Noob intro + baffled by Kovarex

[Pat Healy]

Hello all, first post here...started playing the game a few weeks ago and I've had the common experience of not having seen the sun or or other humans ever since. I'm a tinkerer by nature and insatiably curious, so this game is like crack. Super rewarding to walk around the factory and see things humming away that were completely mysterious a couple weeks ago; I'm really enjoying the process.

I'm slowly working through my first build, working on getting yellow science up and running now. I prefer to figure out things on my own (generally I think overreliance on tutorials ruins games) but I've hit a wall with Kovarex enrichment. I have the process up and running (I'm awash in U235 and fuel cells) but I can't figure out the magic of skimming off that 1 unit of U235 automatically, so periodically I run over to my centrifuges and manually pull it out. I finally broke down and watched several Kovarex tutorials, and honestly I'm baffled. The tutorials usually start out with 20 minutes of basic info followed by a mad clickstorm that skims over all the important stuff in about 45 seconds, and I get lost. I invariably end up with a rat's next of inserters, red and green wire running around, and incomprehensible combinator conditions, the only consistent outcome being that the setup sits there immobile, indifferent to my curses and vile threats, and the U235 just keeps stacking up.

So...can someone please explain to me like I'm a fifth third grader...how do I automate the extraction of the 1 U235?

[GrumpyJoe]

I think you only want a hint and figure out the rest yourself?

Let's try it that way:
Pull out all Uxxx into a chest
Use stack inserters with a stack size set to 10 (something to learn)
Pull out 4x10 U23x
If 1 is left, pull it out with another inserter.

Use the logic, Luke!

[DaveMcW]

Priority splitters.

Put a splitter in the belt between the centrifuge output and centrifuge input. Set the splitter output priority in the direction of the centrifuge input. Any surplus uranium will go in the other direction from the splitter.

[boran_blok]

Here is my very simple and not very efficient Kovarex setup:

SimpleKovarex.png (231.66 KiB) Viewed 5837 times

The key is that the 235 is only ever put on the inside lane and never the outside. The outside lane is for 238 only. The output 238 is priority used as input, and replenished by output from the source centrifuges.

This is not very efficient as it requires a VERY long startup time before any output is generated. But it is relatively easy to understand I hope.

[torne]

The complicated setups to take out 1 each time and only let exactly 40 cycle around aren't strictly necessary; you can just make a looping system with no logic like boran_blok shows and as long as it prioritises keeping kovarex going over outputting 235 for use it will eventually fill the loop and output 235 at exactly the same speed as a clever logic driven setup. Kovarex produces 235 so quickly relative to the rate it's used that the amount that's "tied up" in the loop soon becomes irrelevant.

If you're not approaching the endgame then your reactor is probably small enough (e.g. 4 reactor cores? 6?) that it can run for a very long time just on the 235 being output by centrifuging ore from a single ore patch; you only need approximately 1 centrifuge (and ~9 miners) per reactor for that, and that's assuming your reactor runs continuously; it'll be much less if you only insert fuel when needed. Starting up kovarex is still a good idea unless you are going to stop playing once you launch the rocket because it helps you avoid needing to store ever larger amounts of excess 238 (though using a ton of uranium ammo is another option).. but making sure it "starts up" with the shortest possible delay really doesn't matter 95% of the time.

So if you want to try to build a clever optimal setup for its own sake then see the hints other people suggested, it's an interesting challenge with no entirely trivial solution.. but don't feel like it's vital either, and if it's frustrating you then just make a simple loop and let the 235 pile up while you go do something else

[eradicator]

Like the people said above, you don't need complicated logic as long as you don't care that a bunch of resources are "wasted" on belts. A bunch of belts and splitters is more than enough. Also for power you need really really tiny amounts of U235.

This blueprint is what i used to power my 2GW facility (16 reactor cores). It uses above mentiond 4*10+1*1 stack size trick, but without circuits.

screenshot_000h_07m_02s.jpg (414.93 KiB) Viewed 5778 times

[pichutarius]

Using bot will trivialize the puzzle, if thats ok with you, then:

bot trivialize puzzle

uranium cycle: blue chest → inserter → centrifuge → inserter → red chest → bot → blue chest

Input inserters only work when U235 is below a certain threshold in the logistic network. (1000 in the picture)

Finally, within the logistic network, put red chest full of U238, put buffer chest (not requester) for pulling out U235, done. Buffer chest because it has lower priority than requester chest, so it wont pull out the catalyst U235.

The setup is beacon-able.

[Pat Healy]

Great input, thanks so much to all! I’ve been stumped by this for a few days, admittedly way too focused on something like perfection. Some great quick solutions here to get me over the hump as well as more advanced ideas to come back to. Look forward to playing with these tomorrow.

[Trific]

My centrifuges output to a belt that one space down the line they can input from. Anything that makes it past this (surplus 235, plus the 2 238 that are output from the process), then get side joined to the input belt so they are on the inside lane (inserters preferentially pick from the inside). This gets the output another trip around to ensure the output 238 is preferentially taken over the fresh stuff coming in, plus another shot at the 235. Any 235 left on the input belt at the point just before the output belt joins is separated off (either by splitter or filter inserter) as surplus. It has been absolutely bullet proof so far.

[SoShootMe]

I invariably end up with a rat's next of inserters, red and green wire running around, and incomprehensible combinator conditions, the only consistent outcome being that the setup sits there immobile, indifferent to my curses and vile threats, and the U235 just keeps stacking up.

I have found the Kovarex enrichment process one of the more interesting puzzles in the game, with many possible solutions and design trade-offs, such as:

• General level of complexity.
• Duty cycle.
• Self-starting vs requiring "priming" with U-235.
• Ability to function with productivity modules.
• Ease of scaling up to more centrifuges.
• Ability to use, or effectiveness of, beacons.

I have spent (countless) hours experimenting, getting a complex design using combinators, beacons and productivity modules with 100% duty cycle working along the way. But somehow that complexity was unsatisfying and I currently consider two far simpler designs that don't use any combinators as my favourites: one is self-starting and uses two splitters (for use after first researching the Kovarex enrichment process), and the other is more "beacon-friendly" but requires priming (for later high-output use).

Details

kovarex.PNG (78.8 KiB) Viewed 5652 times

The second is a slight variation of a design I came across (maybe in these forums). It relies on the fact that U-238 will always be removed first when the cycle completes, and sequencing controlled by the circuit network to extract the "extra" U-235. The settings on the linked inserters are:
Stack inserter: enable if U-235=0, read hand contents (pulse)
Fast inserter: enable if U-238>0, read hand contents (hold), stack size 1

[disentius]

I think it was mine.
viewtopic.php?p=391015#p391015

Still using (an updated version) it.
See comment Optera below

EDIT: Clarified the text after Optera's comment

[SoShootMe]

I think it was mine.
viewtopic.php?p=391015#p391015

Still using it.

I can honestly say I came up with my two splitter, circuit network-free design independently of yours (or anyone else's). Great minds think alike (...fools seldom differ). Interestingly, using a pair of filter inserters for the input (as in your version) causes a change in the behaviour: the centrifuge fills with U-235. Presumably because a single, non-filter inserter (as in my version) prefers to pick up U-238 when input space is created by the centrifuge starting.

It was the other that was a tweaked version of a design I came across. I added reading the fast inserter and the condition on the stack inserter, so the centrifuge just stops (with 40 U-235 in its output) if the output belt is backed up.

[Optera]

I think it was mine.
viewtopic.php?p=391015#p391015

Still using it.

That's very outdated, you really use the 2018 setup when you made much more reliable ones since?
one circuit free version: viewtopic.php?p=509952#p509952
and the circuit controlled version I adopted: viewtopic.php?p=529993#p529993

[disentius]

I was compelled by the remark of SoShootMe to dive into kovarex posts on this forums. This one is my oldest ddesign with this mechanic.
I do use the newer versions. Found a lot of other designs too, the kovarex possibilities are endless

[SoShootMe]

viewtopic.php?p=391015#p391015

Still using it.

That's very outdated, you really use the 2018 setup when you made much more reliable ones since?

"Much more reliable" implies there's something unreliable about that design, but I didn't see any indication of that in the topic and I've never managed to break the almost identical one I use. Did I miss something?

[Optera]

Read through the posts between disentius and me about the circuit controlled version.

[SoShootMe]

Ah, this?

When testing with one U325 inserter on the "main"belt, with more modules chained, it can miss picking up all the needed output U235 due to timing issues with U235 arriving on the belts from the previous reactor(s).

Somehow I've not noticed this, but I can see the problem. Loading U-235 from the output (before merging into the input belt) is a simple fix, but without combinators I don't think you can avoid filling the centrifuge input buffer with U-235 before the extra becomes available.

[Optera]

I didn't have issues with inserters miss crabbing the cycling u235 either, but it might be an issue especially in noob friendly builds.
Without circuit control there's no way to limit input inventory to just 40 or 80 u235, but I wouldn't call it noob friendly anymore.

[eradicator]

Without circuit control there's no way to limit input inventory to just 40 or 80 u235, but I wouldn't call it noob friendly anymore.

My primed 4*10+1*1 stack limited build posted above does exactly that I believe. It should only ever fail to do so if the output is blocked.

[SoShootMe]

The (implicit) context was auto-starting. The 4*10+1*1 approach will maintain 40/80 U-235, but I think it is the simplest solution that is guaranteed to work without filling the centrifuge.

...Provided you solve the "output blocked" failure. The simplest way to do this is to prime with exactly 40 U-235 and use the circuit network to enable the inserters that reload the U-235 only when the output inserter is not holding U-235.