters double sided to stack and compress a pair of green belts. That would be my starting point for testing along with 8 carriage trains just because that is what I tend to use to go with the block size I settled on back in the factorio 1.0 days. That is just a rough guess, but I dont expect it to be far off and would take just a couple of minutes to test in creative.
I also use mixed raised/ground rails in my train grid and have a regular grid around my building block grid with vertical rails raised and horizontal rails on the ground just to more or less remove junction as trains hardly ever need to cross a rail which allows a lot more trains on a simple track pair while still keeping very free flow.
Anyway, I wouldn't go for a direct belt to rail replacement, but that is just me. Im lazy and evolved into a very copy/paste centric build style ages ago with a simple fixed grid. Design a simple rail architecture that scales and stick with it so scaling up on just copying and pasting existing block into a good performance rail grid and connecting them. Within rail grid cells ill have a few production units design around trains capacities in and out. Small, simple, easy to scale by replicating. My grid is organized in 100x100 blocks, 3 blocks for a train production unit including rail connections and station etc, then 4 production units to a rail grid cell. Rail entry/exit on the left connecting to raised vertical rail. That is just what I was last using, but if/when I get around to high scale with railways again, it may stay the same or change. Not sure yet. Another advantage of a grid system is in multiplayer, Someone asks me where to build something, I can just tell them some grid cells and they can do whatever they like in there with zero risk of adversely impacting future undecided build plans, so that is another reason I use a strict grid as it keeps the peace
-
- Burner Inserter
- Posts: 14
- Joined: Sun Dec 22, 2024 5:38 pm
- Contact:
Re: 176 to 176 belt balancer
yes, sorry about the size, i'm actually loading from both sides with legendary stack inserter. 4 lanes are actually enough per unit, so here is how one of the unit is build, the total factory has 22 unit on each side, the issue is the size, 4 wagon is the max limit before having an issue with the otherside, as you may see on the last picture i posted, i barely have space to fit both exit and entrance, also due to the amount of train i can't making nay crossing without reducing throughput due to traffic block, so 4 wagon is the limit
AND i just noticed a flaw, i'll fix it
AND i just noticed a flaw, i'll fix it
- Attachments
-
- 12-24-2024, 00-46-46.png (341.44 KiB) Viewed 619 times
-
- 12-24-2024, 00-46-11.png (414.29 KiB) Viewed 619 times
-
- Burner Inserter
- Posts: 14
- Joined: Sun Dec 22, 2024 5:38 pm
- Contact:
Re: 176 to 176 belt balancer
here, fixed the issue, seem my math was slightly off, took a slightly longer than 48 sec to fill a train, i'd say about 50 second
(sent a video of the unit working)
(sent a video of the unit working)
- Attachments
-
- 12-24-2024, 01-09-58.png (241.91 KiB) Viewed 614 times
-
- single unit.mp4
- (22.05 MiB) Downloaded 9 times
Re: 176 to 176 belt balancer
Those legendary inserters will be wasted on that - several of them will be starved and end up idle.
Try these - you will need to alter for your space/belt arrangement. It isnt as compact, but can be made more compact.
They will fill/empty a carriage in around 6 seconds assuming space in trains/buffers. Replicate for an entire train in around that time.
Its uses a box buffer, rare stack inserters on the belt side of the boxes, legendary on the train side of the boxes. The box buffed will allow the boxed to continually fill at belt rates even when no train present. The legendary inserters will very quickly fill a train when present, much faster than direct belt feed.
Reverse it all when emptying.
6 rare on a split/merged belt seems to be all you need to fill/empty a stacked green belt. This isnt directly usable, it just simple load on one side and unload on the other for illustration.
Likely searches though old post would yield some more evenly distributed load/unload patterns as these are simple and will not fill/empty the boxes evenly, just symmetrically. In practice the difference only really apply when machines/trains are not adequate.
I usually use 8 cargo wagon trains for bulk items/fluid, so that would be 16 belts. Stations would need at least one stack space for 1 train stopped while another waits.
Try these - you will need to alter for your space/belt arrangement. It isnt as compact, but can be made more compact.
They will fill/empty a carriage in around 6 seconds assuming space in trains/buffers. Replicate for an entire train in around that time.
Its uses a box buffer, rare stack inserters on the belt side of the boxes, legendary on the train side of the boxes. The box buffed will allow the boxed to continually fill at belt rates even when no train present. The legendary inserters will very quickly fill a train when present, much faster than direct belt feed.
Reverse it all when emptying.
6 rare on a split/merged belt seems to be all you need to fill/empty a stacked green belt. This isnt directly usable, it just simple load on one side and unload on the other for illustration.
Likely searches though old post would yield some more evenly distributed load/unload patterns as these are simple and will not fill/empty the boxes evenly, just symmetrically. In practice the difference only really apply when machines/trains are not adequate.
I usually use 8 cargo wagon trains for bulk items/fluid, so that would be 16 belts. Stations would need at least one stack space for 1 train stopped while another waits.
Re: 176 to 176 belt balancer
I apologize, but I question the whole setup. What is the reason you're creating 2 mio/min green circuits on one spot and transport them away? You will have similar throughput/balancing issues at the receiving side, and you waste a huge part of the space just for collecting/balancing/distributing/filling things and containers. In my opinion, you're going the wrong approach this way. You're trying to solve problems you would not get with a different approach.
Usually, the most part of the green circuits are used to build red and blue circuits. So, if you also create red and blue circuits on the spot (use direct insertion builds) and transport these, you have to transport a fraction of what you're currently transporting.
You need to supply plastic bars (and for blue circuits sulfuric acid), but if you solve that challenge, you're able to reduce the output to 1/2 for every red circuit and 1/24 for every blue circuit you create. For every 24 trains you employ for green circuits to create blue circuits somewhere else, you only need 1 train.
You also seem to completely ignore quality. Everything seems to be of normal quality. By using quality with the assemblers, beacons and modules, the throughput of one machine increases tenfold. It makes everything smaller, because you need 10 times less machines. I recommend you focus on a factory creating quality production items, not on balancing a massive amount of items you will never even need in this amount if you use items with higher quality. That's what quality is mainly for.
If you're on Vulcanus and don't want to consume precious coal for plastic, even more consider quality and productivity research. If you use quality items consequently with mining and all downstream oil and cracking production, you will not really consume up your precious coal fields on Vulcanus any more - they become practically infinite as well. Or transport coal in instead of transporting green circuits out. Use cryogenic plants with lots of productivity modules to create plastic.
Sooner or later you will get into UPS problems if you continue to build setups that huge. If you do it big here, you're doing it big everywhere else as well, and this will overwhelm the game engine sooner or later. And for this again quality is made for: more throughput with less machines.
Last but not least, there is a foundry recipe for directly casting copper wire. It's more efficient to directly cast copper wire instead of casting copper plates, then creating copper wire in an assembling machine - even with production modules. Even if you have infinite molten copper on Vulcanus, that's one assembling machine less to feed and extract items from.
Usually, the most part of the green circuits are used to build red and blue circuits. So, if you also create red and blue circuits on the spot (use direct insertion builds) and transport these, you have to transport a fraction of what you're currently transporting.
You need to supply plastic bars (and for blue circuits sulfuric acid), but if you solve that challenge, you're able to reduce the output to 1/2 for every red circuit and 1/24 for every blue circuit you create. For every 24 trains you employ for green circuits to create blue circuits somewhere else, you only need 1 train.
You also seem to completely ignore quality. Everything seems to be of normal quality. By using quality with the assemblers, beacons and modules, the throughput of one machine increases tenfold. It makes everything smaller, because you need 10 times less machines. I recommend you focus on a factory creating quality production items, not on balancing a massive amount of items you will never even need in this amount if you use items with higher quality. That's what quality is mainly for.
If you're on Vulcanus and don't want to consume precious coal for plastic, even more consider quality and productivity research. If you use quality items consequently with mining and all downstream oil and cracking production, you will not really consume up your precious coal fields on Vulcanus any more - they become practically infinite as well. Or transport coal in instead of transporting green circuits out. Use cryogenic plants with lots of productivity modules to create plastic.
Sooner or later you will get into UPS problems if you continue to build setups that huge. If you do it big here, you're doing it big everywhere else as well, and this will overwhelm the game engine sooner or later. And for this again quality is made for: more throughput with less machines.
Last but not least, there is a foundry recipe for directly casting copper wire. It's more efficient to directly cast copper wire instead of casting copper plates, then creating copper wire in an assembling machine - even with production modules. Even if you have infinite molten copper on Vulcanus, that's one assembling machine less to feed and extract items from.
-
- Burner Inserter
- Posts: 14
- Joined: Sun Dec 22, 2024 5:38 pm
- Contact:
Re: 176 to 176 belt balancer
i am using quality, for beacons, modules and assembler and foundery, also that setup was just for the fun of pushing the game to it's theoretical limit and solving the issues to make it viable, i find it fun to challenge myself in these ways. i don't think anyone would ever use this in a serious playthrough, i did made a way smaller blueprint that does all 3 kinds of circuit if you wish to see it.
also for the copper wire, casting directly into copper wire makes it pass by only the foundry productivity modules, while if it passes by the electromagnetic plant make it have also the plants productivity, making even more wire from the same amount of copper, i did the test myself
also for the copper wire, casting directly into copper wire makes it pass by only the foundry productivity modules, while if it passes by the electromagnetic plant make it have also the plants productivity, making even more wire from the same amount of copper, i did the test myself
- Attachments
-
- 12-24-2024, 17-38-47.png (1.34 MiB) Viewed 442 times
Last edited by Globetrotteur on Tue Dec 24, 2024 5:14 pm, edited 2 times in total.
-
- Burner Inserter
- Posts: 14
- Joined: Sun Dec 22, 2024 5:38 pm
- Contact:
Re: 176 to 176 belt balancer
here is one section of the giant 2mil factory, with component list since it's kinda hard to see when zoomed all the way out
- Attachments
-
- 12-24-2024, 17-49-36.png (411.62 KiB) Viewed 434 times
Re: 176 to 176 belt balancer
Probably a stage people go through to find the game's limits and their own limits of patience - I was tempted to comment on the whole thing, but thought better of it - sometimes it is just better to let people find their own way, make their own mistakes etc. For some it is the journey that mattersTertius wrote: ↑Tue Dec 24, 2024 1:01 pm I apologize, but I question the whole setup. What is the reason you're creating 2 mio/min green circuits on one spot and transport them away? You will have similar throughput/balancing issues at the receiving side, and you waste a huge part of the space just for collecting/balancing/distributing/filling things and containers. In my opinion, you're going the wrong approach this way. You're trying to solve problems you would not get with a different approach.
Usually, the most part of the green circuits are used to build red and blue circuits. So, if you also create red and blue circuits on the spot (use direct insertion builds) and transport these, you have to transport a fraction of what you're currently transporting.
You need to supply plastic bars (and for blue circuits sulfuric acid), but if you solve that challenge, you're able to reduce the output to 1/2 for every red circuit and 1/24 for every blue circuit you create. For every 24 trains you employ for green circuits to create blue circuits somewhere else, you only need 1 train.
You also seem to completely ignore quality. Everything seems to be of normal quality. By using quality with the assemblers, beacons and modules, the throughput of one machine increases tenfold. It makes everything smaller, because you need 10 times less machines. I recommend you focus on a factory creating quality production items, not on balancing a massive amount of items you will never even need in this amount if you use items with higher quality. That's what quality is mainly for.
If you're on Vulcanus and don't want to consume precious coal for plastic, even more consider quality and productivity research. If you use quality items consequently with mining and all downstream oil and cracking production, you will not really consume up your precious coal fields on Vulcanus any more - they become practically infinite as well. Or transport coal in instead of transporting green circuits out. Use cryogenic plants with lots of productivity modules to create plastic.
Sooner or later you will get into UPS problems if you continue to build setups that huge. If you do it big here, you're doing it big everywhere else as well, and this will overwhelm the game engine sooner or later. And for this again quality is made for: more throughput with less machines.
Last but not least, there is a foundry recipe for directly casting copper wire. It's more efficient to directly cast copper wire instead of casting copper plates, then creating copper wire in an assembling machine - even with production modules. Even if you have infinite molten copper on Vulcanus, that's one assembling machine less to feed and extract items from.