Double mechanical H2/O2 filter

[UrmaneHendrake]

I'm moving away from centrally processed gas separation to more filter-at-source, and I'm also trying to reduce power consumption.  I put this oxygen and hydrogen filtering mechanism together for my early-game electrolyzer setup, and thought I'd share for feedback and hopefully answer a few questions before cratering testing on a real base.

double mechanical filter

Piping schematic

Input pipes bottom right.  Output oxygen and hydrogen on top.  Contaminant gases out bottom middle.

The rest is a mirror-imaged setup of two hybrid electric/mechanical filters.

Inside are two mechanical filters a la PVD's post.  Outside that are two gas filters to prime them.  Outside that are two gas valves to bypass the gas filters.

 

Set the inside gas valves to 1g, the outside gas valves to zero, and the filters to hydrogen and oxygen, aligned to which input pipe you normally expect each gas to come in on (I have at least two pumps in my electrolyzer room).

After the inner loops are primed, set the two outer valves to 999g.  This will bypass the gas filters and continue to run both using no power.

Since my electrolyzer gasses always seem to unbalance, the additional bridges in the middle allow oxygen and hydrogen to recombine and properly output if they get pulled in the wrong source gas pump.

Contaminant gases will (usually) cause the gas filters to turn back on until they are flushed.

 

Questions:

- The pipes top left and top right must route around the gas filter output.  If I run it through that square, it locks up - is that normal, and I just never noticed?

- The tiny little bump on the right seems to be required because gasses that fail to go over that bridge won't go travel to the pipe above, instead they lock up ... unless that extra loop is there, then they alternate paths.  Any idea why that is?

 

Like I said, haven't tested it live, only sandbox, but I'm liking it - it's spaghetti, but it's just a bunch of pipes and it'll save a lot of power if it's sturdy.  Thanks!

 

 

 

[Yunru]

Not quite sure where the "mechanical" part comes in, all I'm seeing is powered filters. 

[0xFADE]

You can do that with a set of mechanical filters only.

 

[JRup]

6 hours ago, UrmaneHendrake said:

- The pipes top left and top right must route around the gas filter output.  If I run it through that square, it locks up - is that normal, and I just never noticed?

If you crop the specific section we might be better able to see where this is going.

6 hours ago, UrmaneHendrake said:

- The tiny little bump on the right seems to be required because gasses that fail to go over that bridge won't go travel to the pipe above, instead they lock up ... unless that extra loop is there, then they alternate paths.  Any idea why that is?

You have more than one white (input) & green (output) combination on the same line. It will generate ambiguous paths for the gas and will stall the line. The extra bump kind of helps but wont clear the problem altogether.

Pooling inputs and outputs together on different sides is more stable and will avoid this kind of problems. Plan the route and you'll avoid rebuilds. Spoiler alert: the conveyor rails is less flexible and will give you headaches because of this. With this in mind, just make sure every line has at least one green and white "terminal" and rest assured that greens play together easier than mixing whites and greens.

The pliers mod will help you get experience with this far faster than building and deconstructing pipes every time.

[dallion]

What a mess... (no offence)

I wander what is the difference between this and double gas filter

[Saturnus]

It's already been made before
 

 

[UrmaneHendrake]

Thank you 0xFADE and Saturnus, I will look carefully at those - an automatic splitter with *no* power would be even better

Yunru and dallion, it is a mess     Here's some more pics to try to explain (tho it looks like the other suggestions may be better, so perhaps this is moot).

 

Here we see input bottom right of the picture, and the right-hand gas filter, set to hydrogen, priming the right-hand mechanical filter.  The left-hand filter will do the same with the oxygen when it gets there:

Priming the mechanical filters

 

Here we see the oxygen mechanical filter primed, to the left of the center pipe, and the hydrogen mechanical filter primed to the right.  Inputs are still going thru the gas filters, into the mechanical filters, and O2 and H2 exit the mechanical filters up out the top of this picture.

We also see down the center O2 and H2 that came in the "wrong" input pipe get a second chance to hit the mechanical filters - that big H2 packet right in the middle will go over the bridge to the H2 loop and exit up.  This was a big goal for me in trying to make this.

System primed and running

 

Here the outermost gas valves have been opened, and the path bypasses the gas filters - the system will continue to run with no power, filtering O2 and H2, even if they get mixed in the input pipes.  Any contaminant gases, in either input pipe, exit down (not shown).

Running with no power

 

Thank you for your comments - this was a fun mental exercise, and I appreciate your comments, even if there's already a better way

[UrmaneHendrake]

15 hours ago, JRup said:

If you crop the specific section we might be better able to see where this is going.

You have more than one white (input) & green (output) combination on the same line. It will generate ambiguous paths for the gas and will stall the line. The extra bump kind of helps but wont clear the problem altogether.

Pooling inputs and outputs together on different sides is more stable and will avoid this kind of problems. Plan the route and you'll avoid rebuilds. Spoiler alert: the conveyor rails is less flexible and will give you headaches because of this. With this in mind, just make sure every line has at least one green and white "terminal" and rest assured that greens play together easier than mixing whites and greens.

The pliers mod will help you get experience with this far faster than building and deconstructing pipes every time.

I'll look at the pliers mod, thanks!

Here's a markup showing where I had questions:

markup

The yellow box shows where I need to route around the gas filter output.  I had thought I could run that pipe through that green square, but I guess I just never noticed one cannot?

The red boxes show different behavior - on the left, packets can flow past the bridge entrance if they can't go over.  On the right they cannot, that extra loop is needed.

[Yunru]

Apologies, what I meant was, why is there powered filters?

Does this not suffice?

(Note: Gas travels up the central pipe.)

[JRup]

Hmm, ok... I kind of see where you want to go with this. I'll give a sequence of events for this, starting on the left side (after that you may see what's happening on the right side).

Ok, yellow square's issue is that the routing may be ambiguous for the game. Let's go step by step.

The first valve's output exits in the bridge with the small red square. So far so good.  If the bridge cannot place the gas on the other side it bypasses the bridge and goes towards the second bridge that's connected to the exit of the powered gas filter. And if the second bridge can't place the gas it will bypass into the output from the powered gas filter. And this will also happen the other way around.

So here, a couple of screenshots with what I'm saying:

 

So if you think about it, this is the way the line stalls and won't circulate on this branch.

 

[JRup]

So, I kinda wanted to see how it went editing with GIMP, note the wonderful mock-up. This may give an idea of what could be done with your setup.

You could later make a more compact build when looking into filters that only use valves like others have posted. I only wanted to address the original question of how a particular design was giving issues and possibly suggest a fix or workaround. (Also polish my shoddy image editing capabilities. )

Also, just a couple of arrows into the bridges that were "added" for clarification:

[UrmaneHendrake]

6 hours ago, Yunru said:

Apologies, what I meant was, why is there powered filters?

Does this not suffice?

(Note: Gas travels up the central pipe.)

How does one assure that you get the right gases, and on the right loops?  For instance, what if you get O2 in both?

 

[JRup]

27 minutes ago, UrmaneHendrake said:

How does one assure that you get the right gases, and on the right loops?  For instance, what if you get O2 in both?

You have to pre-seed the valve's loop with the gas type you need before letting it work regularly (don't forget to  set the valve to 1 g/s). Also note that in this picture the gas direction counts, so the entry point for the unfiltered gas will start at the bottom, for example. The exit points for "filtered" gas will be at the valve's input.

BTW: The curved pipe segments are only there to be able to see which gas is being filtered.

A more important thing to take care of is that this kind of filtering setup is more like an extractor of sorts and may present issues to deal with if the lines get full. You won't get the wrong gas in the exits ever, but if any exit is filled then the rest of the gas continues through. A powered gas filter would stop working with a "pipe blocked" status if you don't set alternative routing, for example.

[UrmaneHendrake]

Quote

You have to pre-seed the loop with the gas type you need before letting it work regularly.

aaaand that's why I have powered filters

 

I've had time to look through the thread Saturnus posted, that and Yanru's pictures are far simpler.  The complication I had came from two input pipes - reducing it to one allows the much simpler setup.  I'll use gas filters until the loops are primed, then just deconstruct them and put in two bridges.

 

BTW, I'm building blueprints in my sandbox - that's why I had both paths initially, the idea was build it once and done ... but it went down the rabbit hole

 

Thanks all, I really appreciate it!

 

[KittenIsAGeek]

On 9/29/2020 at 2:07 PM, Yunru said:

Not quite sure where the "mechanical" part comes in, all I'm seeing is powered filters. 

I looked at it for a while. It appears the powered filters were there to make sure the gasses went in the right places to prime the valves.

11 hours ago, JRup said:

So here, a couple of screenshots with what I'm saying:

 

So if you think about it, this is the way the line stalls and won't circulate on this branch.

Correct.  The input (white arrow) in the bridge will take any gas, as long as there is room for it to arrive on the other side (green arrow).  So, if the pipe on the Green side of the bridge is full, then excess gas will proceed up the pipe where you've drawn the arrows.  That section is called an 'overflow' and can be used to detect if your output pipe is too full and you need to stop filtering for a bit.

BTW, you can fix the backflow problem on your overflow by removing your powered filter, since it isn't doing anything now.

EDIT: Nevermind.. don't remove the powered filters.  Your valve filters aren't configured correctly and you'll get mixed gasses.

 

Your input bridge should tie in here:

[Yunru]

7 hours ago, UrmaneHendrake said:

How does one assure that you get the right gases, and on the right loops?  For instance, what if you get O2 in both?

Personally I build a filter, run a gas through it, then build the next filter and repeat.