Jump to content

Fastest Door Pump Possible


Recommended Posts

I've spent a bit of time optimizing one of my favorite door pumps. Credit where it is due, the original automation setup is a 3-door design I shamelessly stole from @goatt from his thread on steam turbine heat deletion. I chose this specific design because it was the most efficient and fastest at pumping, with zero chance of backflow, and is very simple wiring.

I gutted his original design, cut out any possible unneeded inputs and extra wiring, and here it is:

door_pump.png.0d66e7ebde26fd1be262abc44b5e4940.png

Simple explanation, this setup is designed to be powered, and has zero chance of backflow. The signal switch can be toggled to lock it off on demand.

You may be wondering, how is this the fastest possible design? Couldn't it go faster than a 3 second cycle?

The answer is yes.

 

Now presenting, the fastest door pump possible:

door_pump_fast.thumb.png.5a42602bcb8bb0dfc5c89be1bb0d3da3.png

THIS is the true fastest pump on the market. Designing this took some pushing the boundaries of what the doors limits are, and it functions flawlessly. This screenshot is an exploded view of the automation required. It's about twice as fast as @goatt's original at moving gas.

Some technical info, doors open/close in exactly 0.2 seconds. Since Klei has so kindly forgotten how to count higher than 0.1 seconds, or smaller than 1 seconds, stacking buffers and filters set to 0.1 seconds is required.

Condensed Automation overlay:

Spoiler

door_pump_fast_condensed.thumb.png.5d19b7003537aebe53a997d6ca8246e4.png

For those who want a live view of it in action, here's my test bed: Sandbox.sav

 

I did some testing with multiple door compressors, and none of them even came close to the new design. The closest design for pumping speed was @goatt's original setup, with my tweaking. Here's a few 4-door and 3-door designs that I tested it against, the new design is about 2-3 times faster than any of them.

Brothgar 4-door compressor:

Spoiler

Door_Compressor.thumb.png.f24fa2f4d10646c35c471944971a58fa.png

@mathmanican's Steam Behemoth pump

Spoiler

steam_behemoth.thumb.png.18682ac16698830883bdd4c3c9b25d53.png

@Soulwind Compressor

Spoiler

image.thumb.png.1bd2540a61ff27624d49e9b63e573f35.png.2b7a7bcac85a0da7a26a8cf9ec778e80.png

 

Link to comment
Share on other sites

Powered doors for a pump? You're crazy.

In my experience, the speed the doors close at is not important in the slightest, more the amount of vacuum you create before compression begins.

I.e. in my aquatuner build, opening a bay of 18 doors to create a huge vacuum into which the steam floods means that i'm moving ~12kg per tile x 16 tiles (factoring for a complete loss of gas on initial right side closing) = 192kg of steam moved.

image.thumb.png.67ce954c28cd93dae2d845ebc8d1e2f2.png

Basic automation keeps the bottom set of doors closed when the rightmost doors are open.

image.thumb.png.15439bdfe4bb9ffb4d60ce8be2b366dc.png

Rightmost doors close/bottom doors open - compression starts.

image.thumb.png.32330613914104a8b3d0a949ea4233da.png

This is triggered by clock sensors (you can slap as many in as your buffer settings can accommodate) that a few times per cycle move a huge amount of steam. Less operations = less lag. Buffers have 2 seconds between them to make sure no gas is lost. 

image.thumb.png.b9b09dacd505537d9f0c4620968d069d.png

Hope this makes sense - the tl;dr is that the faster/more frequently you close your doors, 1) the higher the chance you delete materials, 2) the more impact you have on game performance.

Link to comment
Share on other sites

14 minutes ago, Lifegrow said:

Hope this makes sense - the tl;dr is that the faster/more frequently you close your doors, 1) the higher the chance you delete materials, 2) the more impact you have on game performance.

I tested for the gas deletion, and over several cycles of running it didn't lose anything, even when save/loading. This particular door compressor is reliable.

You have a fair point that lots of fast automation can have an impact on CPU load. What I'm hoping for is that Klei will fix buffers/filters to give the entire range of 0.1-0.9 selection, and the first setup I posted would be the best, with only 6 active components instead of 23. Showcases like this are in part to get their attention that such things are VERY much needed and requested.

I prefer powered doors because even though they can have an impact on performance, I dislike expansive, sprawling power systems. Density and simplicity are key points in my builds. The power useage is minimal compared to the tradeoff for density.

46 minutes ago, Lifegrow said:

Powered doors for a pump? You're crazy.

In my experience, the speed the doors close at is not important in the slightest, more the amount of vacuum you create before compression begins.

I.e. in my aquatuner build, opening a bay of 18 doors to create a huge vacuum into which the steam floods means that i'm moving ~12kg per tile x 16 tiles (factoring for a complete loss of gas on initial right side closing) = 192kg of steam moved.

Can you post your save file? I'd love to tinker with your design, and see it live. I know you do streams, but I don't often get the privilege to watch due to time and bandwidth constraints.

Link to comment
Share on other sites

The amount moved per second is entirely dependant on the pressure above the doors as that dictates how much gas gets pushed into the pump each cycle. If you're using a steam turbine the whole system naturally self regulates to 10kg/sec as that is how much the turbine moves back up the system. If the doors move less, pressure above builds which makes the doors pump more. If the doors pump more, pressure above drops making them pump less. Just run your turbine chamber at high enough pressure and any simple ass unpowered 3 door design will work. I tend to go for about 40kg/tile steam, personally.

Link to comment
Share on other sites

@crypticorb I copied your design and must say I'm mildly underwhelmed. It deletes 20% of gases and the throughput rate is atrocious in my opinion. It may flap the doors as fast as they can go but unfortunately the gas simply can't keep up filling it unless you have very high pressure levels. It also have significant backdraft at lower pressure levels in my observations.

Here's my set up if you want to trouble shoot it.

image.thumb.png.5157800f66502a81ebfb21bfe4e3e54f.png

5 hours ago, crypticorb said:

I tested for the gas deletion, and over several cycles of running it didn't lose anything, even when save/loading. This particular door compressor is reliable.

Unfortunately, I cannot confirm this. Quite the opposite in fact.

Link to comment
Share on other sites

4 hours ago, Saturnus said:

@crypticorb I copied your design and must say I'm mildly underwhelmed. It deletes 20% of gases and the throughput rate is atrocious in my opinion. It may flap the doors as fast as they can go but unfortunately the gas simply can't keep up filling it unless you have very high pressure levels. It also have significant backdraft at lower pressure levels in my observations.

Here's my set up if you want to trouble shoot it.

image.thumb.png.5157800f66502a81ebfb21bfe4e3e54f.png

Unfortunately, I cannot confirm this. Quite the opposite in fact.

I checked over your wiring, and it looks correct. It should mirror the original timing of the 3 second design, just faster.

I redid my tests. All tests were over a 3 cycle period, at fast speed.

  1.  First test: 1000g pressure at the bottom, testing for backflow and deletion. Zero backflow after test conclusion, even when alt-tabbing away. All gas accounted for.
  2. Second test: 100kg pressure at the top, testing for throughput. The 0.6-second design moved the gas about twice as fast as the 3-second version, and all gas was accounted for to 3 significant digits.
  3. Third test: 100kg at the bottom. I did encounter backflow here, but only after I alt-tabbed away with the camera focused elsewhere, and it was about 500mg of gas total. I double checked the bottom chamber, and all gas was accounted for, even the milligrams of gas that got back to the top.

Did you use super-speed to test the doors? I can't find any issue with the setup. I also plan to build this onto my steam turbine in survival, to see if lag starts becoming a problem.

Here's the save file with the 3 tests on the far left, all set up. Feel free to tinker. Sandbox.sav

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

Please be aware that the content of this thread may be outdated and no longer applicable.

×
×
  • Create New...