The Eggtimer Pacu Flipper: 3 eggs for 3-6kg algae every 5 cycles

[StoneToad]

This build produces (amortized) 3 eggs every 5 cycles for a cost of 3-6kg algae or so per 5 cycles, or 1-2kg of algae per egg.  There's a tiny (0.0064%?) chance of getting too many non-base pacu eggs in a row which I think will require manual restart (but there'll still be plenty of pacu in the basement infinite pacu glum farm so it's not like you'll lose your pacu population.

As a comparison, a naive once daily feeder breeding pacu setup will load the feeder with 1-2kg per day and replace the pacu when it's dead with a (possibly partially incubated) egg.  This would give you 2-3 eggs every 14-19 cycles at cost of 14-28kg of algae depending on the exact details of the design assuming only simple automation or 3-6kg with cycle/egg counting to stop feeding after the 2nd egg.  The biggest loss is the time that the pacu spends starving to death at the end of it's life, since it's only getting 7% reproduction while starving, the same as if it was in a normal non-breeder 1 egg per infinite pacu glum farm, but it is still eating your algae (unless you counted egg output) and taking up your breeder pond.

Since we can't move pacu directly we need a way to 'convince' the pacu to move out of the breeder pond after it's laid 2 eggs.  So we (ab)use the fact that they will count water tiles through mesh tiles, but can't move through them, to drop the water level causing them to flop along and down a disposal shaft into the infinite pacu glum farm where they will lay their 3rd and final egg after ~10 cycles.

The name of the build comes from the fact that I'm using the 5 cycle incubation period of the eggs as the timer to hold the fry until it's an adult before dropping it into the breeding pond, and then flipping them out of the breeding pond after they've laid 2 eggs

Design/build notes:

• you must have the egg's conveyor travel 2 tiles inside the room before the dropper or the sensor won't close the dropper in time.
• the hatching's left room critter sensor is above 0, the main room's is also above 0, but the fry room's is below 1 (negated) for space reasons
• once an egg has left the breeding room, it must never enter the breeding room on a conveyor or the egg counter will be messed up.
• the filter under the egg counter is set to 200, just to make sure we get our last egg before the pacu starves by upping it's reproduction counter 'for free'
• the filter to the egg holding dropper is set to 10 seconds, to make sure the door is shut.
• the buffer to the lower airlock of the reset room is set to 10 seconds, to give the pacu time to realise it in mid air and start falling (Wile E. Pacu?)
• the timer for the algae sweeper is 5s/595s to try and make it only deliver 1kg to the feeder, a 1% cycle timer will also work but will deliver 2kg, 0.5% will need to have the right side sweeper moved up one tile from the picture because it will fail to deliver sometimes.  Might want to move the sweeper up just in case, 3s/597s only loads 1kg reliably, but requires the right sweeper to finish after the algae sweeper drops the ball (of algae, make sure it's up 1 tile from the pictures so it can reach the tile under the algae sweeper).
• the switch in the water tank isn't required, just lets you manually force the water to drain, mostly was there for debugging
• once this is build, you need to set the outer door to deny access in, so that dups don't try to load the feeder manually.
• yes I know the pneumatic door next to the feeder has an automation wire connected to it, it seemed like the best compromise for the layout, and shouldn't cause any problems, during construction you can manually set it to open and it will stay that way until the automation signal changes again.
• you must only feed one egg at a time into the upper area, you can use a conveyor hopper set to sweep only to make the auto-sweeper only pick up one egg even if there's more then one in range.
• the airflow tile next to the mesh tile is required to avoid 2 types of gas blocking the water from coming back up again
• you can build the glumfarm at the bottom by itself and stick all your wild pacu eggs in there, the lower dispenser location is set so there won't be an infinite egg loop.  The upper location near the crusher can only be used with sweep only.  The doors are access restricted for travel in.

 

I'm bad at describing stuff nicely, so if anyone wants to to make a nice video or annotated images whatever, go nuts.  These screenshots were originally taken so I could rebuild this myself in my survival world but figured I've enjoyed reading/watching everyone designs so I should try and share one of mine.

I feel like there should be some way to tidy up the the automation logic using ribbon cable since I've got 3 copies of AND (A) (NOT B) logic, but I want to get back to my survival word.

 

Conveyor view:

Spoiler

Automation view:

Spoiler

edit: hopefully the aspect ratio on the images should be ok now.

[Oozinator]

On 2.4.2020 at 3:17 AM, StoneToad said:

edit: hopefully the aspect ratio on the images should be ok now.

Nope. Sry bruh no cookies for you this time

[Nxf7]

On 4/2/2020 at 12:17 PM, StoneToad said:

3 eggs every 5 cycles for a cost of 3-6kg algae or so per 5 cycles, or 1-2kg of algae per egg

Nice design. Regarding these numbers, perhaps I'm missing something but how can it cost 3kg algae if there is a 5 cycle "lifecycle" and the sweeper is timed to deliver every cycle?

Does the Pacu know that falling through the open door will lead to water or it doesn't and you're hoping it doesn't stay where it is for a long time?

[StoneToad]

The Pacu is only in the breeding chamber for ~3.33 cycles, so only gets to eat from the feeder 3 times. the rest of the time the breeding chamber is empty.

But yes you're right, the MkI design here actually uses 4kg per 5 cycles, since it loads the feeder even when there's no pacu so the first feed is always a double at 2kg.

It also turns out that the MkI is a royal pain to build in survival so I've made a new design using ribbon cable that also fixes most of the issues I've found with the old design, and should be easier to build but I haven't had a chance to do final verification in survival yet.

I'm still can't decide if flopping pacus do some kind of pathing towards ledges and/or water or not, but it doesn't take anywhere near a full cycle in my experience for the pacu to flop off the one tile ledge and fall out of the breeding chamber.

[Prince Mandor]

On 4/11/2020 at 12:11 AM, StoneToad said:

...I've made a new design using ribbon cable that also fixes most of the issues I've found with the old design, and should be easier to build...

Can we see MkII, please?

[StoneToad]

Yes!  I had a weird issue in my survival world where the breeding chamber was cycling even though I hadn't loaded any pacus yet, and finally tracked it down to a bug with critter sensors and rooms on load.

So long story short, the critter sensor is disabled at map load with no valid room message. It's in a room with no eggs, with an air containing wall tile (airflow, pneumatic door) separating it from a room with an egg.  If you watch the room overlay with the critter sensor selected, you'll see as you unpause that the room overlay updates in 2 passes, first it makes all the tiles with gas into rooms, *bug the critter sensors think it has a valid room starts updating and goes green* then trims the rooms at all the airflow and pneumatic doors.  Now the critter sensor is properly reading 0 eggs and showing red.

This spurious  egg reading had me rather stumped as to how to deal with it as closing off the device with airlocks doesn't solve having eggs in transit on the conveyor lines causing problems on save/load.   But I saw a post on the forum where someone suggested using a filter, luckily I managed to find a way to jam one in!

And now on to MkII! special thanks to Better Automation Overlay mod for the automation overlay markup.

The conveyor bridge and line with the loader on the right side can just be ignored.

Note on the automation, the reversed ribbon writer just above the and gate is connected across a single ribbon, it's replicating the signal on bit 2 to bits 3 and 4 (one tick later, but that's fine)  Note that yes that timer is *green* for 596s and red for 4s, the critter sensor is also green if the room is empty, suppressing filling the feeder when there's no pacu.

Note on the conveyors, the top left, there's a up-down conveyor bridge that bridges onto a loop made by the shutoff, bit hard to see in the picture.  There must be 2 tiles when the egg is outside hatching room before it reaches the shutoff so sensor can respond, same for the distance the egg is inside the room before the vent.  Takes about 2 tiles for the sensors to respond.

Do note that I only updated the automation layer image with the filter for the bug workaround, so the counter is in the wrong location in the others.

This should have fixed pretty much all the bugs in the old design, It was also fairly easily buildable in my survival game, though it's currently sitting empty since I discovered I had >130 wild pacu and quite literally a tonne of lime already.