How do transformers actually work?

[Shakie666]

Suppose (for arguments sake) that I had a HV transformer converting 2048eu/t into 512eu/t. What actually happens? Does it just turn 1 packet of 2048eu into 4 packets of 512 eu? Or does it just throw away the other 1536eu per packet?

 

[silenos]

Yes, it converts one higher voltage packet into 4 lower voltage packets, by applying a redstone signal you can reverse its logic so it will take 4 lower voltage packets and bundles it into one higher packet.

 

[Velotican]

More specifically, it has an internal energy buffer (like a BatBox, MFE, etc.) that it stores power in before transmitting it. This allows it to collect a lot of low-charge packets to release as a large packet in situations where power generation is particularly low. Transformers can receive energy up to a specific packet size like a cable but create new packets at the required size.

Generally in vanilla IC2 you will never overload a transformer to the point where its internal buffer capacity actually matters, but GregTech's gigantic superconductor power packets (1 000 000EU/p) are huge enough to make the internal capacity of transformers relevant. Power released into cables with nowhere to go will simply be lost.

 

[DoctorOr]

More specifically, it has an internal energy buffer (like a BatBox, MFE, etc.) that it stores power in before transmitting it. This allows it to collect a lot of low-charge packets to release as a large packet in situations where power generation is particularly low. Transformers can receive energy up to a specific packet size like a cable but create new packets at the required size.

Is there a citation from the author(s) on this? Because that makes no sense and doesn't match what is seen with an eu reader.

If your description were accurate, the output of a transformer would always be exactly a multiple (between 1 and 4) of its packet size, which would make running single machines that use 5eu behind a transformer awful wasteful.

Instead, the "packet size" is just a rating mostly irrelevant to its enery size, except the energy cannot be larger than the declared packet size. Further, a transformers "buffer" is one tick in size, and it receives all the power in one tick and emits up to 4 packets before finishing that tick

Packet sizes are a game restriction, not a simulation of reality. A 2048 packet can contain a single eu.

 

[Velotican]

So apparently I need to explain this thoroughly. Here we go. Much of this is meaningless in standard play anyway because to actually notice this behaviour you have to deliberately flood or overload a power line. If you want to test this, the best way is to connect 8 high-end solar panels to a Mass Fabricator, then insert a transformer and watch the speed of the MF drop accordingly.

As a rule with IC2 power, power isn't drawn from a source until it's needed for something and it can only draw as much as it can find from directly connected sources.

For input to a transformer, as with any other object that requires IC2 power to do anything, the transformer will absorb as many packets as it can find in a single tick until its internal energy buffer is full. If it can't fill its buffer it doesn't matter. The buffer is roughly equivalent to 8 output packets of whatever size the transformer is designed to output. The transformer, like every IC2 block, will try to fill its internal buffer as long as it's not full and is able to find power.

For output, the transformer will transmit as many packets as it needs to from its internal buffer up to its maximum size. It will output every tick so long as something connected to it requires energy, but only as much power as is needed to supply every connected machine, assuming it can satisfy them all. Although it can recharge every tick, it's still limited in each tick to whatever amount of power is stored in the buffer. This is how a poorly positioned transformer can accidentally cripple a power line by throttling the amount of power you can send down it. You're more likely to cripple a power line by using poor quality cable, though.

In practice, provided you either don't flood the cable with excess power (which only GregTech's Fusion Reactor can actually do; the rest of IC2 and its addons are very carefully designed to avoid any power leaks whatsoever) or you don't overextend the limits of a transformer's buffer, you will never notice its throttle effect on your grid, but it's there.

GregTech Fusion Reactors dump 1 000 000 EU/p into attached output cables as soon as they have 1M EU stored internally. Whether you actually get all of that million EU out of your system depends on whether you attach enough transformer buffers to the system or not to process the power as fast as it's being supplied. If you don't do this, it'll get wasted in the cable. The only other power supply that burns excess power the same way is the Nuclear Reactor but because a single HV transformer can cope with its output 99% of the time you never notice!