New power generator setup? *UPDATED*

[Sven_The_Slayer]

Power is still generate the same, a single line of generators will generate the same amount of power as a similar box generator with the same number of modules

Now Thrusters on the other hand... holy crap. I'm starting to feel these may be a little unbalanced. I made a bunch of thruster shapes to see how this affected thrust output. 61 generators in a line put out a little over 100 thrust, that same number of generators in a + shape in all directions will put out over 8000 thrust. I'm all for having smart design giving you better results but that seems unbalanced to me. I kind of feel like large solid packs of thrusters should be a good way to generate thrust, sure it takes the strategy and design out of building but it seems to make more sense to have a giant thruster creating more push than a dinky array.

 

[SkylordLuke]

please dont change it again,i just learned how to use this one xD oh gawd

 

[Lt_Duckweed]

Like I posted in this thread earlier, here are the two thruster equations, old and new.

The old equation was:
( x+y+z-3) + (n^1.125)
Where n is the number of blocks

The new equation is this:
[(x-1)(y-1)(z-1)]+(n^1.125)
Where n is the number of blocks

So in Sven's case,

61 in a line;

[(61-1)(1-1)(1-1)]+(61^1.125) = (60*0*0)+101.98 = 101.98
Which rounds to; 102.0


61 in a cross;

[(21-1)(21-1)(21-1)]+61^1.125 = (20*20*20)+(101.98) = (8000)+(101.98) = 8101.98
Which rounds to; 8102.0

 

[Sven_The_Slayer]

Like I posted in this thread earlier, here are the two thruster equations, old and new.

The old equation was:
( x+y+z-3) + (n^1.125)
Where n is the number of blocks

The new equation is this:
[(x-1)(y-1)(z-1)]+(n^1.125)
Where n is the number of blocks

So in Sven's case,

61 in a line;

[(61-1)(1-1)(1-1)]+(61^1.125) = (60*0*0)+101.98 = 101.98
Which rounds to; 102.0


61 in a cross;

[(21-1)(21-1)(21-1)]+61^1.125 = (20*20*20)+(101.98) = (8000)+(101.98) = 8101.98
Which rounds to; 8102.0

Thanks for the equations, I started a thread specifically to discuss the thrust changes if you would care to pass on some input there. I really wonder what the best solution to thrust is as the current equation seems to be way to easy to reach extremely high thrust.
http://starmadedock.net/threads/how-should-starmade-generate-thrust.251/

 

[Keptick]

I'll make a detailed guide once I get home. Including all the tests made to verify the mechanics.

 

[Holy]

I don't know the maths, but I made a ship the size of a heavy fighter which generates 420k e/sec and is able to permanently jam while shooting and flying with a mass of around 700 units. (to perma jam it only needs around 350k but I also want to fly and shoot at the same time).

Well i see u guys went to thrust but i would like GROOV3ST3R to uplad his Generator image for that heavy fighter

 

[Cablestinks]

Well i see u guys went to thrust but i would like GROOV3ST3R to uplad his Generator image for that heavy fighter

Ditto,

I can get perma, until i hull it. by the time i get done with plain hull it peaks at over 500k requirement, not thrusting, not recharging the jump. I believe my mass was only 300 Power generated was a hair over 400k

 

[Cablestinks]

Like I posted in this thread earlier, here are the two thruster equations, old and new.

The old equation was:
( x+y+z-3) + (n^1.125)
Where n is the number of blocks

The new equation is this:
[(x-1)(y-1)(z-1)]+(n^1.125)
Where n is the number of blocks

So in Sven's case,

61 in a line;

[(61-1)(1-1)(1-1)]+(61^1.125) = (60*0*0)+101.98 = 101.98
Which rounds to; 102.0


61 in a cross;

[(21-1)(21-1)(21-1)]+61^1.125 = (20*20*20)+(101.98) = (8000)+(101.98) = 8101.98
Which rounds to; 8102.0

How did you get that (8k+) i tested 20 on x y z and only got 160 thrust.

Edit
i just tried 62 in a cross (like an 3d plus sign, 160

REedit
62 in a line 160

Im gonna have to asume they nerfed it

 

[Lidren]

It was changed back to big groups rather than line arrays.

 

[Cablestinks]

It was changed back to big groups rather than line arrays.

Yeah , i just noticed the date of the posts:P thank goodness too, the power thing is a big enough headache.

 

[GROOV3ST3R]

Wow, I completely forgot about this thread. Haven't even played SM since, for all I know everything has changed. There is no diagram for the ship I made - it's just crammed full of generators in the longest lines and largest dimensions possible.

Looks like I will have to start playing again and see if anything I know still holds true so don't trust anything I said :P For the record, yes, my ship made a ridiculous amount of power, but that was before jamming got it's requirements slashed - and everything else got ''rebalanced''. It's back to the drawing board.

 

[Snk]

It's a box thingy now. You gotta build your reactors in a box around the core. It's cubilar(?) instead of now linear.

 

[Sven_The_Slayer]

The only thing that maters for reactors is the maximum size in any given direction. A single line of 100 generators gives the same power as one with three arms each 33 meters long (plus the center). You do want to build reactors as large as you can though as the benefit for several large reactors is typically more than single sticks as they are limited to a single plane. Also if you cannot build reactors at least 9 chained try and spread them out as you get more power per block as long as none of the 8 are touching.

 

[GROOV3ST3R]

Ah so not that much changed. That's fairly reassuring.

 

[Bourne223]

As far as i know, If you Build a reactor to the following specifications, it will be the most space-efficient:

And it ends up looking like this:

I call this method the "Cascading Cuboid" method because the center is a cascading series of cube-boundaries:

So you end up with maxed out 3-D "L" shapes for the available space, all nested within one another. I usually fill in the leftover gaps of these power-blocks with power capacitors or shield chargers. Whatever suits your fancy.

And of course this design is not just for a 5x5x5 cube, it can be expanded for any shape or dimension by repeating the pattern. So long as no two groups have block faces touching, you are doing it right.

Is this still the method everyone is describing? I'm fairly certain this will yield good results for small and medium sized ships. My current Space Fairer Mk2 ship has a mass of 539.7, energy production of 70450.3 with only 383 generator blocks. On average that means 183.94 e/sec per block. I know that my energy setup with this ship is not perfect, but for where i have my main power generation set up, it is the maximum space: power generation efficiency.

I would love to hear the community's thoughts on what i have posted.

 

[Sven_The_Slayer]

As far as i know, If you Build a reactor to the following specifications, it will be the most space-efficient:

And it ends up looking like this:

I call this method the "Cascading Cuboid" method because the center is a cascading series of cube-boundaries:

So you end up with maxed out 3-D "L" shapes for the available space, all nested within one another. I usually fill in the leftover gaps of these power-blocks with power capacitors or shield chargers. Whatever suits your fancy.

And of course this design is not just for a 5x5x5 cube, it can be expanded for any shape or dimension by repeating the pattern. So long as no two groups have block faces touching, you are doing it right.

Is this still the method everyone is describing? I'm fairly certain this will yield good results for small and medium sized ships. My current Space Fairer Mk2 ship has a mass of 539.7, energy production of 70450.3 with only 383 generator blocks. On average that means 183.94 e/sec per block. I know that my energy setup with this ship is not perfect, but for where i have my main power generation set up, it is the maximum space: power generation efficiency.

I would love to hear the community's thoughts on what i have posted.

5x5x5 generators are a concept, a way to visualize how the generator systems work. They should never be put into practical use unless your ship has an empty space exactly 5x5x5 for it's power generation. You will always get more power for less room by building lines of power that run the entire length of the ship or as long as you can.

As for the design itself I see a lot of empty space not utilized, the most efficient 5x5x5 concept reactor I've ever seen had 4 arms that utilized the entire 5x5x5 space.

Edit: I followed the design schematic provided and filled in the empty space... that design is actually less efficient than a checkerboard of single reactors.

 

[Bourne223]

Of course 5x5x5 blocks are merely a concept used for illustration purposes ^_^ As for the design in my pictures, there was more added to it that was not pictured. For some reason I uploaded the wrong screenshot.

When you say "Arms" what do you mean? single 1x1x5 stacks? or do you mean snakes of reactor cores that are multidimensional as a group, but only 1 reactor thick? I would also like some numbers from said example.

I utilized 13 groups of 5x1x1 stacks and got an output of 7983.7 e/sec with 65 blocks. This means 122.8 e/sec / block. Simply put, single stack reactors such as this one are outdated and unoptimized. They should not be used. Reactor groupings need to have 3 dimensions maximized in order to fully optimize power output. using only one dimension greatly decreases the potential for power generation.

3D Checker-board yields 8873.0 e/sec, with 101 blocks. The efficiency here is 87.85 e/sec/block. Very low efficiency, but a higher overall power output.

I was able to create a 5x5x5 reactor that generates 9138.2 e/sec using 63 reactor modules. That means 145.05 e/sec / block, 1.65 times the efficiency of a basic checkerboard pattern. This method was similar to my demonstrated cascading cuboid method, but there is a two step cascade on the bottom, and on the top, in a symmetrical manner:

If you overlay Layer 1 and 2, and 4 and 5, you can see this "Double dual cascade" pattern. empty space was filled in the optimal manner.
I tried many different ways of filling empty space until i got the maximum energy output with this Double-dual cascade setup.

In layer 3, if you remove the very center block, you get a very slightly higher efficiency per block (145.12 e/sec/block) but the power decreases to 8997.3 e/sec as a whole.

It is my belief that this post finds that at the very least, 3D cuboid bounds are a more efficient method of generating power within a confined space, such as found in smaller vessels, than chunk-blocks, single-stacks, or checkerboard patterns (In fact, a solid 5x5x5 block of reactors only yields 4912 e/sec with 125 blocks for 39.3 e/sec/block). While not every vessel has exactly a 5x5x5 space available, at the very least a symmetrically designed network of 3d "L" shapes will provide an optimized power generation. Yes, it means a much longer time to build a reactor, and yes, it is really quite complex to construct these types of reactors once you get into non-cube geometry.

However as Sven stated, the most important part is to utilize the maximum dimensions available to create the largest grouping of reactors first, then fill in as you deem necessary.

 

[Sven_The_Slayer]

Checkerboard: all individual blocks, none touching. 63 power modules, 8873.0 E/sec (140.84 e/sec/block)

The best (afaik): 4 reactors snake around to cover the entire length of every dimension plus a center core. 61 power modules, 9727.5 E/sec (159.46 E/sec/block)

Any reactor chain that isn't at least 9 long is better off separate, that's why a 5x5x5 with 13 stick reactors is worse than one with 63 individual reactors.

 

[Bourne223]

Fantastic! is there any way this can be scaled up? Or any way that this principle can be generalized for use in larger-scale setups? Also what are your thoughts on multiple smaller groupings Vs. Fewer larger groupings (on larger scales)?

 

[Sven_The_Slayer]

Fantastic! is there any way this can be scaled up? Or any way that this principle can be generalized for use in larger-scale setups? Also what are your thoughts on multiple smaller groupings Vs. Fewer larger groupings (on larger scales)?

It may be able to be scaled up but I'm not sure. I mostly build as you call them "Cascading Cuboid" in my ships for simplicity sake and ease of building with symmetry planes. While I know more complex power reactors may get me more power I usually just start with a center reactor spanning the space I set aside for reactors and then from there "cascade" down.