This one goes hand in hand with the steel smelting.
- PLATE SMELTING W BEACONS - SCHEMATIC rev.jpg (244.07 KiB) Viewed 13094 times
Final (Maxed) Designs
Forum rules
Circuit-free solutions of basic factory-design to achieve optimal item-throughput
Re: Final (Maxed) Designs
Plate Smelting in Electric Furnaces with Beacons
This one goes hand in hand with the steel smelting.
revised to eliminate 4 splitters per comments by hansinator
This one goes hand in hand with the steel smelting.
revised to eliminate 4 splitters per comments by hansinator
Last edited by DerivePi on Fri Mar 03, 2017 9:47 pm, edited 2 times in total.
Re: Final (Maxed) Designs
Nice one! I really like your series at beaconed designs 
-
hansinator
- Fast Inserter
- Posts: 160
- Joined: Sat Sep 10, 2016 10:42 pm
- Contact:
Re: Final (Maxed) Designs
I like it, too. Good attention to details. Using a two-row design is very clever, it didn't cross my mind before.. I always had one long row. Using two rows eliminates usage of blue belts within the furnace block. It also uses less beacons. Each row adds an additional 15 beacons, whereas a 1-row design like the one from vanatteveldt adds 14 per row. But the 2-row design starts with only 22 beacons instead of 28 for a 1-row design. If you stack 7 of those blocks and share the beacons, they both use an equal count of beacons. Beyond 7 stacked blocks the 1-row design uses less beacons, below the 2-row design from DerivePi is better. I especially like how it gets the furnace speeds right to minimize overproduction. If you have an even amount of furnaces it requires an odd number of beacons which introduces an asymmetry. In a stackable 1-row design that's a problem, but here DerivePi has presented a very nice solution. It has an asymmetric layout but the sides that connect to the next block are still consistent. Nice! I personally dislike the balancing splitters but it does it's job I guess. I would have tried to avoid splitters and put the output of the first 3 furnaces from the left in each row onto the other belt side to make space for the output of the last 4 furnaces. The last 4 furnaces should produce 691.2 plates per minute, which is manageable for one side of a red belt.
However, there is one thing that is inherent to all those beacon setups. They either over- or under-produce. When you add productivity modules you can't precisely output 2400 plates/min because all the math results in fractional numbers of furnaces you'd need. This setup produces 2460 plates/min, so inevitably one or two furnaces will buffer up until their output buffer is full.
However, there is one thing that is inherent to all those beacon setups. They either over- or under-produce. When you add productivity modules you can't precisely output 2400 plates/min because all the math results in fractional numbers of furnaces you'd need. This setup produces 2460 plates/min, so inevitably one or two furnaces will buffer up until their output buffer is full.
-
d4rkpl4y3r
- Long Handed Inserter
- Posts: 66
- Joined: Wed Apr 27, 2016 8:24 pm
- Contact:
Re: Final (Maxed) Designs
I posted a green circuit design on reddit a while back and just now found this thread. With its 40x17 footprint it is quite a bit smaller than the ones here.
Blueprint string
Re: Final (Maxed) Designs
1) Is a red long inserter fast enough to keep copper wire production running full speed?
2) Why the buffer chests on the copper wire output? I now inserters can hand stuff directly to the next inserter. Does that not work with inserter bonuses later on?
2) Why the buffer chests on the copper wire output? I now inserters can hand stuff directly to the next inserter. Does that not work with inserter bonuses later on?
-
d4rkpl4y3r
- Long Handed Inserter
- Posts: 66
- Joined: Wed Apr 27, 2016 8:24 pm
- Contact:
Re: Final (Maxed) Designs
1) Yes, requires both stack boni for non stack inserters and is highly unstable. Even slight disturbances can drop production down to 2.3k/min.
Production stats
2) Inserters can drop items on the ground and pick up items from the ground, they don't give it directly to the next inserter as far as I know. With the stack bonus that should make the buffer chests better, but I haven't tested it.
Re: Final (Maxed) Designs
Circuits with Beacons
Since there is so much demand for basic circuits and you get a production multiplier on copper plate twice (once going to copper cables and then again going to circuits) this setup represents one of the more advantageous uses of productivity modules and speed beacons (cuts required copper plate by almost half). I also found it to be one of the more difficult layouts to puzzle together. I haven't play tested it yet but have a look.
Schematic
Since there is so much demand for basic circuits and you get a production multiplier on copper plate twice (once going to copper cables and then again going to circuits) this setup represents one of the more advantageous uses of productivity modules and speed beacons (cuts required copper plate by almost half). I also found it to be one of the more difficult layouts to puzzle together. I haven't play tested it yet but have a look.
Schematic
- CIRCUITS W BEACONS SCHEMATIC.jpg (320.64 KiB) Viewed 12961 times
-
d4rkpl4y3r
- Long Handed Inserter
- Posts: 66
- Joined: Wed Apr 27, 2016 8:24 pm
- Contact:
Re: Final (Maxed) Designs
133k/h green circuits. If you replace one of the fast inserter per copper cable assembler with a stack inserter it goes up to 143k/h green circuits. That is less than 1% off from the full 40/s limit of the output belt, very impressive.DerivePi wrote:I haven't play tested it yet but have a look.
Blueprint of the 143kph version
-
Ranakastrasz
- Smart Inserter
- Posts: 2173
- Joined: Thu Jun 12, 2014 3:05 am
- Contact:
Re: Final (Maxed) Designs
How did you make the schematics? Is it a program, texture pack, or paint?
Or something else?
Or something else?
My Mods:
Modular Armor Revamp - V16
Large Chests - V16
Agent Orange - V16
Flare - V16
Easy Refineries - V16
Modular Armor Revamp - V16
Large Chests - V16
Agent Orange - V16
Flare - V16
Easy Refineries - V16
-
vanatteveldt
- Filter Inserter
- Posts: 947
- Joined: Wed Nov 25, 2015 11:44 am
- Contact:
Re: Final (Maxed) Designs
It is well known that 3 pairs of plants (with full prod3+sp3) can saturate a blue belt. The challenge is ensuring enough throughout in a limited space, since you have 3 belts, need multiple inserters per connection, and only limited space between the beacons. This solution uses belt braiding, which allows for a very elegant design, but is frowned upon by some as an 'exploit'.d4rkpl4y3r wrote:133k/h green circuits. If you replace one of the fast inserter per copper cable assembler with a stack inserter it goes up to 143k/h green circuits. That is less than 1% off from the full 40/s limit of the output belt, very impressive.DerivePi wrote:I haven't play tested it yet but have a look.
Also, this solution is non-optimal in the sense that it uses more beacons than strictly needed as there are gaps between the assemblers. If you have at most one 1-tile gap, all beacons reach a maximal number of assemblers. However, without gaps and without belt braiding it becomes a lot trickier to find an elegant solution as suddenly the wire->circuit connection also has to happen in the two tiles between the assemblers and the beacons.
More importantly, @OP: can you please post blueprints of your designs, and preferably screenshots as well. Your CAD plots look very swanky, but they are much harder to read than a screenshot and to duplicate/check/improve designs it is annoying if there is no blueprint.Ranakastrasz wrote:How did you make the schematics? Is it a program, texture pack, or paint?
Or something else?
Re: Final (Maxed) Designs
When approaching EC production from input sides I ended with this monstrosity.
I didn't bother balancing the output from 6 to 5 belts since I liked the symmetry and In my base it feeds into a warehouse as buffer anyway.
I didn't bother balancing the output from 6 to 5 belts since I liked the symmetry and In my base it feeds into a warehouse as buffer anyway.
- EC 11k_min.jpg (144.1 KiB) Viewed 12932 times
My Mods: mods.factorio.com
Re: Final (Maxed) Designs
Hey Optera, is that a mod that gives you those stats on the left side of your screenshot? That looks very useful if so.
Re: Final (Maxed) Designs
Sadly it's nothing fancy automated. Only smart display connected to a constant combinator with manual set production values.TI-89 wrote:Hey Optera, is that a mod that gives you those stats on the left side of your screenshot? That looks very useful if so.
My Mods: mods.factorio.com
Re: Final (Maxed) Designs
Ah, oh well. Still looks useful, thanks.Optera wrote:Sadly it's nothing fancy automated. Only smart display connected to a constant combinator with manual set production values.TI-89 wrote:Hey Optera, is that a mod that gives you those stats on the left side of your screenshot? That looks very useful if so.
Re: Final (Maxed) Designs
Ok, so I'm back with my efficiency module smelting layout 
. But this time I've got some interesting points of comparison with the beacon smelting layouts. Plus I refined mine a bit.
First off, I based these designs around the belt throughput rates so that each furnace column receives the exact amount it needs. This means that it will not back up unless the output is blocked, even when using productivity modules (though furnace utilization is slightly less than 100%). That being said, I have found that optimizing belt usage matters very little other than for aesthetics. But then my goal was to make the trade offs in module usage more apparent, rather than to minimize one cost or another. I hope this will serve that purpose.
With the iron/copper layout, mine uses about 1/4 the power of DerivePi's layout (2584.8 KW vs ~10.5 MW), with only a slightly larger footprint (37x38 vs 23x27). Obviously if you really need to minimize the space of your smelting beacons are still the way to go, but I think most of the time that is not the case. Obviously the productivity modules change things substantially, so I don't claim it's a direct comparison. But with iron/copper being 1:1 from ore to plates, I don't find productivity to be as useful.
. My first layout used 25 columns of 15 electric furnaces, with a beacon on the last four in each stack. This requires 187 MW to produce a blue belt of steel. By comparison, DerivePi's beacon layout uses 130.4 MW. I also built one which uses no beacons at all, which uses ~198.5 MW. I was planning to work from here to figure out beacons, but I think the one posted is a lot better than I would have gotten. I'll keep my belt splitting versions for the efficiency modules, but if using prod3 it seems you need to prioritize beacon placement from the start.
. But this time I've got some interesting points of comparison with the beacon smelting layouts. Plus I refined mine a bit.First off, I based these designs around the belt throughput rates so that each furnace column receives the exact amount it needs. This means that it will not back up unless the output is blocked, even when using productivity modules (though furnace utilization is slightly less than 100%). That being said, I have found that optimizing belt usage matters very little other than for aesthetics. But then my goal was to make the trade offs in module usage more apparent, rather than to minimize one cost or another. I hope this will serve that purpose.
With the iron/copper layout, mine uses about 1/4 the power of DerivePi's layout (2584.8 KW vs ~10.5 MW), with only a slightly larger footprint (37x38 vs 23x27). Obviously if you really need to minimize the space of your smelting beacons are still the way to go, but I think most of the time that is not the case. Obviously the productivity modules change things substantially, so I don't claim it's a direct comparison. But with iron/copper being 1:1 from ore to plates, I don't find productivity to be as useful.
Iron Plate Smelter
For steel production, it is an entirely different case. As I said I was doing my belt ratios first and so I didn't prioritize the beacon placement. It makes quite a difference. Apart from being absolutely massive, my steel smelter used substantially more power. I don't claim this is new info, it was said earlier in this thread that beacons need to be used to offset productivity modules. But I have a few numbers and pictures now to back it up 
. My first layout used 25 columns of 15 electric furnaces, with a beacon on the last four in each stack. This requires 187 MW to produce a blue belt of steel. By comparison, DerivePi's beacon layout uses 130.4 MW. I also built one which uses no beacons at all, which uses ~198.5 MW. I was planning to work from here to figure out beacons, but I think the one posted is a lot better than I would have gotten. I'll keep my belt splitting versions for the efficiency modules, but if using prod3 it seems you need to prioritize beacon placement from the start.Steel Smelters
-
iceman_1212
- Filter Inserter
- Posts: 256
- Joined: Wed Aug 17, 2016 9:49 am
- Contact:
Re: Final (Maxed) Designs
The new extended underground belt lengths allow for new constructs in 0.15. Here is one that I liked for beaconed red-circuit production, which includes "on-site" production of plastic and green circuits. *Insert belt-braiding warning here*
Five of these give a blue belt. Am including only the minimal modular piece here. It was nice to see the ratios work out rather closely for multiple ingredients across multiple steps of a production chain in a beaconed build.
Five of these give a blue belt. Am including only the minimal modular piece here. It was nice to see the ratios work out rather closely for multiple ingredients across multiple steps of a production chain in a beaconed build.
Production
Power
Blueprint String
Re: Final (Maxed) Designs
Love this thread. Hate the Cad drawings...just do it in game.
Re: Final (Maxed) Designs
I confess to being a little bemused. Around 12 months ago, when I was fresh-faced and still quite new to Factorio, I did a LOT of research on the forums and Reddit to gain some thoughts and insights into designs for various parts of my bus and rail setup. This solution for green circuits was posted (I think I actually tweaked another design, similar to one of the ones already mentioned in this thread), but it doesn't seem to have caught on. It should be made with Assembler 3s (not 2s) for max throughput, IIRC, but it certainly seems to me to be a pretty decent "maxed" design - input 2 express belts of copper, split an express or feed two fast belts of iron, and presto! Out pops 1 express belt of Electronic Circuits. Did I miss something, however? Is there an issue with this design?
Blueprint
Re: Final (Maxed) Designs
Wasteful to use blue belts all the way I think. You can go to red and yellow belts as copper gets used up and go from yellow to red to blue as green cards are added.
Re: Final (Maxed) Designs
A very good point - now that you have mentioned it, I vaguely recollect that the original design may well have done that & changed to red belt where the balancer is.