Battery switch automation stopped working

[Denisetwin]

Battery switch automation ( @Saturnus design ) stopped working after working just fine for dozens of cycles.  I saved and reloaded which usually fixes weird automation issues for me, but no go.

Can someone take a peek and see if they can see what is wrong?

[FiannaTiger]

Was this the switching that gave you infinite free power?

[Saturnus]

59 minutes ago, Denisetwin said:

Battery switch automation ( @Saturnus design ) stopped working after working just fine for dozens of cycles.

I have retracted the design as it doesn't do what it's supposed to do. Namely to automatically unstuck itself.

Apologize if you weren't aware of that.

[Zarquan]

1 hour ago, FiannaTiger said:

Was this the switching that gave you infinite free power?

No, this is the switching that lets you avoid using heavi watt wires by having no consumers on the grid, as batteries aren't considered consumers.  Instead, it switches the battery from on the larger power grid to charge and the smaller grid with the consumers such that the smaller grid is never in contact with the larger grid.  It avoids overloads.

[FiannaTiger]

On 4/3/2019 at 4:32 PM, Ipsquiggle said:

Dead batteries will no longer provide magical power to transformers (the "Battery Transformer Infinite Power Loop" bug)

It is possible this fix may have stepped on that as well as fixing the free power.  Not sure.

 

[Saturnus]

15 minutes ago, FiannaTiger said:

It is possible this fix may have stepped on that as well as fixing the free power.  Not sure.

No. Battery switching still works fine. It just that set up does not auto-unstuck itself as it should.

[Nitroturtle]

This doesn't look right to me.  Are both batteries supposed to be connected to the same automation wire?  I know getting stuck is a bug, but this doesn't look stuck to me, it just looked like it's implemented incorrectly.

 

[Mullematsch]

This happens occasionally. Might be connected to saving / loading the world during the design switching, can't say for sure.

Haven't found a setup which does not mess up at all. This is what I am currently using and every so often the power switches get stuck:

I usually have battery one with my normal settings (95/85) and battery two with "reset" settings (0/0).

If the power switches get stuck, just copy the batter two settings to battery one to reset it and then go back to normal. 

[Gus Smedstad]

5 hours ago, Saturnus said:

No. Battery switching still works fine. It just that set up does not auto-unstuck itself as it should.

My experience with the battery-switching transformers is that sometimes the shutoff switches stop responding to automation. The automation logic is correct, it's the shutoffs that behave incorrectly. Destroying and rebuilding the shutoff doesn't work, either. The only thing that has unstuck it is destroying the cable and rebuilding it (it's a separate entity from the shutoff).

 

[Saturnus]

7 hours ago, Gus Smedstad said:

My experience with the battery-switching transformers is that sometimes the shutoff switches stop responding to automation. The automation logic is correct, it's the shutoffs that behave incorrectly. Destroying and rebuilding the shutoff doesn't work, either. The only thing that has unstuck it is destroying the cable and rebuilding it (it's a separate entity from the shutoff).

 

I don't think you get the point here. The purpose of that design was for the automation to get itself to get itself unstuck without having to deconstruct anything or copy settings back and forth but it's not reliable.

You can use my older manual unstucking version. If it gets stuck just flick the switch and it'll run again.

How it works is that one battery becomes the dominant battery but if something happens and a shut off gets stuck then it'll drain that battery. Regardless of which battery is is drained, one of the batteries will always output a positive signal when stuck. Going through a NOT gate that means the control line is always negative when it is stuck. So all we have to do is flick the switch to switch the control line to positive briefly, and the set up is unstuck.

Note, always set both batteries at 75/25 in a bi-directional battery switching set up like this.

 

[Saturnus]

If you're plagued by the shut off getting stuck a lot and it gets annoying to flick the switch yourself then there is a way to detect if there's no power on a network with a shut off valve loop and an element sensor set to a buffer.

As long as there's power one blob of liquid keeps circling around and resetting the buffer but if at any point the circuit loses power then the shut off valve shuts down and the loop stops which triggers a negative edge detector to send a positive pulse to reset the battery switching automation by itself replacing the manual switch.

[Gus Smedstad]

4 hours ago, Saturnus said:

You can use my older manual unstucking version. If it gets stuck just flick the switch and it'll run again.

 

My experience was that I couldn't get the power shutoff to start responding to automation again, period. If I'm reading you correctly, you found that if you forced the signal to the shutoff switch off and then on again, it got "unstuck." I couldn't get that to work. Once stuck, the shutoff ignored automation from then on no matter what I signal I sent to the shutoff. Only deconstructing the cable and reconstructing it worked.

Just FYI, I've seen shutoffs get stuck in either state. Stuck turned on when they should be off, and stuck turned off when they should be on. The former's problematic because it created a direct line between the generators and the consumers, since both shutoffs were on even though the automation logic was supposed to make that impossible.