Ok so it's been a very long time since I've taken a math class. This took a while but with the help of my good friend who has a PhD in a STEM field, I'm a college drop out, we came up with this iterative math equation. X is the output and Y is the initial input: X_n = (x_(n-1) + y)/6 I asked ChatGPT what the limit would be if y=100 as the input value like I used in the video. "what is the limit of x_n = (x_(n-1)+100)/6?" The limit of the sequence = 20 Unfortunately the math doesn't copy and paste well here so you will have to verify yourself. If there is enough interest I can try my best at explaining this in a future video.
I will watch this video every single time I need one of these, until I finally understand it, or just have to accept I will never understand it, but be grateful someone else does.
Generally I find it best to use the smallest whole numbers of the ratios involved and assume that the system has converged. To work out an arrangement that distributes a single input to 5 equal outputs, make each output 1, which means the input will be 5. 5 can't be divided by multiples of 2 or 3 to give 1, so a merged loop back is required. With a loop back of 1 the merged output will be 6, which can be divided as 2 x 3 way splits or 3 x 2 way splits to give 6 lines of 1, one of which provides the loop back and the remaining 5 of 1 the required outputs. The same numbers for the non bottleneck version. The 5 input is split to 2 lines of 2.5, the loop back of 1 split for 2 lines of .5, each 2.5 and .5 merged to give 2 lines of 3, each split 3 ways to give 6 lines of 1, again, one of which provides the loop back and the remaining 5 of 1 the required outputs.
The best explanation of 5-way splits I've seen so far. =) Seems like some people in the comments didn't watch the video all the way through for the non-bottleneck version of the 5-way split.
This is so helpful, I literally just updated my rotor production for use with steel screws and the output numbers required a 5 way split for the inputs. Thanks for the very clear tutorial!
I would love to see a 2 to 5 split, as condensed as possible. Hear me out, Industrial Storage Container, two outputs to five dimensional depots. Easy way would be 1 to 2, and 1 to 3 separate, but just thought it'd be neat to see it all combined.
Hello, love your videos and specially the bluepritns, thanks for shring for free. Only thing is, I fiind extremely difficult to find the input & outputs for the resources in the machines, you could add signs or colour them in a llamative way
Thanks so much! I totally agree. I made them and even I have a hard time finding the right input or output at times. A simple sign with an arrow would go a long ways. Thanks for the suggestion!
I’m glad you said this, these are the videos I’m planning on making next. They will have their own playlist and folder in the Google drive to hopefully keep things organized.
I see many problems with the "back loop" on a load balancer. 1) If you don't limit your primary input at (conveyor line item/min max - 1/6 of conveyor line item/min max), it will not work. Example : if you input 101 instead of 100. The system will jam at one moment and the back loop will stop having a function of balancing the load. 2) Your back loop need to feed the 1/6 of items at the same speed as the input conveyor belt at the merger otherwise the conveyor belt after the merger will not have a stable total item/min and when it fluctuates so will the load balancing The only good 5 ways split requires much more parts than that and has no "back loop" Let's say that it's a work around to make it function with less parts.
I'm guessing the values used in method 1 were to show a possible bottleneck issue, method 2 overcomes that. Anything within a load balancer assembly is always going to a be a fraction of the incoming based on its speed, so isn't a "problem". Splitters give the option of splitting to 2 or 3, with combinations of splitters giving multiples of these. Using a loop back is the only way ratios involving prime numbers can be done because they have no factors other that themselves or 1.
YOUR MATH IS TEH VOODOO MAGICS UND EBILS!!!!!!!! 💩☠🤡 Hey, this got me to thinking about the merger on prime balancers. You have the odd out leg coming back to the input. Fine. But, if the input belt is at capacity, does the merger stagger the input for the odd out? Your iteration suggests that it does. Would a next higher belt for a few meters just before the main input leaving a gap be better than the addition of another merger and splitter? Don't forget to show your work! 🎓👍😁
Hahaha the voodoo of fractions! Good question! I think if the incoming belt is at capacity then the next belt after the merger would have to be the next speed up. The second method would be the only way I know of to split a full 1200 belt.
It's a complete bollocks. This splitter can only work at the very beginning of the game. After 10-15 hours of play you will be feeding so much ore/details to one incoming branch that the return branch returning to the main branch will slow down your production or mining located on the main branch.
There are those of us who play the game to exercise extreme levels of efficiency. We play the game very differently from you. There's nothing wrong with the way we play the game. If there is something wrong with the way we play the game, there is something wrong with the way you play the game.
Ok so it's been a very long time since I've taken a math class. This took a while but with the help of my good friend who has a PhD in a STEM field, I'm a college drop out, we came up with this iterative math equation.
X is the output and Y is the initial input:
X_n = (x_(n-1) + y)/6
I asked ChatGPT what the limit would be if y=100 as the input value like I used in the video.
"what is the limit of x_n = (x_(n-1)+100)/6?"
The limit of the sequence = 20
Unfortunately the math doesn't copy and paste well here so you will have to verify yourself.
If there is enough interest I can try my best at explaining this in a future video.
I will watch this video every single time I need one of these, until I finally understand it, or just have to accept I will never understand it, but be grateful someone else does.
Generally I find it best to use the smallest whole numbers of the ratios involved and assume that the system has converged.
To work out an arrangement that distributes a single input to 5 equal outputs, make each output 1, which means the input will be 5. 5 can't be divided by multiples of 2 or 3 to give 1, so a merged loop back is required. With a loop back of 1 the merged output will be 6, which can be divided as 2 x 3 way splits or 3 x 2 way splits to give 6 lines of 1, one of which provides the loop back and the remaining 5 of 1 the required outputs.
The same numbers for the non bottleneck version. The 5 input is split to 2 lines of 2.5, the loop back of 1 split for 2 lines of .5, each 2.5 and .5 merged to give 2 lines of 3, each split 3 ways to give 6 lines of 1, again, one of which provides the loop back and the remaining 5 of 1 the required outputs.
The best explanation of 5-way splits I've seen so far. =) Seems like some people in the comments didn't watch the video all the way through for the non-bottleneck version of the 5-way split.
This is so helpful, I literally just updated my rotor production for use with steel screws and the output numbers required a 5 way split for the inputs. Thanks for the very clear tutorial!
I’m glad it was useful! Thanks!
Best 5 way split load balancer explanation I've seen! Well done sir.
I would love to see a 2 to 5 split, as condensed as possible.
Hear me out, Industrial Storage Container, two outputs to five dimensional depots. Easy way would be 1 to 2, and 1 to 3 separate, but just thought it'd be neat to see it all combined.
That is a really interesting idea! I'm going to give that some thought and see if I can come up with anything, thanks!!
@@moo121 You've already got a "method 1" version. Remove the input splitter in method 2 and the inputs go direct to each merger.
I also have a 2 to 5 non bottleneck version. The splitters and mergers occupy 3.5x1.5 tiles - is this considered compact enough?
That is a great solution! Thanks again for helping me with that method 2!!
The video i need, thanks!
Hello, love your videos and specially the bluepritns, thanks for shring for free. Only thing is, I fiind extremely difficult to find the input & outputs for the resources in the machines, you could add signs or colour them in a llamative way
Thanks so much! I totally agree. I made them and even I have a hard time finding the right input or output at times. A simple sign with an arrow would go a long ways. Thanks for the suggestion!
need a video showing common prime load splitters, like the 1/6th 5/6th you briefly showed at 8:00, i know ive seen in your vids 7/15th, 8/ 15ths etc
I’m glad you said this, these are the videos I’m planning on making next. They will have their own playlist and folder in the Google drive to hopefully keep things organized.
I'm a manifold guy myself but this is an excellent demonstration of belt balancing! Well done!
I see many problems with the "back loop" on a load balancer.
1) If you don't limit your primary input at (conveyor line item/min max - 1/6 of conveyor line item/min max), it will not work. Example : if you input 101 instead of 100. The system will jam at one moment and the back loop will stop having a function of balancing the load.
2) Your back loop need to feed the 1/6 of items at the same speed as the input conveyor belt at the merger otherwise the conveyor belt after the merger will not have a stable total item/min and when it fluctuates so will the load balancing
The only good 5 ways split requires much more parts than that and has no "back loop"
Let's say that it's a work around to make it function with less parts.
I'm guessing the values used in method 1 were to show a possible bottleneck issue, method 2 overcomes that. Anything within a load balancer assembly is always going to a be a fraction of the incoming based on its speed, so isn't a "problem".
Splitters give the option of splitting to 2 or 3, with combinations of splitters giving multiples of these. Using a loop back is the only way ratios involving prime numbers can be done because they have no factors other that themselves or 1.
@gamer61k3 infinity :D
YOUR MATH IS TEH VOODOO MAGICS UND EBILS!!!!!!!! 💩☠🤡
Hey, this got me to thinking about the merger on prime balancers.
You have the odd out leg coming back to the input. Fine. But, if the input belt is at capacity, does the merger stagger the input for the odd out? Your iteration suggests that it does.
Would a next higher belt for a few meters just before the main input leaving a gap be better than the addition of another merger and splitter? Don't forget to show your work! 🎓👍😁
lol @ "Don't forget to show your work!"
Hahaha the voodoo of fractions!
Good question! I think if the incoming belt is at capacity then the next belt after the merger would have to be the next speed up. The second method would be the only way I know of to split a full 1200 belt.
It's a complete bollocks.
This splitter can only work at the very beginning of the game.
After 10-15 hours of play you will be feeding so much ore/details to one incoming branch that the return branch returning to the main branch will slow down your production or mining located on the main branch.
There are those of us who play the game to exercise extreme levels of efficiency. We play the game very differently from you. There's nothing wrong with the way we play the game. If there is something wrong with the way we play the game, there is something wrong with the way you play the game.
@Andrew_EfimOvich Did you watch the video? The return branch is only being returned to the incoming branch, it's not affecting anything further back.