Microcycle reactors

Microcycle reactors are reactors that works in small bursts, then cool down for a while. It allows to have little or no internal cooling components, usually just being a big pile of uranium in thin sheet of metal.

What is good about them?

First, they are most efficient. You can even reach theoretical maximum of 4.44 effectiveness, though it is not really practical.
Second, they are cheap in everything, except uranium. For them, uranium is not just fuel - it is constructive part, because they can run 80+ hours from one load.
Third, low maintenance. As I said, 80+hours per load, so you just make it, connect and forget about it. Like a solar/wind/water, but much more beefy.
And last, they are very safe. CASUC reactors have a lot of parts where something can go wrong. Mark-2 reactors can blow up if you forget to wait for it to cool down. For microcycle reactors, wiring is extremely simple, and human factor is minimum.

And there is one, but serious drawback. It is amount of uranium you need to get it started. Recommended is 27 per core, and it produces just 24.8 Eu/t (but for a long long time). So, if you need your Quantum here and now, it is not a best option.

Now about how it is done.

Simplest scheme is flow-limiter. You make a reactor with average 32Eu/t, or little more, or very little bellow (everything above 31.5 is ok). Then you filter output of it through single BatBox, not letting reactor run until BatBox is empty. EU part is simple - all you need is downconvert ouput to low voltage without loss. Redstone part is even simpler and consisting of one redstone torch and two redstone dusts. Batbox is set to "emit if empty". Redstone torch is sticked into it, effectively being lit when batbox is NOT empty. And redstone wire goes straight into reactor chamber. As a bonus, it also switches reactor off if there is nothing on the other side that can take energy.

On first picture there is a two-chamber flow-limited reactor. Load is following -
http://www.talonfiremage.pwp.b…n2kxe1ezyn7nl3wesro3ptsn4

There are two drawbacks, though. First, you are limited to 32Eu/t designs. You can make it work with other, of cause, with some splitomancy, but it defeats the idea simplicity.
Second - it does not work well with more than one reactor. Main reason is that reactors update once per second, and not synchronously with each other. So, you have to use more complex wiring to extend working period enough to allow all rectors to emit energy. And that, again, defeats the simplicity.

Nevertheless, I could make a quad core one that worked for many hours without exploding - see picture 2. 4 cores make nearly 100 EU/t, which is then cplit by 3, with "control" batbox receiving 33EU/t. Three repeaters extend "on" time to about one second.

So, now about "real" way to do microcycle. And that is - run a plant of many cores from one redwire and fed into one system of transformers/storages.

It can be done simply with timer. It should give "off" output for short period of time (couple of second), and then "on" for much longer period of time.

For example, for chamberless full-out uranium reactor ratio is 29 (divide "Active" EU/t by "Effective")
http://www.talonfiremage.pwp.b…vpaz1bc5mrejn58u1tja9kif4

So, if you want "on" period to be 2 seconds, "off" should be 2*29=58 seconds.

There are different ways to do that. If you have redpower, just use it's timers. Picture 3 has an example schematics. Timer of the left is "long" one (58 seconds in our example), timer on the right is "short" (2 seconds).
Thingie on the far right is just a "or" gate that can be replaced with one redstone wire. It is needed because redpower wire does not go "into" blocks. I learned it hard way - it was only time reactor blew in my face without me intending it:(

Or you can make timer from just Industrialcraft compenents. And it will look very cool.

I present you a Cubeclock! See pictures 4 and 5.

Because of oh so informative looks of Industrialcraft storage and transformers, it is better to tell what it is in words.

It is two parallel rings of BatBoxes and LVTs. They are connected so electricity can go only in one direction - clockwise, for example. So, in both rings direction of BB output should be clockwise, and direction of LVTs should be counter-clockwise (because it's their "input").

One ring is one-quarter rotated to another, so there is a BB of one rings against LVT of another.

Each BB is set to "Emit if empty" and have a torch sticked to it that transmit signal to LVT of other ring.

Rings "work" in order. Usually there are three of four BBs non-empty, so there is only one LVT allowing EU through. When all EU transimted, then BB before that LVTs becomes empty, unlocking other ring.

So, long story short, put 1200 EU in one ring and 18000 in another. Batteries give energy in 100 EU packets, so it is relatively easy to do it with them.

Then connect torch sticking of "lower EU" ring to zero-core all-uranium reactors. Mission accomplished.

Math for it is following. Complete cycle for Cubeclock consists of for cycles - two for one ring and two for other. Length of each ring cycle is 1 tic per 32 eu in it, or 1 second per 640 eu.
Torch on one BB is unlit for length, of it's ring cycle, plus three or four lag.

So, if we want two second "on" cycle, we put 2*640 = 1280 EU. Then we remove 80 EU to compensate that lag.
Now, "off" cycle should be 2*29=58 seconds. But it consists of two "big" ring and one "small" ring cycle. So, we should substract from that time duration of "off" time and then divide remainder by two. (58-2)/2=28 seconds. After multiplying by 640 and rounding we have 18000 EU.

Hopefully someone can understand it:)

Next time may be I'll discuss best way to place dozens of reactors in small space. "Forest" allows better wiring, but is hard to water properly. "Wall" is much easier to water, but harder to wire, etc.

EDIT: Any idea why that goddamn url BBCodes are not working?

Quote from jppk1

Because there's a link button?
IMO you're overcomplicating stuff, two times, a few wire and a toggle latch allow you to make a system that inputs a redstone signal for n seconds every n seconds. I knew most of the other stuff, I just don't use this exact same method as you.

I have vanishingly small experience with Redpower, so any improvements on design are welcomed.
I know that if you want simple n/n or n/3n ratios, it can be done with one sequencer. But I need more complex ratios, such as 2/58. Do you have example of simpler designs that do that?

Quote from Rick

If a nuclear reactor is making 25eu/tick then why do you even make a nuclear reactor in the first place?

Because nuclear plant can have dozens of reactors.

EDIT: and obviously, BBCode hates me now

jppk1
Isn't it exactly same as my design (third picture)? Two timers, one latch.

GregoriusT
Yes, I don't trust RP a lot too, but they seem to be ok in simple designs, that can start from any position. I am not expert, though.

Eutronic timers are reliable in everything, except for measuring tick-precise amounts of time. Small amounts of EU keep getting stuck in transformers and storages, throwing off calculations. And it is much worse with slowires. Cubeclock is extremely simple design, but even it fluctuates a little. That is why it took much longer for me to make adequate scheme for energy reactor, than for breeder.

Quote from Rick

Geothermal works too. You can also use wind or solar. Its pretty expensive to make a reactor just for 25eu/tick.

This reactor is just one core without chambers or internal cooling parts. It costs only 10x of solar in metal and only 4x in resin. So, it provides 5x more energy per metal.
Only expensive construction part is uranium, but you can use mostly-empty uranium cells for it. And good breeder can make enough uranium cells from four uranium ores to feed one microcycle core for a lifetime.

Quote from Gaxx

na metal becomes cheap once you get your miner running. the really expensice materials are lapis and redstone, maybe glowdust, which you all need to craft those chambers and the reactor. also don't forget that you mentioned a breeder, which is another reactor that costs materials.
the design itself is interesting, but i'd prefer 4 watercells (24EU/t) with power converters for this which would run infinitly.
your eutronic experiments are more interesting
on a side note: just learn some boolean algebra and logic, then RP2 circuits will do just what you want. playing logic based games like "codex of alchemical engineering" or "spacechem" helps a lot to improve logical thinking in bigger circuits

I think hardest part in making many reactors is not getting metal, or uranium, or lapis. Real PAIN is juggling all those non-stackable components when you craft it. Especially coolant cells, that you also have to take one by one from condenser.

Experiments are fun when you have some objective. Such as making effective, reliable and elegant reactor. This goal made me explore IC electricity, redstone, nuclear explosions (and protection from them), other reactor quirks, of cause, and even water flow rules much deeper than I knew it before.

And, of cause, I wouldn't be able to do it without knowing "some boolean algebra and logic"

Thanks for links. I'll try them.