Redstone energy cell and conduit behavior?

[Deadgye]

So I've been using Redstone energy cells and conduits to transfer power in my base on my DireWolf20 1.5.2 server. However, their behavior seems to be weird to me. I have a set up that is basically like this:

          |---< Cell <---|  |<--- Cell <--- Engines
Output <---|              |---|
          |---< Cell <---|  |<--- Cell <--- Engines

My intention was that the output would evenly split its power draw to the two cells on the left, and those cells would evenly draw power from the two cells on the right. This apparently isn't how it works. It seemed like all the power draw from the output was being provided by one cell on the left while the other cell on the left did nothing. And then that cell on the left that was providing power only drew power from the top cell on the right. When the top cell on the right ran out of energy the bottom cell on the right started providing energy.

It didn't work how I wanted, but I was able to deal and set up my engines to start when the bottom right cell ran out of power and stop when the bottom cell became full. This worked and energy would drain from the top right cell until it emptied, then energy would drain from the bottom right cell until it emptied. My engines would then start filling both right cells, energy would start being drained from the top right again and not the bottom right, and the engines would stop when the bottom right filled up.

It was weird, but as long as the weirdness was consistent I was able to plan around it. However today after I added some more machines to the output, or perhaps after a server reset, energy is now drawing from the bottom right cell first, completely ruining the logic and efficiency of my setup.

Can anybody explain what is happening? Why is the output bypassing the buffer energy cells and drawing energy directly from the right cells instead of drawing energy from the left cells and the left cells drawing energy from the right cells? And how is this decision that chooses which cell on the right is used made? Nothing I can think of makes any sense.

 

[Samueldmm]

I'm not 100% on how you've go the power running there. If you provide a screenshot I'll be able to explain a bit better. What i've found with energy cells is that it's better to run them in a line like so. This way it draws from both of them and you can dictate how much you want to be drawn from either one. I have it so that all the power drains from the left most one quicker. Then you could have it so that when the left one runs out the engines start again. (If you have more to the right each one staggered to have less out a tick you create a buffer)

 

[Seraph089]

Why is the output bypassing the buffer energy cells and drawing energy directly from the right cells instead of drawing energy from the left cells and the left cells drawing energy from the right cells? And how is this decision that chooses which cell on the right is used made?

Most likely, the output is drawing from the buffer cells as you want but they're being recharged so fast you don't see it. RECs in series transfer charge instantly if input is greater than output, if you throttled the output speed on the right cells to less than what's being used you would be able to see the left cells being drained first and the right cells charging them.

I can't say for certain why the cells are being chosen at random between the top or bottom set, but it could be the way the conduits are run. Possible fixes depend on your power needs though

 

[Deadgye]

Indeed when I throttle the input low enough on the left cells I was able to see the power draw. The multimeter apparently shows the energy requested as "0.0" when the input to the cell is greater than the output. That clears that part up at least.

Now if I can only figure out why one of the left cells and one of the right cells aren't used until the other same-sided cell is empty.

I'm not 100% on how you've go the power running there. If you provide a screenshot I'll be able to explain a bit better. What i've found with energy cells is that it's better to run them in a line like so. This way it draws from both of them and you can dictate how much you want to be drawn from either one. I have it so that all the power drains from the left most one quicker. Then you could have it so that when the left one runs out the engines start again. (If you have more to the right each one staggered to have less out a tick you create a buffer)

Those cells are connected in parallel to the line (although set on input?), which should be how my engines are configured as well.

 

[cynric]

Sadly, energy conduits don't balance the power draw if they have multiple suppliers so having multiple energy cells in parallel doesn't really work as expected. And the order seems to change randomly, even by just restarting the server.

Only workarounds I've found are to just not rely on energy storage if your network requires more throughput than what one cell can deliver.

 

[Seraph089]

Seeing a screenshot definitely helps. Since it looks like each cell has it's own bank of engines, it may be a good idea to set up a gate on each cell to start it's bank of engines. That way it doesn't matter as much which one is taking more of the load. It would also be much easier to just use one set of RECs (top set or bottom set) and remove the other, unless you need more than the 50 mj/t one set of cells can provide.

 

[Samueldmm]

Those cells are connected in parallel to the line (although set on input?), which should be how my engines are configured as well.

Yeah, I was storing energy.

Only way I can see around this is to get rid of the connecting part in the middle. So it's like this: Clicky


It's pulling energy steadily out of the right most two first. Then when they deplete it moves on. it seems to be pulling from both the top and bottom equally.

 

[Deadgye]

I originally had a gate on each cell for each engine bank, but since power is only drawn from one at a time doing so will result in only one line of engines running every now and then and the other line never running at all. The reason I have two is because the engines all-together create more than 100MJ/t which results in wasted power if I only have one.

If the order didn't randomly switch for seemingly no reason my setup would have worked. I can 'fix' my setup by moving the gates to the top energy cell right now.. but then if it decides to switch again I'd have to move the gates back down. It's looking like the only solution is connecting both with gates and then configuring some extra dynamic logic with the rednet controller.

Yeah, I was storing energy.

Only way I can see around this is to get rid of the connecting part in the middle. So it's like this: Clicky

It's pulling energy steadily out of the right most two first. Then when they deplete it moves on. it seems to be pulling from both the top and bottom equally.

When I do that with my setup nothing changes. :/ It just continues to exhibit the exact same behavior as before.

From what you've described, I'm beginning to make the conclusion that when cells are in parallel they will only both be active if the draw is higher than what one cell by itself it allowed to output. In order for me to set up my cells for that I'd have to test and change the output limits every time I add something to my system, which is highly undesirable.

 

[Samueldmm]

When I do that with my setup nothing changes. :/ It just continues to exhibit the exact same behavior as before.

From what you've described, I'm beginning to make the conclusion that when cells are in parallel they will only both be active if the draw is higher than what one cell by itself it allowed to output. In order for me to set up my cells for that I'd have to test and change the output limits every time I add something to my system, which is highly undesirable.

Would it not just be easier to have it all input into one, then that output to multiple and then when that one runs out you've got the engines going?

 

[Deadgye]

That'd work if I only had a handful of engines that only created 100MJ/t or less. But since I generate more than 100MJ/t energy would just be wasted pretty badly.

I think I've found a solution for the mean time. I've added the gates to the top cell as well but I've configured my logic differently. Instead of each cell being an RS latch that triggers its respective bank of engines I've configured each cell as an RS latch that outputs to a variable and then AND the variables so that both banks of engines trigger when both cells reach empty and stop running when any one of the two cells reach full.

I was hoping to find a solution within the redstone energy cells/conduits themselves or to make sense of them, but with a behavior that I can only describe as buggy it looks like it had to be handled this way.

 

[Samueldmm]

I'm confused. Why'd you make so many engines if you didn't need all the power? if you don't want to waste power then build more cells and have a bank of them for input and storage, or reduce the amount of engines. I was assuming that your output was a constant. I see multiple different ways to go around making it so that the engines only turned on if you needed the EXTRA power to recharge your storage. You can use the energy empty on a buildcraft gate. The energy empty being on just one specific cell put you have input into multiple.

 

[Deadgye]

I don't use all the power during idle time, but when my system is under full load with multiple machines running and my quarry running I burn through energy pretty fast. The plan was to make the system so that I can provide enough throughput during peak, but not have it waste power when I don't need it. Making a energy system for an entire base with a constant output sounds not only silly, but impossible to me. It should always be dynamic, less you have to manually change settings every time you want to use your machines, or if you add/remove things to your infrastructure.

I have always had the system set up so that the engines only turn on when I need the extra power to recharge my storage. When the cells randomly changed their behavior though I was confused and wanted to see if there was a reason why it happened, but apparently there isn't and it's just buggy or unpredictable behavior.

Also I can confirm that my solution mentioned in my previous post worked. The system is now behaving the same way it did before. And if the cell order switches for no reason after a server restart it should still continue to work.

 

[Samueldmm]

I constantly have output because I'm always running multiple quarries and the sorting to have ores pulverised and smelted. I have a rough output of about 140mj/t but I generate way more than that, roughly 208mj/t. The rest goes into redcells MFSUs and a Massfab. That way I can create a massive store of energy and keep the simplicity of a constant energy input.