EU/t is not as expected, what don't I get?

[Lawlcat]

So I've got this setup in my machine shop:

HV power comes in on the left into the MV transformer, which then goes into the LV transformer, then copper cables come out. If my understanding is correct, the HV (512) should be converted to MV (128) and then to LV (32) so only 32 is put out the copper cables...

but when I hook a batbox up to the copper cables and run my EU-reader on the wires I get an average of 120 EU/t going INTO the batbox.

Thankfully the batbox then properly puts out 32 so my machines don't explode, but I was expecting the entire copper piping to be 32. What gives? Also, the extractor there is running off the 120 eu/T copper wire and doesn't seem to explode.

Here is my power generation system

The 4 geothermals go into an MFE which goes into a MV transformer pushing out to HV... this is so I can run it a long distance without losing too much.

So, is my understanding of how the transformers work incorrect? Why are my copper cables getting 120eu/T on the output side of an LV transformer? Why is my Extractor not exploding?

[Lothos]

i don't think the EU meter is accurate myself.

[Industrial Miner]

As the way I think that transformers work is that each transformer can only transform a packet every tick. So try adding more transformers, hoping that the packets will get transformed faster?

[Lothos]

As the way I think that transformers work is that each transformer can only transform a packet every tick. So try adding more transformers, hoping that the packets will get transformed faster?

does that mean 1 higher voltage packet input and 4 lower output per tick? or does it throw out the other 3?

[charcharmunro]

The copper cable probably has some power loss, there, each packet has individual power loss, so each packet is losing some EU.

[Lothos]

The copper cable probably has some power loss, there, each packet has individual power loss, so each packet is losing some EU.

that would correspond to the mv/lv transform of 4 32EU packets at a time with loss for 120 EU on the measurement.

[Lawlcat]

So it's sending multiple packets over the time of a tick? I guess I don't understand the packet/tick system. I figured it was just 32/t out means it can supply 32 per tick, so I cannot have more machines than would draw 32 a tick. What's the packets?

[Lothos]

well, forgive me if I am wrong, but my understanding is a cable can handle multiple packets of the designated size at any given period. So, in this case the EU transmitted is 32EU in packet size, but its not restricted to 32 per tick.

[nedned2k]

Ok, people seem to be confused, so I'll try to explain IC2 power.

Everything that outputs energy, outputs it in what will be referred to as packets. A packet is sent out every tick (~20 ticks/sec) with a certain value of electricity, or EU. Let's say we have this scenario above, where a 512-eu packet is sent out on HV cable every tick. It goes into the MV transformer, and splits each 512-eu packet into four 128-eu packets. At this point, there's still 512eu/t, but it's not all one packet, but four 128-eu packets. Then, the LV transformer splits it further into 16 32-eu packets per tick. So, in a perfect system, you should have 512et/t safely going across the copper cable into the batbox in the form of 16 32-eu packets per tick.

However, copper cable loses one eu for every three or four (to lazy to confirm on IC2 wiki) blocks of cable, and each of the 16 packets are affected, causing a total loss of 16 eu per three or four blocks.

So basically, you have this system set up perfectly, and it's working as it should, except that you are actually losing 3/4 of your power. I would suggest placing the transformers closer to the batbox (so that only one packet is affected by loss) or using Glass Fibre Cable, which is very expensive but has a loss of one eu per ~20 or so blocks.

Edit: for the record, the person above me posted while I was typing this. Move along.

[Industrial Miner]

nedned, that is basically what I've thought too, but then after reading the IC2 wiki topic about transformers, it really said each transformer can only upscale and downscale one packet per tick. I've now tried some stuff in a testing world about the transformers.

My setup was one Hybrid solar panel outputting 1280 EU/t (IC2 addon and generation and output rates are configurable) - HV transformer - some glass fibre cable - Massfab. The massfab gets the full 1280 EU/t. Then I added in a MV transformer in the line and that dropped down the EU flowing through the wire to 512 EU/t. And then the LV transformer came into the line, which dropped it to 128 EU/t. What surprised me was that EU was not getting lost. The EU in the hybrid solar panel was building up it's internal charge.

So what I mean is, if you really want many small packets of 32 EU/t, you've gotta use more transformers. Otherwise upgrade the machines with transformer upgrades and try to stay in the MV/HV/EV or even >EV range if you can.

[Lothos]

how are you measuring the EU/t though? Is it smart enough to measure it in such a manner of 16 32EU/t packets?

[Industrial Miner]

The EU/t flowing through a cable is all counted up together. So if 16 32EU/t packets would flow through it, you would read 512EU/t on the EUreader.

[nedned2k]

Industrial Miner, I suppose I was a little wrong there. I've looked at the wiki some more, and have come up with an alternate statement.

From what I understand, the transformers still do split the packets, but not as I thought. The MV transformer takes each packet per tick, splitting it in four and sending it on it's way, but then the LV transformer is bombarded with packets, (4/t) and is only able to do one a tick, and thus outputs four 32eu-packets, which then loses energy. The rest of the power "backs up", I suppose. I don't know where it goes, I'll have to do more testing for that, but I'm sure it doesn't vanish into thin air.

It makes a lot more sense that you are losing 8 eu/t from cable loss than 3/4 of the entire power.

[Lothos]

i wonder if you're losing power because it outputs to all the faces at all times regardless of connections :/

[Lawlcat]

Nedned, but that comes back around to my question. If the power in the LV transformer "backs up" and only emits 1 32eu packet a ticket, why is my wiring downstream from the LV transformer reading at 120eu/t draw?

[BLAZE MkIV]

The batbox can accept more than 32eu/t it can only accept 32eu/packet. So its receiving multiple packets per tick totalling 120eu/t.

[nedned2k]

Nedned, but that comes back around to my question. If the power in the LV transformer "backs up" and only emits 1 32eu packet a ticket, why is my wiring downstream from the LV transformer reading at 120eu/t draw?

Sorry if I was unclear.

It's transforming one packet a tick, and four 32-eu packets are going out, totalling 128 eu/t, and then, the copper wire is making each of the four packets lose eu, thus you getting 120eu/t.

[lordlundar]

Something that I have noticed is that there is a lot of HV cabling running around and it looks like most, if not all of it is for areas not hitting above 512 EU/t packets (running into an MV transformer). You might want to change those up for glass fiber cables if you can afford it as HV cables have an insane amount of EU bleed.

[aMUSiC_gr]

One common misconception is that power generators and transformers and power storages "Send" energy out. That's not the case. The way this works is that power goes through the cables if there is something that needs it. i.e. if you put a power source and a transformer and nothing else, since the transformer is not consuming or storing energy, nothing will go through the wires. As soon as you stick a batbox or a macerator after the transformer then the macerator/batbox will "pull" energy from the transformer at the maximum packets per tick that the machine is designed to accept, then the transformer in turn will pull energy from the powersource at the rate it can handle it (1 packet per tick). So no, you do not lose power this way, but you definitely do not make the optimal use of the power output.

I'll have to agree with the guys above though, the best approach is to invest into glass fiber cables and minimize power loss (remember the higher your packet size is, the less the loss you suffer), use MSFU's to store all that excess power (and also use them as "repeaters" to minimize the power loss from the cables) and use transformer upgrades on your machine. A machine with 2 transformer upgrades, 13 overclocker upgrades and 64 power upgrades is a beast. You can work full batches in just 3 seconds :)

[TheTaleteller]

IC2 electricity is pretty tricky. Each of your (6?) geothermals send packages of 20EU every tick towards the transformer. Each tick 120EU come in but the "voltage" still is 20EU because its the max packet size.

The transformer (lv->mv) accumulates 128EU and sends out 1 packet that "voltage" about every 6 ticks in this example. same game goes for mv->lv. Transforming down it gets a package of 128 and sends 4 packages of 4 32EU in a single tick.

An often mistake is to hook up a single BatBox at the end of such a line. It will get charged by 128EU each tick but can only discharge at 32EU in one tick. Therefore the BatBox will run full, jamming the energy output at the generators, which is pretty bad for nuclear reactors because those just tick down wasting uranium. You can fix this by paralleling energy storage, i.e. 6 GeoGen feed 4 BatBox.

Another confusion is energy loss. The energy loss is calculated for each energy source individually at the FULL cable length. Say you get a MV-Solarpanel and a LV-Solarpanel hooked up at a gold cable feeding a MFE at 8 blocks and something else at 16 blocks, the Energy loss for the MV-Panel wil be 8EU even at the MFE and the loss for the LV-Panel will be the full 8EU it supplies. Counter it by using short cables chaining storage devices.

[Industrial Miner]

So I am still right about the amount of transformers! :D

[Nolz]

Nevermind

[Industrial Miner]

Something that got edited.

Instead of batboxes, use Transformers instead. After every transformer, the packet of 32 EU/t gets downgraded again to LV, again to 32EU/t. And they're cheaper and easier to make too.

EDIT: Yes I did read that.

[TheTaleteller]

Take an experiment. GeoGens have a known output of 20k EU per cell at 20EU/t. Take 6 of them in a row and hook them up on a single transformer chain LV->MV->HV->HV->MV->LV feeding an empty MFE. Beside that take a single GeoGen feeding a MFE directly. Prime the generators while the cables are disconnected and close both circuits. This setup with transformers will charge up the MFE up to 120k EU in exactly the same time the single generator charges its MFE directly.

Now alter the construction and give every generator its own up chain LV-MV-HV feeding together a single down chain of HV-MV-LV. The MFE will end up with about 120k EU again, but it will take a little longer. Some generators will stop producing energy from time to time because the down chain can't take all the packages at this rate. Same yield lower throughput. Therefore use the same amount of transformers on each end, but unless you have a lot of sources a single chain will do, and having MFS units saves you stepping down below 512EU for the incoming feed anyway.

[aceclown]

^^ The problem is the thick HV cables, that image shows how long you can have a cable before it loses EU.

Swap the thick HV cable out for glass fibre and you should see an improvement straight away.