Divide 4kw to two 2kw

[FIXBUGFIXBUGFIX]

transformer.sav

 

Maybe someone need it

Spoiler

 

Divide two 4kw to four 2kw

All gates are one second.

 

 

[Tunechi_sama]

Im a noob ,explain how the automation works. Does this cycle through the circuits and charge small battery at certain intervals?

[abud]

4 minutes ago, Tunechi_sama said:

explain how the automation works

AFAIK, those logic gates will alternate on and off in fast rates.

[martosss]

Another alternative is if you read my tutorial(also in signature), where I've shown how to separate each 1kW network and expose just 1 battery for charging. You need more batteries, but less logic + no conductive wire(or conductive wire only behind the power shutoffs in case of 2kW circuits as in the example above). It could be done with Smart battery + another battery per circuit.

• adding 2 batteries will ensure more stable power even if the power source blacks out for a little, so batteries take the load(20kW or 40kW can last for 10-20 sec vs 5 sec for small battery, you can potentially add more batteries to make a certain network safer, same for the example above).
• This solution is loading all cables between the power transformers and the power shutoffs with 1920W all the time, which means you can't connect more networks to the power transformers, so you're not using the full potential of Power transformers, which supply up to 4kW of Power, so 8kW per 2 Power transformers(you're using only 2kW of those).

[Promethien]

1 hour ago, Tunechi_sama said:

Im a noob ,explain how the automation works. Does this cycle through the circuits and charge small battery at certain intervals?

You got the gist of it. The transformer can pull 4k from the heavy duti but conductive can only handle 2k. However charging batteries doesn't count towards a wire's draw. So the set up alternates between charging the batteries while fully utilizing the transformer's full 4k draw potential. The problem is there is no point. Hes taking up just as much space and generating just as much heat as if he had just used multiple transformers. Also hes losing power to the batteries that have a natural drain.

Still, the concept is sound but I bet it could be more elegantly done with smart batteries with their built in toggle thresholds and an exclusion gate so they can't activate together.

[FIXBUGFIXBUGFIX]

42 minutes ago, martosss said:

1kW network and expose just 1 battery for charging.

I got your idea. Won't your design face the problem like this?

Quote

NOT GATE has a slight delay, hence when the signal changes, input and output of the NOT GATE are likely to be false at the same time. In this time, no power are supplied to refrigerators.

 

 

[martosss]

11 hours ago, R9MX4 said:

I got your idea. Won't your design face the problem like this?

 

 

I'd say the NOT GATE delay is a bug and has been fixed(or not completely? see below).

• You can't see the power change in 1 picture, but here's is an example of my general setup:
• 

Currently the timing is as follows:

• 0sec: Smart battery is drained to 5% so it sends 1.
  • Actually before it outputs 0, there is 1 on both sides, here's a pic where I managed to catch it!   If you're curious, the circuit is fully connected at this instant and shows consumers belonging to the circuit on the other side:
  • 
  • However, no overloading/brownout ever occurs, since it is probably calculated at whole second intervals, not at every tick(and this is obviously the first tick of that second). So as long as that circuit receives its 1kJ power juice during that 1 second(4-5 ticks, I think it was 5, but in the latest update they changed it to 4), it shouldn't brownout and everyone is happy.
• 0.1 sec(or 0.25 - the 2nd tick?): Not gate receives 1 and outputs 0.
  • 
  • Here you already have 2 circuits, so everything is fine and dandy - battery supplies consumers, smart battery goes to charging mode.
  • As you can see, although animations haven't changed yet, the circuit logic has been updated instantly, so even thought the bottom left and top right Power shutoffs look "on", they're actually "off" already, same for bottom right and top left, which seem "off", but are "on".
• 0.25-1.0 ... potentially more logic propagation might occur, I suspect 4 steps(5 in older versions as they have 5 ticks?).
• 1 sec: animations of power shutoffs update, but the action is already long over.

So the NOT gate is still buggy, but the fact that brownouts and circuit overloading is calculated at whole seconds save me. The only issue might occur if the smart battery is at 0% charge before switching. But then it's your fault for not setting it at 5% instead of 0%.

Just for completeness(and fun) here's what happens when the Smart battery is charged fully:

• 0 sec: it sends 0 and not gate has red on both sides!
  •  
  • 
  • Notice the fridge doesn't care that there is no power - it's still working!
• 0.1 sec: Circuit is updated and action is over in an instant, I hope you didn't miss anything!
  • 
  • The fridge, however, certainly missed the moment without power, as it's still happily working
  • The shutoff on the top left is already disabled, although we can still see it as enabled

Dissection complete!

I looked at your proposal and:

• it seems to be using a glitch with the Power shutoff that is totally unconnected to anything except 1 wire(that probably is/should be also fixed?)?
• really nice wiring! Here's an idea to use even less wiring - a bridge instead of going around the left battery. I might use this in my big-battery setup, I like the singular wire connecting the left battery:
  • 

As a conclusion, I'd say that because of those nasty logic delays and the general complicated nature of logic I'd like to avoid using logic as much as possible. If there's a simple and clean way to do it, you should try to use that instead of a spaghetti bowl of wires and logic circuitry. It uses less materials and is much easier for maintenance.

[FIXBUGFIXBUGFIX]

4 hours ago, martosss said:

I'd say the NOT GATE delay is a bug and has been fixed

Not a bug nor be fixed. The delay between input and output of gate is very realistic, because logic gate need time to refresh.

https://en.wikipedia.org/wiki/Race_condition

4 hours ago, martosss said:

• • However, no overloading/brownout ever occurs, since it is probably calculated at whole second intervals, not at every tick(and this is obviously the first tick of that second). So as long as that circuit receives its 1kJ power juice during that 1 second(4-5 ticks, I think it was 5, but in the latest update they changed it to 4), it shouldn't brownout and everyone is happy.
• 0.1 sec(or 0.25 - the 2nd tick?): Not gate receives 1 and outputs 0.

Power simulation refresh every 200 ms. Logic signal simulation refresh every 33 ms. The change of transformer didn't mean the power simulation refresh time also changed.

4 hours ago, martosss said:

0 sec: it sends 0 and not gate has red on both sides!

•  
• 
• Notice the fridge doesn't care that there is no power - it's still working!

The logic signal and the power detection also have delay.

No power sign and logic cutoff don't match.

It's hard to say, but I guess the delay is between animation and the real-time status.

5 hours ago, martosss said:

it seems to be using a glitch with the Power shutoff that is totally unconnected to anything except 1 wire(that probably is/should be also fixed?)?

That is an exploit and has already been fixed long long ago.

Actually I hope you see this.

or this

 

[martosss]

So the top one produces false and true and switches them fast.

As far as I can tell, the bottom is switching the batteries(probably fast), which would allow batteries to charge and not expose consumers to the rest of the circuit. However, You need to connect each shutoff with all the logic gates, or build logic gates for a circuit that is far away.

So I'm wondering why are they so complicated? Can't you achieve the same fast-switching effect with fewer logic elements?

A smart battery has comparable cost and provides more storing power. + generates less heat than small batteries.

[FIXBUGFIXBUGFIX]

49 minutes ago, martosss said:

So the top one produces false and true and switches them fast.

As far as I can tell, the bottom is switching the batteries(probably fast), which would allow batteries to charge and not expose consumers to the rest of the circuit. However, You need to connect each shutoff with all the logic gates, or build logic gates for a circuit that is far away.

So I'm wondering why are they so complicated? Can't you achieve the same fast-switching effect with fewer logic elements?

A smart battery has comparable cost and provides more storing power. + generates less heat than small batteries.

It's OK to use smart battery instead of a periodic signal source. When I post that thread, there was no smart battery in game, that's why you can't see any smart battery in that thread.

 

I hope you notice this

You use the input and the output of NOT gate as a pair of signals to control shutoffs, but there is delay between your signals.

I use the output of NOT gate and the output of AND gate as a pair of signals to eliminate the delay and avoid Logic Hazard.

If you feel it's too complicated and worthless, please ignore it.

[martosss]

48 minutes ago, R9MX4 said:

It's OK to use smart battery instead of a periodic signal source. When I post that thread, there was no smart battery in game, that's why you can't see any smart battery in that thread.

 

I hope you notice this

You use the input and the output of NOT gate as a pair of signals to control shutoffs, but there is delay between your signals.

I use the output of NOT gate and the output of AND gate as a pair of signals to eliminate the delay and avoid Logic Hazard.

If you feel it's too complicated and worthless, please ignore it.

Ah, I see, so instead of using the input of the NOT directly, you're slowing it down through the AND gate. So you could basically use any gate to fix the delay, as long as they have the same delay(which they probably do). Unfortunately the FIlter/BUffer gates don't have fine time tuning(only 0.1sec or >=1sec), so you can't use those to sync delay of multiple logic gates - you need 1 gate to sync another one. OK, that explains a lot about how logic operates ... It would've been easier to figure logic setups if this was explained somewhere.

And the lack of Smart batteries also explains the use of logic gates and small batteries.

[Albryant]

I'm curious if this change to the power transformers is indicative of future potential buildings that have higher than 2kw draws.

[psusi]

Those are water heaters aren't they?  The only way to cool water is with the hydro climatiser or whatever with water being pumped through it isn't it?  And it getting very hot in the process.