I did not want to miss some 235 without the input filter, but practicality is not really the goal here anyway.quyxkh wrote: Mon Mar 16, 2026 2:19 pm Take the filter off the '235 loader, replace the '238 loader with an underground exit, the back '238 unloader fills, that saves the side lane.
I like this, but it isn't perfect. It will backlog on 238 and stop producing 235.Loewchen wrote: Mon Mar 16, 2026 12:55 pm Might still not be the smallest setup using belts:
grafik.png
I don't like using unconnected undergrounds because I have a habit of accidentally connecting to them when building something else nearby and causing a huge mess.quyxkh wrote: Mon Mar 16, 2026 2:19 pm Take the filter off the '235 loader, replace the '238 loader with an underground exit, the back '238 unloader fills, that saves the side lane.
I haven't seen something like this before and it makes me think, but there's a couple problems with this design. Most importantly it loses 235 on top that it doesn't recapture. You only want excess 235 to leave or it will shutdown prematurely. I think there's probably issues with 238 too, but I stopped analyzing once I saw the problem with 235.inefficient wrote: Mon Mar 16, 2026 8:38 pm I have this one:
I think thats the smallest possible, got it from the forum I think, you dont need the beacons and can tear the units apart, this is just a small module that is balanced, ie it makes enough uranium to fuel the kovarex.
03-16-2026, 20-39-10.png
My first thought in seeing your design is that I do the same thing with recipes on Gleba that consume their own outputs like nutrients and bacteria. If there's one flaw with my design, it's that you wait a long time for the 235 to loop around which is agonizing when first starting to run. I could use this to redesign mine to go faster, even if it takes up a little more space it will be a win. I also think that incoporating this into my design will allow me to fit the max # of beacons which is a huge win.quyxkh wrote: Mon Mar 16, 2026 10:45 pm Really, though, I think you can start with the one on the left here and extend to the row as desired:
pic
. This is somewhat close to optimal, it uses quyxkh's trick of using a low speed belt to force ejection. I did not know that one before. Is it cheating if I use a filtering stacker? It takes a wire and configuration, so it might count as unacceptable.Dude, finish that, you can get it into 5m high + any beacons and everything but the initial input prep fits in the beacon box.coffee-factorio wrote: Tue Mar 17, 2026 11:44 pm Note that the red belt demonstrates a shorter pathway that can be used when the machine is less motivated.
Screenshot 2026-03-17 173351.png
Ah, the belt gave out
Okay, you'll have some versions after my shift. Sometimes I leave things so that others can pick it up and improvise off it. I can just, see a line of kovarex when you don't have beacons. But when you do, you actually have to have the extra arms to get the full potential out of the machine.quyxkh wrote: Wed Mar 18, 2026 10:37 am Dude, finish that, you can get it into 5m high + any beacons and everything but the initial input prep fits in the beacon box.
That's definitely way more elegant than mine! awesome!quyxkh wrote: Tue Mar 17, 2026 8:20 pm I hadn't bothered going for a do-it-all 'cause as you say the fixes are easy, but giving it a shot led me to this, which surprised and pleased me with how well it works, so, thanks for the prodding!
pic
I decided to try it out and see if I could get it fully beaconed. It jammed on the output bulk inserter on my first test.quyxkh wrote: Tue Mar 17, 2026 8:20 pm I hadn't bothered going for a do-it-all 'cause as you say the fixes are easy, but giving it a shot led me to this, which surprised and pleased me with how well it works, so, thanks for the prodding!
pic
I fixed it, this is going in my blueprint book for sure. I couldn't figure out how to get it to come out on the opposite side, but it's no big deal to use transport belts to bring it around and chain it's output into the next one.royce3 wrote: Wed Mar 18, 2026 12:25 pmI decided to try it out and see if I could get it fully beaconed. It jammed on the output bulk inserter on my first test.quyxkh wrote: Tue Mar 17, 2026 8:20 pm I hadn't bothered going for a do-it-all 'cause as you say the fixes are easy, but giving it a shot led me to this, which surprised and pleased me with how well it works, so, thanks for the prodding!
pic
03-18-2026, 20-24-24.png
