Merge/Split Conveyors & Mk.1 <--> Mk.2

[Dayradon]

If I use 2x Conveyor Mk.1 (60 units) to input into a merger, and have the output connected to a Conveyor Mk.2 (120 units), will it add the two inputs to a final output of 120 per stack, or will it just spit out 60's? Also conversely with a splitter, if I split a Conveyor Mk.2 will I get 40 per output?

Just trying to figure out if I can eliminate gaps etc.

[DroneBoat]

YES to all

[Grandaddypurple]

If both of your Mk1 are full and convey 60 parts/min and you merge them, then yes your mk2 will be full as well and convey 120 parts/min
Conversely, if you split a full Mk2 in three, you'll get an even 40 indeed

[Steeleye]

splitters and such are only as fast as the conveyor belts attached, so the splitters do not have a set speed

[Dayradon]

awesome ty all

[Dayradon]

ok. This makes optimization kinda wonky. Or at least the merger/splitters need to be reworked IMHO.

If I have a mine spitting out ore at 60/min and I attach mk.1 conveyors (60/min) I achieve 100% efficiency to destination.

Since smelters only have 30/min intake & 30/min ouput. I split the line using only two outputs and should achieve 30/min on each line to the smelter.

Iron plates consume 30/min of iron bars but only output 20/min. of plates.

If I run three of these setups (40 plates/min per line) I should achieve a final output of 120/min.

Merging the 3 lines should allow me to make use of a final MK.2 conveyor at the end going to storage/next production chain with 100% efficiency.

I tested this, and still see gaps between plates at the final output. If I'm producing 120/min and moving them at 120/min there should be no gaps.

I even tried backing up all the conveyors first to allow for the 60/min conveyors with only 40/min production. And there are still gaps in the final output which shouldn't be there.

It feels like the merger is operating at the input belt speed instead of the output belt speed. I suspect maybe the splitter could also be reversed and operating at the output belt speed rather than the input belts speed.

[Edmund Greyfox]

So why not just use the mk2 belts for all of it? I know you're trying to balance everything perfectly, but there really isn't any need to do that. If you're merging then there really isn't any issue with one of the input lines moving faster then what you can produce. All you really care about is that the final belt gets the full 120.

[Dayradon]

but that's the problem, the final belt has gaps. so anything at the end that could utilize 120/min (full production chain), will pause briefly every time a gap occurs

** EDIT: May need to clarify this better.

Output: 40 per minute
Belt speed 60/minute

60 segment long belt = 40 segments with product & 20 segments empty

If end point can accept 60/min, it will sit waiting 20s (total combined) every time it hits a gap
= 66% efficiency

[Grandaddypurple]

The fact that your belt isn't full doesn't mean it's not efficient
I'm having a hard time understanding your issue, I think you have a misconception of how belts are supposed to be used

[Air Bear]

feel like your answer has something to do with fractions of a second and your belts not being in sync with each other. like perhaps your items are hitting the belt a 1/3 of a second and 2/3 of a second but your last item isnt there yet, so the merger has nothing to merge and that makes a gap? I dont know if this is actually the case, I'm just guessing. Something about the producing 40/min strikes me as 'well that means that for 20 seconds, that belt is empty' and even though you're merging multiple belts together, you could end up with gaps because they are simply not in sync with each other.

[Edmund Greyfox]

That's why i said you should just use 120 belts for everything. If you are producing 120/minute and they flow down nothing but 120 belts then whatever is at the end is going to get 120/minute. Whether or not there are gaps on the belt is irrelevant as long as the total flow is what the machine needs.

[Dayradon]

@Grandaddypurple:
@ Edmund Greyfox

Think of it like this. A single belt segment (1 unit) is capable of moving 60 items over it in a minute or 1 item per second. If I only place 40 items on it in 60 seconds, then for 20 seconds it's spining with nothing on it. Sure it may only be 2 seconds at a time which is not alot.

If you take Iron rods for example.

Uses 1 ingot - which can fill easily 30/min - produces 1 rod every 4s (15/minute)

Turn into screw: 1 rod in - 4 screws out every 6s (40/min)

Turn into rotor: 25 screws per item needed - produces 4 items per minute (100 screws/min)

I'm only producing 40 screws/min of the needed 100/min.

I make 2 more screw lines (3 in total) = 120 screws per minute

Using a merger and a mk.2 conveyor on output, should get a constant stream of screws and actually 20 more than is required/min so it should pauses occasionaly as the assembler is full and needs to wait for room before it can load in another screw.

Instead, the line runs constantly and about every 13 screws is a 2 item gap which equates to 2s and over a minute is 8s down time.

In the assembler, 13 screws in, wait 2s over 15 seconds, 13 screws in + (still 2s gap), produce 1 item over 15s

during that 15s repeat above... but, due to 26 screws in and only 25 consumed, over 25 productions, you'll end up with 25 screws already loaded and no 2s gap (at end or front) before production starts on #26 (bonus production).

2 seconds every 15s, over 25 productions = 50s over 375 seconds or 6 mins 15 seconds.

That's a loss in productivity of 13.3%

a) I haven't even taken into account the 2nd ingredient - iron rods and it's loss

b) things catch up once the power goes down, due to the conveyors still moving but over enough time the gap eventually reappears.

[Grandaddypurple]

Originally posted by Dayradon:

Instead, the line runs constantly and about every 13 screws is a 2 item gap which equates to 2s and over a minute is 8s down time.

If your setup is made to produce 120 screws/min and you have gaps on your Mk2 belt after merging three mk1 belts conveying 40/min, then you have a problem somewhere upstream, let alone if it's dedicated for a machine consuming only 100/min. Monitor your constructors and see if one isn't starving of iron rods. If constructors are running fine, then your issue might be the following:
If you placed your merger by snapping it directly on an mk1 belt and then upgraded the output to mk2 (i.e you built the belt before the merger), there currently is a bug where a tiny piece of Mk1 is left inside the merger thus screwing your throughput.
If that's not the issue either, could you provide a top-down screenshot of your setup?

[Dayradon]

that bug you mentioned may be the cause. I tend to build sequentially forward. All planned out in head before I place anything. I'll see if that improves things.