Am I missing something with the latest patch?

[Guest aresd]

Why is the pump now pumping 700-900 grams of gas? has the mechanics changed somehow? Why can the pump now accumulate up to 1000 grams before it sends gas to the pipe?

 

[Yoma_Nosme]

They've always worked this way...if you want full packets of any kind you have to "stack" them with a pipe loop or use a reservoir as a buffer

[Guest]

2 minutes ago, Yoma_Nosme said:

They've always worked this way...if you want full packets of any kind you have to "stack" them with a pipe loop or use a reservoir as a buffer

This is the first time I see a pump accumulating gas. I always built the oxygen generator in this way and there were no problems. And now I go back into the game and the old circuit already crashes, because instead of 500 grams of gas, the pump delivers a whole kilogram into the pipe

[Yoma_Nosme]

Btw...or use a atmo sensor and automate the pumps to only pump over 1500g for example you have 1 electrolyzer producing a max of 1000g ... maximum because of deletion there's less 

[Guest]

1 minute ago, Yoma_Nosme said:

Btw...or use a atmo sensor and automate the pumps to only pump over 1500g for example you have 1 electrolyzer producing a max of 1000g ... maximum because of deletion there's less 

Only one pump works in the screenshot. I specifically turned off all unnecessary for the test.

I just checked in the sandbox. While the pump pumps only one type of gas, it works normally and steadily pumps, not exceeding the limit of 500 grams. If you add a second gas, the pump starts to work according to another mechanism - it gives from 500 to 1000 grams to the pipe. What's happening?

[werecat]

35 minutes ago, aresd said:

If you add a second gas, the pump starts to work according to another mechanism - it gives from 500 to 1000 grams to the pipe. What's happening?

It's still the same mechanic of pumping a maximum of 500 g/s. What is happening is the pipe sometimes gets blocked because the other gas is in the way, so the gas pump can grab another packet of the same gas and it stacks on to the previous packet

[Guest]

One pump - one pipe to ensure that the queue is excluded. However, when adding three times the gas, the pump starts pumping from 500 to 1000 grams of gas instead of the maximum 500 grams.

3 minutes ago, werecat said:

It's still the same mechanic of pumping a maximum of 500 g/s. What is happening is the pipe sometimes gets blocked because the other gas is in the way, so the gas pump can grab another packet of the same gas and it stacks on to the previous packet

In total, I play about 2 thousand hours and the first time I see such a model of pump behavior. 

5 minutes ago, werecat said:

It's still the same mechanic of pumping a maximum of 500 g/s. What is happening is the pipe sometimes gets blocked because the other gas is in the way, so the gas pump can grab another packet of the same gas and it stacks on to the previous packet

So the pump is now not working accurately enough? If before the limit was 500 grams, now this limit can vary? Increased pump buffer?

I do not exclude that all this time I used some kind of bug and now this bug has been fixed. I have already met more than once with the fact that it works for me, that for others it works in a different way or breaks

[SharraShimada]

Whats the gas pressure in the chamber with the vent?

No, the max gas pressure a pump can build up with one cycle is 500g. Thats a fixed value. But it can add 500g to already existing 500g up to a maximum of 1000g, what is the max-pressure of a pipe segment.

If SOMETHING interferes with gas flow, the gas packets will stop for a brief moment. And thats the chance for the pump to add more gas to a already existing packet. 

Imagine a conveyor with buckets. Every bucket has a capacity of 1 liter. 

You have a smaller bucket in your hand with 1/2 liter.

Now let the conveyor run steady, and you fill 1/2 liter in every bucket on the conveyor. 

Now the conyeor stops (because reasons) for a second or two, and you keep filling stubbornly. 

What happesn is, you fill the same 1L bucket with your 1/2L bucket twice, before it moves again. 

You´d never filled a bucket with 1 liter at once, because you cant. YOUR bucket is too small to do this.

 

 

 

[Gurgel]

The routing in the pipes has changed somewhat and that can lead to gases stacking because of blockages that were not there before or changes in flow priority.

[nakomaru]

3 hours ago, aresd said:

it gives from 500 to 1000 grams to the pipe. What's happening?

I also believe it has always worked like this (at least 2 years). My understanding is pumps are storage containers with about 1kg capacity, and do the following:

T+0s: Store up to 500g (pump throughput) of stuff from the 5 tile plus sign. For example 50g H2 400g O2
T+0s: Output up to 1000g (pipe throughput) of the first element on the list. For example, 50g H2.
T+1s: Take in another 50g / 400g O2. Now we are at 50g H2 / 800g O2
T+1s: Output up to 1000g (pipe limit) of the next element on the list. 800g O2.
T+2s: 100g H2
T+3s: 800g O2
etc

[wachunga]

It's been like this as long as I can remember. If you search well enough you'll find people arguing about it from years ago.

 

The general idea of it is that pumps absorb up to 100 grams from a 5 cell + shaped area 5 times a second into an internal storage. First from the top cell, then middle, bottom, left, and finally right. If any of those cells are blocked, then one of the previous cells is reused. Once per second the pump ejects from the internal storage into the pipe at it's output. The ejection is a round robin type of system, but it can go a bit wonky at times.

Imagine a pump at a gas boundary such that hydrogen is in 1 of the absorption cells and oxygen in the 4 other cells. After one second, it's storage will be 100g of hydrogen and 400g of oxygen. The pump can only eject one type of gas into the pipe, let's say it is hydrogen. There will now be 0g of hydrogen and still 400g of oxygen in the storage. After the next second, the internal storage will be 100g of hydrogen and 800g of oxygen. The pump will then eject the 800g into the pipe. This cycle repeats.

Note that sometimes the 5 times per second absorptions don't always happen. This seems related to how much is already in the internal storage, but not as a simple total. Some other odd metric is used. Most noticeable when the pump is taking in many different types of gas, rather byzantine to nail down exactly what's going on. This wonkiness is probably why you never noticed the behavior before.

Also note this 5 times per second absorption vs 1 time per second ejection is why pumps usually give small packets when they first start. Depending on how the timings line up, the pump may have only taken in 1-4 times before it ejects into the pipe.

 

Use alt- in debug to advance the game by one tick (1/5 second) to observe all these behaviors yourself.

[Guest]

Thank you all for your answers. I even found the game from the old patches and checked that everything works as you all say. Apparently, this was my computer that gave out tricks, because I used this scheme of oxygen generation many times and it did not pass oxygen into the hydrogen generator. Now, this leads to permanent breakdowns. I thought that after a long break, I did not notice any strong changes. But now I see that I just have hands out of my ass.
Took the advice of @Yoma_Nosme and used the tank as a buffer. That even greatly facilitated the construction of an oxygen generation station, there was no need to control overfilling of the system, the automation of the tank does everything.

[psusi]

4 hours ago, nakomaru said:

I also believe it has always worked like this (at least 2 years). My understanding is pumps are storage containers with about 1kg capacity, and do the following:

Nope.  I has not worked like this until recently where I believe they mentioned in the patch notes that machines will no longer block because the output pipe is only partially full.  In other words, if the gas pump wants to add 500g to the pipe and it currently only has 500g in it, then it will now do so, but before it would refuse.  You know how you used to have to NOT run a line of pipe straight over the outputs of your toilets or they would fail because a small packet of pwater from one toilet would block the next as it passed it by?  This no longer happens, but the same change affects gas pumps as well so they can also now add more to the pipe if it is somehow blocked.

 

What I don't understand is how the pipe is ever getting blocked in this case.

4 hours ago, wachunga said:

Imagine a pump at a gas boundary such that hydrogen is in 1 of the absorption cells and oxygen in the 4 other cells. After one second, it's storage will be 100g of hydrogen and 400g of oxygen. The pump can only eject one type of gas into the pipe, let's say it is hydrogen. There will now be 0g of hydrogen and still 400g of oxygen in the storage. After the next second, the internal storage will be 100g of hydrogen and 800g of oxygen. The pump will then eject the 800g into the pipe. This cycle repeats.

I've never seen that happen before.

 

[nakomaru]

10 hours ago, psusi said:

if the gas pump wants to add 500g to the pipe and it currently only has 500g in it, then it will now do so, but before it would refuse.

1. Not what happens (never has), and doesn't explain what OP sees.

10 hours ago, psusi said:

I've never seen that happen before.

Ever pumped a mixed gas? See previous post for why this happens.

[psusi]

34 minutes ago, nakomaru said:

1. Not what happens (never has), and doesn't explain what OP sees.

You appear to be showing how it used to work.  If I read the patch notes correctly, now the pump will add 500g to the pipe and you will end up with 600g in there instead of the pump saying "blocked".

36 minutes ago, nakomaru said:

Ever pumped a mixed gas? See previous post for why this happens.

Yea, and I've never noticed that happening.  I could have sworn it always resulted in 500g or less.

 

[beowulf2010]

1 hour ago, psusi said:

You appear to be showing how it used to work.  If I read the patch notes correctly, now the pump will add 500g to the pipe and you will end up with 600g in there instead of the pump saying "blocked".

I've actually had more instances of "pipe blocked" recently than I have in the past and have seen the behavior in Nakemaru's video in my current base as recently as 2 days ago.

I've just gone back to using bridges for all connections.  It's just safer for me and I still don't trust T junctions without bridging into them.

1 hour ago, psusi said:

Yea, and I've never noticed that happening.  I could have sworn it always resulted in 500g or less.

I've personally never paid enough attention to know if I've seen it before. I just know I've seen the occasional topic about it for the past year+ I've played the game.