Factorio Forums

I did some measurements and here are results (all AMs are without modules).
This produces ES at the speed of 96 items/minute:

And this produces ES at the speed of 264 items/minute:


Number of inserters is the same, except in later case 4 are red instead of blue, and it takes less space

It's very often discussed here, that copper wires on belts is a bad idea.

ssilk wrote:It's very often discussed here, that copper wires on belts is a bad idea.

Well, you can speed up the first case to the same speed as the second by putting 3 blue inserters in front of AMs that are making ESs.
But yeah, this scales extremely poorly.

I would think the 4 red inserters operating at 1 unit per second would limit the top amount of CKT production to 240 per minute.

Every time you put a wire on a belt god removes1 line of code for the next update.

^ if the above is true. We might not have a game very very soon.

ssilk wrote:It's very often discussed here, that copper wires on belts is a bad idea.

Generally yes. However, this is usually not a problem if you are using multiple inserters. If you use 3 inserters, you can, generally, move all the needs at once. Upgrade to fast/express belts and fast inserters, and things should chug along nicely. If you do it this way, you will potentially need to double-bandwith the wires for larger circuit layouts.

If you take the extra space to go belt->inserter->chest->inserter->factory, you can get the best of both worlds if you hook up a smart inserter on the output side to turn the assembly off (overproduction is bad, mmkay?) and allow wires to stockpile (which can also be throttled if it's a smart chest fed by a smart inserter on a local circuit network)

Im surprised the difference is so large.

I was under the impression that long-handed inserters are bad for factories due to their slower speed?

Indeed, iron input can become a bottleneck. Easily resolved by inputing the iron from the sides:

Nevermind the stray normal inserter. Pretend it's blue.

Penrif wrote:Indeed, iron input can become a bottleneck. Easily resolved by inputing the iron from the sides:

Nevermind the stray normal inserter. Pretend it's blue.

This looks like a great design, I'll go with this, thanks! If anyone has any other efficient layouts please poat

Personally, if I were doing a bot layout I would still have the iron on a belt, but I would have the assemblers on the copper side spaced one apart from each other with the circuits being taken out towards the copper side, (into a chest of course), with the iron on the opposite side. (Much closer to being the 2nd pic in OP, than the one Penrif posted.)
Otherwise, I usually aim for something like shown in the OP's last pic.

But in either case, the 2nd pic doesn't have iron as the bottleneck. It may have a problem with inserting copper with only one blue inserter, and it may have a problem with extracting the circuits if the AMs get module'd, but otherwise it's fine.

n9103 wrote:Personally, if I were doing a bot layout I would still have the iron on a belt, but I would have the assemblers on the copper side spaced one apart from each other with the circuits being taken out towards the copper side, (into a chest of course), with the iron on the opposite side. (Much closer to being the 2nd pic in OP, than the one Penrif posted.)
Otherwise, I usually aim for something like shown in the OP's last pic.

But in either case, the 2nd pic doesn't have iron as the bottleneck. It may have a problem with inserting copper with only one blue inserter, and it may have a problem with extracting the circuits if the AMs get module'd, but otherwise it's fine.

I am haivng trouble visualizing what you are describing, but i agree that iron plate isnt the botteneck in in the OPs second pic, the bottleneck would be the long-handed inserters being too slow, correct?

So what is the "best" electronic circuit design that is easily expandable and efficient?

Damn, was afraid my imagery wasn't up to the task.
Incoming screeenshots shortly.

silman wrote:I am haivng trouble visualizing what you are describing, but i agree that iron plate isnt the botteneck in in the OPs second pic, the bottleneck would be the long-handed inserters being too slow, correct?

So what is the "best" electronic circuit design that is easily expandable and efficient?

As far as extendibility goes, This variant --- with the iron supply between the wire assemblers and the circuit assemblers --- is what I've been experimenting with so far. I haven't really been able to put it through it's paces since smelting and belt throughput seem to be the bottlenecks for me.

Once I get the normal inserters replaced with fast inserters I am assuming here that the stack size bonus for moving from assembler to assembler will cancel out the reduced speed of long handed inserters.

I definitely prefer the OP's second pic as my preferred EC build. You can build multiple setups side by side without having to worry about or deal with splitters and underground belts to get iron to the Electronic Circuits, and in one of my factories, I mirrored the EC build on the other side of the Copper line. While I tend to only use 1 Long Handed Inserter to output Electronic Circuits, if I wanted to increase their output there is space for a second Long Handed inserter.

Alternatively, if you want multiple fast inserters to output Electronic Circuits, output them to the side. Leave a single tile between builds if putting them side by side, similar to Penrif's build, except flip your inputs and outputs.

The best build ultimately depends on how you plan to build your factory and where you need to get resources to (or where they are coming from). Fortunately there are many ways to do this.

I'm between games, so these are only mockups.
No modules. 2nd red inserter removing is a must to keep up with Mk2 AMs.
If you're using speed modules, then this wouldn't keep up.
But if you're using speed modules on general circuit production, your factory needs an overhaul.

Crud. I meant to test if that 2nd blue inserter for copper was required to keep up or not.
Anyways, there's what I was talking about.

n9103 wrote:I'm between games, so these are only mockups.
No modules.

With logistics

Without

2nd red inserter removing is a must to keep up with Mk2 AMs.
If you're using speed modules, then this wouldn't keep up.
But if you're using speed modules on general circuit production, your factory needs an overhaul.

As far as those long handed inserters go, how well do you think it would work if you replaced them with fast inserters and used underground belts to free up the path in front of them?

Crud. I meant to test if that 2nd blue inserter for copper was required to keep up or not.
Anyways, there's what I was talking about.

Given that this machine uses a 2:3 ratio of iron plates to copper plates, I would assume that, since one fast inserter per assembler is all that is needed for iron plates the same would be true of copper as well.

I'm sure it works better with the underground belts and blue inserters.
But it's overkill.
You're spending A LOT more if you try to speed up the process more than plain AM 2s can manage. There's few logical reasons for having to overclock such a simple process, and none I can really come up with, especially if logistic networks are in place. 2 ingredients in 2 steps. As simple as it gets other than 1-step items, and therefore highly expandable.

I suppose that's not really the point being made by some of the posters, but it's one I'll make.

As an aside, I would say that if you needed the most compact possible production, you would probably abandon the 3:2 ratio, and just make it linear.
There's an old Super Compact Blues thread somewhere here. It's outdated, since a couple recipes have been altered, but mostly still applies, just needs a little expansion. It's not fast at all, but it gets the job done in the smallest area possible before logistic bots. (And probably competes for the title even then.)

It's oxymoronic/ironic really, since the need for smallest possible isn't really useful until you have automated construction, but in order to get to the point where you'd be doing this, you've already done most of the research that needs blues. (Assuming non-easy settings)

I just checked. 2 red inserters per AM2 is more then enough to get EC to belt.
Replacing them with fast inserter placing thing sideways have no impact on production speed.

In case of AM2 iron is not even close to be a bottleneck.

As for AM3s... I'm not entirely sure that it's worth to use them for ECs.
When you really need to mass produce ECs, the space is usually not really an issue.
So, it's much more effective to setup another set of AM2s then bother with AM3s.

Speeding up production with modded AM3s for such a quick produced thing as EC is a straight way to hit transport belt bottleneck

I don't like to have two different factory inputs for the green circuit, so I just use 1 copper wire production for each green circuit. But you have to make sure that your copper wire production is 1.5 times faster than the green circuit. Like using factory 3 for copper wire and then factory 2. If you really do maximize your output, then just build more factories next to them.