Challenge: belt balancers

[quinor]

While discussion in topic https://forums.factorio.com/forum/vie ... f=8&t=9071 I came with an idea of new contest: building effective belt balancers for multi-belt systems. If it goes well this topic will become quite nice knowledge base of good designs, but my intention is just a good puzzle for competitors.

I'd like to see designs in 2 categories:

• Compressed belts input
• Uncompressed belts input

First one is about designs, where on input there is some amount of fully compressed belts and the design should preserve this full compression. Second one is less restrictive, it allows decompression caused by single-belt turns (inner belt speed lose). It will be fixed anyway in one of future versions and this kind of design is still useful sometimes anyway.

Obviously, we'd like to optimize space which is used (both width and length), but I'd also like to see maybye less efficient, but easier to remember designs.

Some other rules (but it's more like a convention I'd suggest, feel free to add even completely different designs):

• Input should be all parallell with no space between them, output in the same style.
• Input and output should be in one axis, so you can build balancer on top of existing straight line

There is a chance I'll ad more convention suggestions as progression of contest goes.

About naming: X-Y balancer takes X inputs and gives Y even ouptuts.

If you want you can also post an explanation how specific balancer is designed, so it's easier to analyse by other players.

And finally, my own entries:

16-16 balancer

It's one of my early designs, it's for uncompressed input. I'm not proud of 4-tile shift, but I couldn't get rid of it. 24x35 in size.

8-8 balancer

Not very symmetric and intuitive, for uncompressed input, but quite small - I doubt you can build much smaller. Parallel inputs and outputs, 9x13 in size.

4-4 balancer

I'm very much sure it's the smallest possible 4-4 balancer. Often part of my bigger designs. It's 100% compressed, with 4x7 footprint.

[dee-]

I'll start by throwing in an old design of mine, too
Nothing too spiffy and totally not space efficient but it can not only balance belts but also their lanes. Nice when you have an uneven consumption on these you want to even out.
Can easily be upgraded to double blue.


This is the rebalancer I use for my standard "two-lanes-of-fast-belts"-system. (Edit: bad style, use the ones from the posts above or below)

image

It is able to rebalance every single one of the incoming four lanes equally onto the four outgoing lanes:

image

If one or more of the four outgoing lanes are blocked the other not blocked lanes still get filled with more items.

It also preserves the full item throughput if no rebalancing is needed:

image

Red corners are needed to keep the full thoughput of the yellow lanes.
(I think the red corners in the center merging area might be downgraded to yellow as the throughput there is only 1/2 yellow belt. I haven't tested it yet because it just works However keep in mind 180° u-turns behave differently when they side-feed onto a belt)

Red belts on the field where left and right belts merge are needed to have the side-merges working. Yellow belts would be too slow to get all incoming items on the merged belt.

The rebalancer also only needs equal or lower technology of your usual belt system.

[MadZuri]

I'll start by throwing in an old design of mine, too

Those are line balancers, not belt balancers. Two belts is literally the easiest design to "belt balance".... can be done with a single splitter with no turns.
If I can banish one design in this game forever, it is that line reblancer design. It is horrible and I hate it. (it's the design I hate, not you, dee-; just thought I'd make that clear) I consider it a failure if I need a line balancer. But, I do have designs for those as well:

small balancers

And before anyone chimes in that those aren't "true" line balancers because they don't evenly split the sides of the line.... you will NEVER need a true line balancer. What matters is belt throughput. Does the input of the balancer clog if one side of the line (for line balancer) or one belt (for belt balancer) is completely backed up? That mass tangle of belts is big, wasteful, and ugly.

What I linked was a 2-lane and 4-lane combined belt/line balancer. Please, for the love of god, use this instead of that horrible knot.

[dee-]

I'm really not fond of that brute-force design myself and do not use it anymore.
I switched for OP's 4-belt-balancer, it's really good (kudos to psorek!) and will try your design if I find the time (hopefully over the upcoming 4-day-long-weekend here).

[aka13]

What is the point of these contraptions? Usually you just have a consumer eating 1 line away, Then, add a splitter at that consumer which would balance 1 line to the other line. Is it for beauty purposes only?

[quinor]

I'm mainly using it on the output of rail stations, where I need to even big amounts of belts to split them furher and direct into different parts of the factory, but I can easily imagine different uses like transporting stuff with multiple parallell belts.

[MadZuri]

16

This design assumes that each set of 4 lanes are already balanced, and balances the rest in a count-perfect manner. It is designed for 0.12 belt mechanics. Why sets of 4 lanes? So you can cross it with another set of belts.

[kdadrummer]

Those are line balancers, not belt balancers. Two belts is literally the easiest design to "belt balance".... can be done with a single splitter with no turns.
If I can banish one design in this game forever, it is that line reblancer design. It is horrible and I hate it. (it's the design I hate, not you, dee-; just thought I'd make that clear) I consider it a failure if I need a line balancer. But, I do have designs for those as well:
+ SMALL BALANCERS

And before anyone chimes in that those aren't "true" line balancers because they don't evenly split the sides of the line.... you will NEVER need a true line balancer. What matters is belt throughput. Does the input of the balancer clog if one side of the line (for line balancer) or one belt (for belt balancer) is completely backed up? That mass tangle of belts is big, wasteful, and ugly.

What I linked was a 2-lane and 4-lane combined belt/line balancer. Please, for the love of god, use this instead of that horrible knot.

that dosnt balance between the sides of belts the way he is asking for. lets say there are 8 inputs (2 sides per belt, 4 input belts), certain inputs only go to a few outputs as list below
inputs/outputs in picture listed left to right (1 2 3 4 5 6 7 8)
i got the rations by puting 100 items in one input and counting where they came out (cause i was board as hell and science nerd things)
*NOTE: I "primed" the balancer first buy running some items through it so any dead spots where an item would get stuck were filled before the tests where done
listed as follows

input #: output #-ratio of items recieved

input 1: 1-1/4 2-none 3-1/4 4-none 5- 3/20 6- 3/25 7- none 8-23/100
input 2: 1-none 2-1/4 3-none 4-1/4 5- 1/4 6-none 7-13/100 8-3/25
input 3: 1-1/4 2-none 3-1/4 4-none 5-3/25 6-13/100 7-1/4 8-none
input 4: 1- none 2-1/4 3-none 4- 1/4 5-1/4 6-none 7-3/25 8-13/25

at this point i became board because i say a pattern, as well as this layout is mirrored right to left, so these results will be mirrored for inputs 5 - 8.
as you can see, this does no accomplish the task the person requested initially.

it's also worth saying that it probably wont matter at all but this is far from a perfect balancer

[Qon]

decompression caused by single-belt turns (inner belt speed lose). It will be fixed anyway in one of future versions and this kind of design is still useful sometimes anyway.

Source? Or is this already fixed? I haven't noticed any slowdown or decompression from turns.

[aka13]

decompression caused by single-belt turns (inner belt speed lose). It will be fixed anyway in one of future versions and this kind of design is still useful sometimes anyway.

Source? Or is this already fixed? I haven't noticed any slowdown or decompression from turns.

It has been fixed, yes.