Power Transfer Issue Can someone verify?

[GRHayes]

Theoretically this module should be about the most effective in producing power per block count or mass.
Supposedly Power Supply modules provide 1200 power at the cost of 1500 power each.
As you can see it is producing 755K e/sec and has 759K capacity.
The PS isn't running out of power or anything. In fact it only drains between 1/3 and 1/2.
Which means I should be able to take some of those off technically.
I ran time tests using this to see how long it would take to charge 50Megs with variable results the best results being between 2:11 and 2:20 min. That only equates to about 357K a second or about 714K per a beam. Where it should produce about 1.2M a beam. Firing it manually does produce the full beam strength and can be sustained switching back and forth between beams. Technically if we went by specks alone it should take 83 seconds to charge the 50M.

Maybe I am missing something. If so please point it out. If you don't mind testing it I would appreciate that as well. The module is available below.

 

[Mork2]

Found some of the issues.

For this module the while the power supply is 755,136 e/sec the power supply beams are only consuming 603,600 energy per second and supplying 482,880 power per second.

Simply two alternating power supply beams staggered like that over 5 seconds provide constant power. Their tick rate is 0.5 seconds so the power usage and transfer per second is double the tick rate. So each transfer beam should consume half the power per second of output per tick so divide the power supply by 600 to get the number of blocks required per power supply beam.


To increase efficiency of this module
You could make this core more efficient by using 1258 power supply beam modules each you would have a constant output of 603,840 e/second using 754,800 e/sec maybe a little less just encase. You can also reduce blocks required by reducing power capacity to around 400k energy total storage as you only need a bit more storage than half a second. Using 1258 power transfer beam blocks per beam should bring the recharge time of 50mil energy in theory back down to 83 seconds.


Back to the bug finding
A 50Million block should take over 103 seconds to charge with the existing block you posted. I timed one and it took 140seconds give or take a couple.

Did a little testing and found one of the problems.

If you use your hotbar to control the power supply beam and manually fire it, it will transfer the full 1.2mil energy into an empty power capacitor block. But if you fire it using logic only 966k energy goes in. Only 4 out of the 5 ticks transfer energy happen when activating using logic. In theory with this penalty though it should take 129 seconds to recharge the 50mil power block. Note sure why its taking even longer than that. It might be because I'm lagging my game somewhat at 45% cpu usage maybe. Except I just noticed Lancake's post and were both getting 2minutes 20 seconds...

Can you confirm that the logic driven power transfer beam is only putting 4 out of 5 ticks of energy in?

And if so do you want to post the bug into the bug tracker or should I?

 

[GRHayes]

Found some of the issues.

For this module the while the power supply is 755,136 e/sec the power supply beams are only consuming 603,600 energy per second and supplying 482,880 power per second.

Simply two alternating power supply beams staggered like that over 5 seconds provide constant power. Their tick rate is 0.5 seconds so the power usage and transfer per second is double the tick rate. So each transfer beam should consume half the power per second of output per tick so divide the power supply by 600 to get the number of blocks required per power supply beam.


To increase efficiency of this module
You could make this core more efficient by using 1258 power supply beam modules each you would have a constant output of 603,840 e/second using 754,800 e/sec maybe a little less just encase. You can also reduce blocks required by reducing power capacity to around 400k energy total storage as you only need a bit more storage than half a second. Using 1258 power transfer beam blocks per beam should bring the recharge time of 50mil energy in theory back down to 83 seconds.


Back to the bug finding
A 50Million block should take over 103 seconds to charge with the existing block you posted. I timed one and it took 140seconds give or take a couple.

Did a little testing and found one of the problems.

If you use your hotbar to control the power supply beam and manually fire it, it will transfer the full 1.2mil energy into an empty power capacitor block. But if you fire it using logic only 966k energy goes in. Only 4 out of the 5 ticks transfer energy happen when activating using logic. In theory with this penalty though it should take 129 seconds to recharge the 50mil power block. Note sure why its taking even longer than that. It might be because I'm lagging my game somewhat at 45% cpu usage maybe.

Can you confirm that the logic driven power transfer beam is only putting 4 out of 5 ticks of energy in?

And if so do you want to post the bug into the bug tracker or should I?

Correct under the 2.5 second cubes need to be increased form 1006 to 1258 or so. I may have to so do something to ensure the last tick is captured. which also seems to be an issue. Another activator may do the trick it is shorter than .5 seconds. from what I can see. That or the dev team may find a solution with the bug report being filed.
The two issues combined seem to add up to the missed power being transferred.

 

[lilheartless]

Correct under the 2.5 second cubes need to be increased form 1006 to 1258 or so. I may have to so do something to ensure the last tick is captured. which also seems to be an issue. Another activator may do the trick it is shorter than .5 seconds. from what I can see. That or the dev team may find a solution with the bug report being filed.
The two issues combined seem to add up to the missed power being transferred.

you could try adding cannon slave to the system that might negate some of the negative effects although if its literally just the last tick that is "not fireing" the cannon might not do anything.

 

[GRHayes]

you could try adding cannon slave to the system that might negate some of the negative effects although if its literally just the last tick that is "not fireing" the cannon might not do anything.

It never hurts to try though.
I probably will tomorrow. Just to see if it solves the issue without waiting for a bug to be fixed.
As long as I can keep the block count the same or close then its worth putting the cannons on.

 

[lilheartless]

It never hurts to try though.
I probably will tomorrow. Just to see if it solves the issue without waiting for a bug to be fixed.
As long as I can keep the block count the same or close then its worth putting the cannons on.

I'm assuming right now you are running alternating beams right? If so the cannons should be identical lol ;p

 

[GRHayes]

I'm assuming right now you are running alternating beams right? If so the cannons should be identical lol ;p

Yes, Just I made a miscalculation. I based it on 2 seconds vs 2.5 seconds. So my modules were off 20% but that didn't account for the full amount not being charged. Either way currently worth it. far better than 25 a cube. Its in the picture above as well and in the download here.

 

[lilheartless]

Yes, Just I made a miscalculation. I based it on 2 seconds vs 2.5 seconds. So my modules were off 20% but that didn't account for the full amount not being charged. Either way currently worth it. far better than 25 a cube. Its in the picture above as well and in the download here.

you've probably considered this but with the mis design the second beam fires early causeing your system to temporarily go into power drain mode or at least it should if both beams fired at any one time maybe that is the issue.

 

[GRHayes]

you've probably considered this but with the mis design the second beam fires early causeing your system to temporarily go into power drain mode or at least it should if both beams fired at any one time maybe that is the issue.

Yep considered but it didn't pull power down even 25% that is because it only takes away 20% every tick. that tick is every 0.5 seconds.

 

[Mork2]

With the current bug not putting the 1 out of 5 ticks on energy on target if it doesn't consume the extra power wouldn't it be possible to upsize the power transfer beam by 25% to make up for it? Or does it consume all 5 ticks power anyway?

 

[Mork2]

Checked it is also only using 4 ticks worth of energy.