I've updated the blueprint due to a brilliant commenter jarring my brain. factoriobin.com/post/uf60id The belt sushi is now more accurate and more compact design. Thanks!
Decent tutorial. Some thoughts: 1. To have precise amounts being added to the belt, add an arithmetic combinator to each inserter that subracts the current count from desired amount and outputs a signal that the inserter uses to set the stack amount. The inserter will then transfer the maximum amount until less than that is needed, whereupon it will transfer the exact amount needed to match the desired amount in one final pull. 2. For higher throughput, I just use splitters pulling from feed belts, where the belt right before the splitter only turns on when more items are needed. It isn''t as precise, but can dump up to 60 items per second onto the belt. (I also prioritize adding new items to the belt, holding up prior items as necessary.) 3. I also add lane balancers at the sushi section's input and output to ensure that one lane doesn't jam. 4. For more throughput, you can run multiple belts. This multiplies the chance that the desired item will be available as soon as needed. (i.e. instead of 8 chances, you have 16, etc.) You can get *REALLY* crazy throughput if you weave belts. For example, weaving the first three tiers will provide 90 items per second, all of which can be accessed by bulk/stack inserters. Placing three inserters--pulling from each belt colour--will triple the chance of the desired item being available on top of potentially tripling the quantity pulled. 5. Running the belt back past the labs/etc. will give them a second chance at pulling the items needed. 6. That logistics version is very clever but I can't see it keeping up with later-stage science. (The stack size trick would work well with it.) Cheers!
I wish there was circuit wire option which would return "belt leftover capacity" or "belt total capacity" for the "hold belt" area, so you could just do the item calculations using math combinators.
That would mean the inserters have to be on for signals that are below zero, which afaik doesn't work for setting filters. (Unless you're doing some integer overflow shenanigans; but the 2 extra spaces for the combinator and the 1 tick signal delay that you'd avoid are not worth _that_ headache.)
you could cut the decider combinators by using filters an the inserters and enable/disable them based on the capacity. i probably would still use a constant combinator to be more flexible in case you need to expand your science setup
yo dang SvenWM!! Great Idea! I set 1 of each of the inserters to filter only one type of science, then I set the enable condition being that type of science > 0, AND I set the stack size to the same science signal, so it won't ever pick up more than is needed. Now it's even more compact, simple, and accurate. Brilliant! Now this video is old news lmaoooo
This is good, but I just use a sideloader belt setup.. I check the content of the single belt tile being loaded + one tile after it, and enable the sideloading belt if it doesn't contain the pack you're loading. No inserters, no capacity-specific logic. So in short, just add flasks when the sushi belt has a gap.
@@TheJarGames Unfortunately I can't give you a link, the URL gets my post deleted. All I can say is, take a standard sideload belt setup. Wire up three belt tiles. Two on the sushi belt, one on the sideloader. Enable the sideloader belt if (sci) == 0. Read+hold the two tiles of sushi belt. Now it only runs if there's a gap. Sorry I can't show my example, but I swear it is very simple.
@@creamydonk yeah I think I understand but it seems like you would need one belt for each science which doubles the input belts... anyways you could post the link but break it up like factoriobin . com / post/2345lkj45 so that it doesn't get deleted and I can construct it and view it. all good though either way, I appreciate the idea just wanted to see it in action
@@TheJarGames Ah good point. In that case I put it on factoriobin, the post id is g279nu And yes, one belt per thing. That is how I usually do things though, since I don't mix the outputs until they get to the sushi ring. I make a fairly huge ring behind the base and just tap everything into it.
Because of filters: "inserters will only put stuff out of the box that's in the box..." - would like to see the inserters putting things out of the box that are not in the box... 🧐🤭
you can make a parameterized BP which changes the capacity amounts relative to your currentbelt tech lvl, research spd lvl, and also amount of labs being fed too, and feed it into the new combinators to make a dynamic sushi belt which overstocks thebelt or understocks it with ceetain sciences relative to what is beingconsumed and in what quantities
Glad you enjoyed it David! Thing is, it only really matters what your capacity is - based on your belt length in your specific setup. The length of your belt and what inserters you use is really the only things that matter. Join up the discord and let's continue the conversation! Maybe you can take the blueprint and parameterize it to improve it :)
You can do this type of thing in vanilla, but I enable a mod called "editor tools" and start a game in the testing environment. feel free to join discord for more help
2 sciences per belt, several belts all leading to one head lab, then that lab passes the sciences down to the next in the lab chain to the rest. I don't see why anyone needs a setup other than this.
This chain filling often leads to heavy lab flickering which signals the research process slowing down due to science shortage for small intervals. It is only a good choice because it is easier to build and takes up less space. If you can do it, it is always better to directly feed each lab.
@@narancs5 yeah 100% - I am actually not really a fan of sushi belts, and I don't use them in my gamesaves... but I felt like trying it out from a theory perspective
I liked your throughput test videos but I feel like this tutorial is not good. The belt needs red science? Good grab a green one. All those combinators doing nothing that could not be set on the inserter or belt itself. My early game was simple, belt disabled if sushi ring contents for that science is greater than a value S from constant combinator, since the labs will expand inevitably. Every second science is sideloaded on the other side of the belt with either underground or inserter.
I've updated the blueprint due to a brilliant commenter jarring my brain.
factoriobin.com/post/uf60id
The belt sushi is now more accurate and more compact design.
Thanks!
Decent tutorial. Some thoughts:
1. To have precise amounts being added to the belt, add an arithmetic combinator to each inserter that subracts the current count from desired amount and outputs a signal that the inserter uses to set the stack amount. The inserter will then transfer the maximum amount until less than that is needed, whereupon it will transfer the exact amount needed to match the desired amount in one final pull.
2. For higher throughput, I just use splitters pulling from feed belts, where the belt right before the splitter only turns on when more items are needed. It isn''t as precise, but can dump up to 60 items per second onto the belt. (I also prioritize adding new items to the belt, holding up prior items as necessary.)
3. I also add lane balancers at the sushi section's input and output to ensure that one lane doesn't jam.
4. For more throughput, you can run multiple belts. This multiplies the chance that the desired item will be available as soon as needed. (i.e. instead of 8 chances, you have 16, etc.) You can get *REALLY* crazy throughput if you weave belts. For example, weaving the first three tiers will provide 90 items per second, all of which can be accessed by bulk/stack inserters. Placing three inserters--pulling from each belt colour--will triple the chance of the desired item being available on top of potentially tripling the quantity pulled.
5. Running the belt back past the labs/etc. will give them a second chance at pulling the items needed.
6. That logistics version is very clever but I can't see it keeping up with later-stage science. (The stack size trick would work well with it.)
Cheers!
I wish there was circuit wire option which would return "belt leftover capacity" or "belt total capacity" for the "hold belt" area, so you could just do the item calculations using math combinators.
that would be awesome, yeah "read belt capacity" signal
You don’t need to use the *-1 arithmetic combinator, the constant combinator can also multiply by -1
That would mean the inserters have to be on for signals that are below zero, which afaik doesn't work for setting filters.
(Unless you're doing some integer overflow shenanigans; but the 2 extra spaces for the combinator and the 1 tick signal delay that you'd avoid are not worth _that_ headache.)
Great video, Jar🎉
nice vod, ty for tips
you could cut the decider combinators by using filters an the inserters and enable/disable them based on the capacity. i probably would still use a constant combinator to be more flexible in case you need to expand your science setup
yo dang SvenWM!!
Great Idea!
I set 1 of each of the inserters to filter only one type of science, then I set the enable condition being that type of science > 0, AND I set the stack size to the same science signal, so it won't ever pick up more than is needed.
Now it's even more compact, simple, and accurate. Brilliant! Now this video is old news lmaoooo
This is good, but I just use a sideloader belt setup.. I check the content of the single belt tile being loaded + one tile after it, and enable the sideloading belt if it doesn't contain the pack you're loading. No inserters, no capacity-specific logic. So in short, just add flasks when the sushi belt has a gap.
So much this, inserters are unnecessary.
@@AlexanderBukh have a blueprint with this design? I would like to see it
@@TheJarGames Unfortunately I can't give you a link, the URL gets my post deleted. All I can say is, take a standard sideload belt setup. Wire up three belt tiles. Two on the sushi belt, one on the sideloader. Enable the sideloader belt if (sci) == 0. Read+hold the two tiles of sushi belt. Now it only runs if there's a gap. Sorry I can't show my example, but I swear it is very simple.
@@creamydonk yeah I think I understand but it seems like you would need one belt for each science which doubles the input belts...
anyways you could post the link but break it up like
factoriobin . com / post/2345lkj45
so that it doesn't get deleted and I can construct it and view it.
all good though either way, I appreciate the idea just wanted to see it in action
@@TheJarGames Ah good point. In that case I put it on factoriobin, the post id is g279nu
And yes, one belt per thing. That is how I usually do things though, since I don't mix the outputs until they get to the sushi ring. I make a fairly huge ring behind the base and just tap everything into it.
Thank you for making this. Not many videos on sushi belts, and the few that you find are not very clear.
Always happy to help!
Because of filters: "inserters will only put stuff out of the box that's in the box..." - would like to see the inserters putting things out of the box that are not in the box... 🧐🤭
Looks good.
Thanks Fumbler! The blueprint I added actually is even better than the video... but yeah,
you can make a parameterized BP which changes the capacity amounts relative to your currentbelt tech lvl, research spd lvl, and also amount of labs being fed too, and feed it into the new combinators to make a dynamic sushi belt which overstocks thebelt or understocks it with ceetain sciences relative to what is beingconsumed and in what quantities
Glad you enjoyed it David!
Thing is, it only really matters what your capacity is - based on your belt length in your specific setup.
The length of your belt and what inserters you use is really the only things that matter.
Join up the discord and let's continue the conversation!
Maybe you can take the blueprint and parameterize it to improve it :)
Nice setup!
Are you playing a mod or can you configure something in the world settings to get to this sort of editor?
You can do this type of thing in vanilla, but I enable a mod called "editor tools" and start a game in the testing environment. feel free to join discord for more help
I love SUSHI!!!
Now were thinking with FISH
edit. traincar buffering science? hrmmmmm :p
I'm going to send you a traincar of fish for your birthday
@@TheJarGames yoo!
2 sciences per belt, several belts all leading to one head lab, then that lab passes the sciences down to the next in the lab chain to the rest. I don't see why anyone needs a setup other than this.
because, sushi
This chain filling often leads to heavy lab flickering which signals the research process slowing down due to science shortage for small intervals. It is only a good choice because it is easier to build and takes up less space. If you can do it, it is always better to directly feed each lab.
@@narancs5 yeah 100% - I am actually not really a fan of sushi belts, and I don't use them in my gamesaves... but I felt like trying it out from a theory perspective
I liked your throughput test videos but I feel like this tutorial is not good. The belt needs red science? Good grab a green one. All those combinators doing nothing that could not be set on the inserter or belt itself. My early game was simple, belt disabled if sushi ring contents for that science is greater than a value S from constant combinator, since the labs will expand inevitably. Every second science is sideloaded on the other side of the belt with either underground or inserter.