Noob intro + baffled by Kovarex

Don't know how to use a machine? Looking for efficient setups? Stuck in a mission?
Laie
Long Handed Inserter
Long Handed Inserter
Posts: 60
Joined: Fri Dec 25, 2020 6:18 pm
Contact:

Re: Noob intro + baffled by Kovarex

Post by Laie »

Pat Healy wrote:
Thu Mar 18, 2021 5:39 pm
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
The simple method is to use a standing train as a big box. It has to include a locomotive and needs to be parked at a station, though, just so you can read the train contents. Up to six kovarex machines can work on the contents a wagon, and you still have to spots at the rear where you can place inserters to refill / skimm off. As usual with simple setups, you won't get anything out of it until all Kovarex machines have filled their buffer.

Another variant to the sushi belts above:
Image
U235 is fed back into the machine just as it comes out. With hand_size = 2, that takes exactly ten swings until it has enough to start the next batch, at which point it demands U238 -- this takes another three swings, while the excess item of U-235 happily slides past. So even without any circuit magic, I can rake off some U-235 right from the start.

However, without circuit magic I cannot ever be 100% perfectly positively sure that it will always work like that; so I'm careful to only take what I really need while the remainder goes into the machine's buffer. And anyway, that scheme will no longer work as soon as I add a second centrifuge. A this point, i install a throughput limiter for U-235. One item/tile is allowed back into the loop, anything that piles up at that chokepoint is siphoned off.

Backstop5
Manual Inserter
Manual Inserter
Posts: 2
Joined: Fri May 31, 2019 11:08 am
Contact:

Re: Noob intro + baffled by Kovarex

Post by Backstop5 »

NZ2020_002.jpg
NZ2020_002.jpg (171.65 KiB) Viewed 1130 times

(Disclaimer: I did not come up with this one, but I can't find on the forum who did. Whoever you are, thank you for this simple solution!)

This is what I use. It only uses one wire.

It works because of a quirk in the way the uranium is extracted after the process is done.
The (bad) U-238 always gets extracted first, in this set-up by the left green inserter. This inserter is wired to the yellow inserter at the bottom, and is set to read contents and wait.
The yellow inserter is set to activate once it gets the signal a U-238 piece is picked up by the green inserter. Because all the U-238 is already removed from the centrifuge once this inserter gets activated, it will pick up ONE piece of U-235. This inserter MUST be set to stack size one. (it can be any inserter, as long as it is set to stack size 1)

All the uranium (2 U-238 and 40 U-235) is put on a short belt by the left inserter, and is picked up and put back into the centrifuge by the right green inserter. The U-238 is complemented by the long arm red inserter from a belt with U-238, and the process starts again.

Few notes:
- you cannot use productivity modules in this set up. Because th bottom inserter only picks up one U-235, the extra produced U-235 from the productivity modules will just accumulate over time and clog the centrifuge eventually.
-as with all kovarex, you need 40 U-235 to start it up. Put it in manually to start it up.
-the top belt in this picture is not important for this set-up

Serenity
Filter Inserter
Filter Inserter
Posts: 961
Joined: Fri Apr 15, 2016 6:16 am
Contact:

Re: Noob intro + baffled by Kovarex

Post by Serenity »

Backstop5 wrote:
Sun Mar 21, 2021 7:34 pm
you cannot use productivity modules in this set up.
I used that for quite some time. Worked great. Then when they re-enabled productivity modules for Kovarex I was very sad when it broke :(

SoShootMe
Fast Inserter
Fast Inserter
Posts: 186
Joined: Mon Aug 03, 2020 4:16 pm
Contact:

Re: Noob intro + baffled by Kovarex

Post by SoShootMe »

Backstop5 wrote:
Sun Mar 21, 2021 7:34 pm
(Disclaimer: I did not come up with this one, but I can't find on the forum who did. Whoever you are, thank you for this simple solution!)

This is what I use. It only uses one wire.
It's the same idea as the second one I posted earlier in this topic. But I also did not come up with it, and can't find who did either.
Backstop5 wrote:
Sun Mar 21, 2021 7:34 pm
- you cannot use productivity modules in this set up
-as with all kovarex, you need 40 U-235 to start it up. Put it in manually to start it up.
Like most things, it is a matter of opinion, but I decided speed modules are overall more useful than productivity modules for enrichment, due to the long crafting time, limited bonus (two module slots) and relatively low raw resource consumption.

With a source of U-235 in reach, an inserter (to add the U-235) and an arithmetic combinator (to count how many have been added, used to signal the inserter to turn off forever) are sufficient to make it self-starting.

User avatar
eradicator
Smart Inserter
Smart Inserter
Posts: 5190
Joined: Tue Jul 12, 2016 9:03 am
Contact:

Re: Noob intro + baffled by Kovarex

Post by eradicator »

Regarding simplicity I've always thought that well ordered inserters should be sufficient, but I've never actually made a design out of the idea before. So on the occasion of this thread I tried and it works:
  • No circuits.
  • No inserter filters.
  • No stack size override.
  • Works with automatic (buffer size 80) or manual priming (buffer size 13!).
  • Works with productivity.
The filter splitters are mostly there to push the U238 down to the lower belt. I'll leave trying to squish it into a 12 beacon square to someone else.

Auto-/manual priming:

Manual: Upper belt lower lane = no input.
Auto: Upper belt lower lane = U235 input.

The upper belts upper lane must always be empty to prevent inserters from missing U235 due to lane-switching delay.


About U235 buffering:

When U235 is supplied via the upper belts lower lane the centrifuges automatially prime themselfs but each centrifuge permanently buffers 80 additional items of U235.

When the centrifuges are instead manually primed and nothing is supplied on the upper belt, then each centrifuge will buffer up to 40 items and then on the next cycle 27 of them will be released so that only 13 additional items are permanently buffered per centrifuge. (Measured with max research. The exact numbers here will vary with inserter type, stack size research and belt speed).


2021-03-22_SimplestKovarex_v1.1.png
2021-03-22_SimplestKovarex_v1.1.png (513.24 KiB) Viewed 1067 times
[Edit1:] Added explanation about priming.
Last edited by eradicator on Tue Mar 23, 2021 3:02 pm, edited 1 time in total.
Author of: Belt Planner, Hand Crank Generator, Screenshot Maker, /sudo and more.
Mod support languages: 日本語, Deutsch, English
My code in the post above is dedicated to the public domain under CC0.

SoShootMe
Fast Inserter
Fast Inserter
Posts: 186
Joined: Mon Aug 03, 2020 4:16 pm
Contact:

Re: Noob intro + baffled by Kovarex

Post by SoShootMe »

eradicator wrote:
Mon Mar 22, 2021 5:11 pm
Regarding simplicity I've always thought that well ordered inserters should be sufficient, but I've never actually made a design out of the idea before. So on the occasion of this thread I tried and it works:
Mostly... But U-238 jams the splitter unless the right lane of the far belt is moving enough. And if you have occasional U-235 on the near belt (eg from uranium processing) and are unlucky, U-235 from the output can be missed, causing a stall (like mentioned before), which I don't think can be solved while maintaining all the properties you listed :(.

User avatar
eradicator
Smart Inserter
Smart Inserter
Posts: 5190
Joined: Tue Jul 12, 2016 9:03 am
Contact:

Re: Noob intro + baffled by Kovarex

Post by eradicator »

SoShootMe wrote:
Tue Mar 23, 2021 12:44 pm
And if you have occasional U-235 on the near belt
That is invalid usage. The upper ("near") lane must be empty at all times as stated in the original post. I've edited the post to be more explicit about that.
eradicator wrote:
Mon Mar 22, 2021 5:11 pm
When U235 is supplied via the upper belts lower lane
___
SoShootMe wrote:
Tue Mar 23, 2021 12:44 pm
Mostly... But U-238 jams the splitter unless the right lane of the far belt is moving enough.
Hm...true, my test setup had a full moving belt. I have an idea how to fix it but i can't play for a few days.
Author of: Belt Planner, Hand Crank Generator, Screenshot Maker, /sudo and more.
Mod support languages: 日本語, Deutsch, English
My code in the post above is dedicated to the public domain under CC0.

SoShootMe
Fast Inserter
Fast Inserter
Posts: 186
Joined: Mon Aug 03, 2020 4:16 pm
Contact:

Re: Noob intro + baffled by Kovarex

Post by SoShootMe »

eradicator wrote:
Tue Mar 23, 2021 2:54 pm
SoShootMe wrote:
Tue Mar 23, 2021 12:44 pm
And if you have occasional U-235 on the near belt
That is invalid usage. The upper ("near") lane must be empty at all times as stated in the original post. I've edited the post to be more explicit about that.
I didn't specify but I meant the right (far) lane of the near belt. Assuming stack size 12 and express transport belt, put some U-238 and 40 U-235 in a centrifuge. A few ticks after the inserter picks up the output U-238, place a single U-235 where the output will be dropped. The inserter that should reload the U-235 will pick up that U-235, start waiting for more, let the U-238 pass under it, and time out just before the first batch of U-235 output arrives. By the time it returns to pick up more, some is missed, so it ends up with less than 40.

The underlying issue is the reloading being "disturbed" by other U-235, which is also the root of the problem you mention with inserter lane switching. Belt speed and stack size will change the timing and exactly what happens, but I don't think you can guarantee to avoid it causing the stall I described without compromising at least some of your design properties. Always ready to be proved wrong though!

User avatar
eradicator
Smart Inserter
Smart Inserter
Posts: 5190
Joined: Tue Jul 12, 2016 9:03 am
Contact:

Re: Noob intro + baffled by Kovarex

Post by eradicator »

Like I said can't test right now :|... But I don't think that even if what you describe would happen that it would be a problem. The presence of any U235 at all on the upper lane means that the user wants automatic mode and thus there should be more U235 incoming which will eventually fix the problem automatically. Further more in automatic mode each fully buffered centrifuge contains 121 items of U235 (80 in input and 41 being currently processed), so it would have to happen a dozen times in a row for even a single centrifuge to run completely dry. So the chance that even a single centrifuge in the row temporarily stops is imho minuscule. And as it's stackable the user will have more than one centrifuge. After all the design does not aim for perfect ratio or 100% machine utilization, it's meant to be "fire and forgot".
Author of: Belt Planner, Hand Crank Generator, Screenshot Maker, /sudo and more.
Mod support languages: 日本語, Deutsch, English
My code in the post above is dedicated to the public domain under CC0.

SoShootMe
Fast Inserter
Fast Inserter
Posts: 186
Joined: Mon Aug 03, 2020 4:16 pm
Contact:

Re: Noob intro + baffled by Kovarex

Post by SoShootMe »

It will recover without intervention, but consider: you've got most of your first 40 U-235 and set up a few enrichment centrifuges to build up some U-235 (or to use it to make uranium fuel cells or whatever), and leave it to it while you do something else. When you finally get round to checking on it, an hour or three later, instead of a nice stash of U-235, you find the first centrifuge stalled 10 minutes after you built it and you've produced little or no U-235 since.

Fed with "occasional U-235" I found the buffer for the first centrifuge basically followed what you described for "manually primed". So it takes just one badly timed U-235 when it's near the lower end of the 13, 14, ..., 40, 13, ... pattern to cause the stall. Definitely very unlikely (or unlucky) but that just means it's guaranteed to happen when it's most annoying ;).

steve7100
Burner Inserter
Burner Inserter
Posts: 17
Joined: Tue Mar 30, 2021 7:39 pm
Contact:

Re: Noob intro + baffled by Kovarex

Post by steve7100 »

Here is my nuclear setup, I'm sure it's not optimized but it works for me.
After watching Yama Kara's video on youtube it kinda inspired me to come up with
my Kovarex Go-Round setup. the outer loop had 238 and the inner loop is for 235 & 238 put out
by the Kovarex centrifuge. I put the splitter on the inner belt to force the 238 to the inside leaving the
outside clear for the 235 coming out. The filter inserter picks up the excess 235 after the second blue
inserter puts the 235 back into the centrifuge.
Attachments
Factorio -Nuclear.jpg
Factorio -Nuclear.jpg (2.56 MiB) Viewed 879 times

User avatar
ptx0
Smart Inserter
Smart Inserter
Posts: 1080
Joined: Wed Jan 01, 2020 7:16 pm
Contact:

Re: Noob intro + baffled by Kovarex

Post by ptx0 »

wonder if you all scared the OP off :lol:
My Mods - Fish Per Minute base size metric - Use your crashed spaceship as a belt balancer?
• • •
Base: Bob's @ 1 Million SPM
• • •
Linear search and overflows are indicative of sloppy coding practices.

User avatar
eradicator
Smart Inserter
Smart Inserter
Posts: 5190
Joined: Tue Jul 12, 2016 9:03 am
Contact:

Re: Noob intro + baffled by Kovarex

Post by eradicator »

SoShootMe wrote:
Tue Mar 23, 2021 5:00 pm
The underlying issue is the reloading being "disturbed" by other U-235,
Gotta admit I don't have the motivation to invent a circuit contraption that generates a single-tick-precision disturbance like that. If someone can supply me with a test setup I'll happily use that.

In the meantime I did several hours (+ several hours to get it started in the first place) trial run on a tiny (~200k) uranium field, using both the Version 1 design posted here, and my this-should-fix-it Version 2 unpublished design.
screenshot_078h_24m_11s.png
screenshot_078h_24m_11s.png (5.87 MiB) Viewed 688 times

They both ran without disturbance until the U238 belt got stuck due to full chests. Then the V1 stopped and a few minutes later V2 stopped too. The problem was the 2*U238 being output by the recipe blocking the splitter and thus the outflow of U235. The only fix I could think of was to limit the U238 belt to one lane, so that the longserter prioritizes the locally produced 2*U238.

Also important: Manually feeding U238 to the centrifuges in V2 will cause blockage because it stops the longserter from un-stucking the splitter. (Manual priming with U235 is fine.)

[Edit: Well, that fix didn't last long. Perhaps obvious to someone more awake than me the V2 now loses U235 when it waits for more U238 at the end of a cycle. Setting both stackserters to stack override 2 (:/) should fix that i think...]

screenshot_081h_46m_42s_copy.png
screenshot_081h_46m_42s_copy.png (1 MiB) Viewed 688 times
Last edited by eradicator on Tue May 11, 2021 9:30 pm, edited 1 time in total.
Author of: Belt Planner, Hand Crank Generator, Screenshot Maker, /sudo and more.
Mod support languages: 日本語, Deutsch, English
My code in the post above is dedicated to the public domain under CC0.

User avatar
disentius
Filter Inserter
Filter Inserter
Posts: 571
Joined: Fri May 12, 2017 3:17 pm
Contact:

Re: Noob intro + baffled by Kovarex

Post by disentius »

Well... here is a very simple version. Input/output left, expandable right, room for beacons included.
Autostart. Doesn't care if you stop it with backpressure or by cutting ore. basically some splitter shenanigans.
2021-05-11 22_34_12-Window.png
2021-05-11 22_34_12-Window.png (2.16 MiB) Viewed 680 times
[EDIT]
Misaligned one splitter... here is the beaconed version, with a fuel cell recycler added.
2021-05-11 22_55_21-Window.png
2021-05-11 22_55_21-Window.png (1.34 MiB) Viewed 678 times

User avatar
eradicator
Smart Inserter
Smart Inserter
Posts: 5190
Joined: Tue Jul 12, 2016 9:03 am
Contact:

Re: Noob intro + baffled by Kovarex

Post by eradicator »

disentius wrote:
Tue May 11, 2021 8:49 pm
Well... here is a very simple version.
Kudos. Your designs are superb as always. Except for the larger buffer size it does everything better than mine :D.
Author of: Belt Planner, Hand Crank Generator, Screenshot Maker, /sudo and more.
Mod support languages: 日本語, Deutsch, English
My code in the post above is dedicated to the public domain under CC0.

Premu
Long Handed Inserter
Long Handed Inserter
Posts: 79
Joined: Wed Nov 13, 2019 4:40 pm
Contact:

Re: Noob intro + baffled by Kovarex

Post by Premu »

Kovarex is one of the things where you can build simple things, but also highly optimized.

My latest design used some new tricks I learned and is beaconized, tileable, and starts itself while only taking exactly 80 U-235 for its processing per centrifuge.

Image

The upper belt provides the U-235, the lower one the U-238.

To get the U-235 to start the process, inserter B pulls it from the upper belt. A memory cell counts the amount of U-235 received by substracting it from the signal U. This signal U is remapped to the value V, and the inserter B takes this value V together with value 80 from constant combinator and only puts in more U-235 if V is positive. V is also used to set the stack limit of that inserter, so if 79 U-235 have been loaded in, and two arrive close at the belt, only one will be loaded. After 80 U-235 have been loaded in, this section won't do anything at all and could be removed completely if you're frugal.

The blue inserter A is only active if its input "U-235" is larger than 0. This same value is used to set its stack size. It receives this value by reading both the content from the crate and a constant signal for U-235 which is set -40. So once a production cycle is finished and all the U-235 is dumped into that crate, either one or two U-235, depending if the productivity bonus gave something on top, will be put onto the belt. After that the inserter is inactive, and exactly 40 U-235 remain in the chest.

The inserters C and D are controlled by RS latches in the orange box. D will get activated once the box is empty and deactivated once it contains more than 40 U-235. C gets active once inserter A grabs something and deactivated once the crate is empty.

The U-238 is a lot simpler - the inserters feeding the lower crate are just limited to put no more than 20 U-238 into that crate, while no limitation is there for the output of the Kovarex ernrichment. Thus the latter is priorized and won't block anything.

Overall, these centrifuges will fill up on their own, and they will automatically shutdown if the output belt backs up, keeping exactly 80 U-235 in its system.

I have to admit it's kind of overengineered for later in the game, but it's always fun to make "optimal" and nice looking builds.

Post Reply

Return to “Gameplay Help”