Page 1 of 1

Direct inserters, not really good.

Posted: Mon Apr 29, 2019 10:08 am
by drNokard
Hi!
I post here instead of general discussion, because this post involves a link to a design I posted on reddit, but I'd like to hear from the "official community". :)

Maybe many of you already know this "effect" tht I call "the inserter effect", but while designing this thing (https://www.reddit.com/r/factorio/comme ... ompressed/) I noticed that trying to insert the output directly on the output belt have great impact on performance.
This is why in that design I opted for using side belts loader that increase performance. In this design I use 35 beacons and 16 furnaces, while in the common designs I see around at least 50 beacons.
The inserter has great loss of efficiency because over time it will miss a couple of spots, and that will make the belt uncompressed. You will have then to add more production entities to cover those spots.
The side loading belts, instead, will act as buffer and fill completely the belt.
These are my conclusions after some tests in sandbox mode.
For those not used on blueprints on Reddit, I add a link to the resulting image (https://cdn.discordapp.com/attachments/ ... eprint.png) and a link to the blueprint in pastebin (https://pastebin.com/YSjPefBF)
What do you think?

Re: Direct inserters, not really good.

Posted: Mon Apr 29, 2019 11:01 am
by Bauer
For direct machine-belt insertion:
Use 2 inserters to load to the belt for the last furnace.

Re: Direct inserters, not really good.

Posted: Mon Apr 29, 2019 11:07 am
by mrvn
I think that splitting a blue belt into 2 belts should not require blue belts anymore.

Otherwise I'm not sure what you expect. Yes, inserters take time to swing back and forth and any gap passing by in that time is lost. Using side loading belts as buffers can cover that time and keep your belt compressed. It also means the inserter can drop it's hand faster onto the buffer belt without having to wait for gaps in the main belt. So twice the bonus.

Re: Direct inserters, not really good.

Posted: Mon Apr 29, 2019 11:54 am
by drNokard
mrvn wrote:
Mon Apr 29, 2019 11:07 am
I think that splitting a blue belt into 2 belts should not require blue belts anymore.

Otherwise I'm not sure what you expect. Yes, inserters take time to swing back and forth and any gap passing by in that time is lost. Using side loading belts as buffers can cover that time and keep your belt compressed. It also means the inserter can drop it's hand faster onto the buffer belt without having to wait for gaps in the main belt. So twice the bonus.
You are right, a couple of red belts should be enough, but I like it all blue :)
Nothing, I discovered this thing and I thought that it would be useful to share it for someone that didn't realize it yet.
That, and to share my design. :)