Increasing pipe pressure using feedback loops

[Mijae]

There have been a few posts asking how to increase pressure in pipes.  That is, right now gas valves can reduce packet sizes, but not increase them (by merging them together).  However, this can be achieved using feedback loops.  This design is related to the solution I posted for increasing efficiency of thermoregulators prior to TU.  i realized the flaw in that design (hitting pipe size limit) was actually the feature people were now requesting.

This is 3 nested loops each of which doubles the size of output packets.  They stack on each other so this increases packet sizes by 8x.  It requires a constant supply of same-size input packets for stability. You'll need to be able to estimate the average flow of incoming packets and use a valve to normalize the stream.  In this example, the valve is set to 100g/s on the right and packets come out at 800g on the left (every 8s).  Getting the packet sizes you want takes some math to set the valve limit and number of loops based on your expected input rate.

Examples:

1. Turning this valve up to 125 g/s will max the pipe out at 1000g packets (if your input is at least 125 g/s).
2. If your input is less than 125g/s and you want 1000g packets, you can add another outer feedback loop to increase the packet sizes to 16x (every 16s).  This will max out the pipe with the valve set up to 1000/16 = 62.5 g/s.
3. You can reduce the setup to 2 loops to increase packet sizes to 4x and set the valve up to 250 g/s.
4. You can split your input prior to the valve and have parallel recursive loops depending on input rate or overflow into some other system

Setting valve higher than (1000/multiplier) will either backup the system to that rate or could cause packets to emit faster than the multiplier rate with some at 1000g and some less than 1000g.

[Kasuha]

That's pretty neat idea. Here's a more compact arrangement in action:

I found this very useful thing to do if you're making vacuum in a room and are pumping the contents to a filter for further processing. Saves a ton of power on the filter.

[Risu]

Figured I'd just try out fixing the valve and I can see why they opted not to have it working correctly.

As you can see it blocks the pipe while it waits for more mass. In a mixed system this will just break it entirely.
 

[Kasuha]

2 hours ago, Risu said:

As you can see it blocks the pipe while it waits for more mass. In a mixed system this will just break it entirely.

It would need to store the unfinished packet in itself and if there is a different material on input, release that packet even if it's not the set amount.

 

[Risu]

1 hour ago, Kasuha said:

It would need to store the unfinished packet in itself and if there is a different material on input, release that packet even if it's not the set amount.

 

Adding a storage would only complicate things. People already can't comprehend there being packets under the icons in the overlay.

Edit: So much clearer.

 

[MoMuffins]

On 4/18/2017 at 11:49 PM, Risu said:

Adding a storage would only complicate things. People already can't comprehend there being packets under the icons in the overlay.

There's always been a packet under the overlay whether you can see it or not.  If you mouse it just right on the valve, you can see it.  They should just show the internal storage in the detail window.

On 4/11/2017 at 10:35 PM, Mijae said:

There have been a few posts asking how to increase pressure in pipes.  That is, right now gas valves can reduce packet sizes, but not increase them (by merging them together).  However, this can be achieved using feedback loops.  This design is related to the solution I posted for increasing efficiency of thermoregulators prior to TU.  i realized the flaw in that design (hitting pipe size limit) was actually the feature people were now requesting.

This is 3 nested loops each of which doubles the size of output packets.  They stack on each other so this increases packet sizes by 8x.  It requires a constant supply of same-size input packets for stability. You'll need to be able to estimate the average flow of incoming packets and use a valve to normalize the stream.  In this example, the valve is set to 100g/s on the right and packets come out at 800g on the left (every 8s).  Getting the packet sizes you want takes some math to set the valve limit and number of loops based on your expected input rate.

Examples:

1. Turning this valve up to 125 g/s will max the pipe out at 1000g packets (if your input is at least 125 g/s).
2. If your input is less than 125g/s and you want 1000g packets, you can add another outer feedback loop to increase the packet sizes to 16x (every 16s).  This will max out the pipe with the valve set up to 1000/16 = 62.5 g/s.
3. You can reduce the setup to 2 loops to increase packet sizes to 4x and set the valve up to 250 g/s.
4. You can split your input prior to the valve and have parallel recursive loops depending on input rate or overflow into some other system

Setting valve higher than (1000/multiplier) will either backup the system to that rate or could cause packets to emit faster than the multiplier rate with some at 1000g and some less than 1000g.

Yes, this is exactly what I've been wanting.  I tried to do it with pipe valves and the solution was rather unstable.  I never thought of doing it with pipe bridges.