No the 128 Packet will be split up in 4 32 EU Packet. So if the Transformer receives 128 Eu it will spit them up to 4 EU packets of 32 EU.
no power will be wasted, the transformer will only request a packet the size that the machines it supplies require
So even if the LV transformer is hoked up to a MFE unit, which outputs 128 EU per tick, and to a macerator wich consumes 2 EU per tick, there wouldn't be any energy wasted?
Edit: I just did a few test in creative mode and found the answer to my question. Thanks for your help nevertheless! 
Please remember to put the answer to your question in the original post, and put "[Answered]" in front of the title, so that other people with the same question can find the answer more quickly. Thanks!
i have a similar questions concerning transformers.
i have a cesu feeding into a lv-transformer(step-down), followed by a batbox.
there is also a centrifuge between the cesu and the transformer.
when the centrifuge is running full steam, i imagine it takes either every other packet(pulse? not sure what the unit is now), or half the currently transmitted one.
but when it is warming up, it only takes a very small amount of power.
does the machine only draw power when there is a need for a full packet?
what does a transformer do with a packets less than it's input (ie, a 5eu packet from a wind turbine, or a packet transmitted by the cesu, but drawn from by the centrifuge)? does it buffer until it is a full transmittable packet?
Energy storage units buffer until they have a full packet (unless you charge a tool in the GUI, then they don't, because there is no packet size there). So I would assume transformers do the same with their internal buffer.
A machine asks for power whenever there is room in its internal buffer. Then it gets a burst of energy that will fill the internal buffer over the threshold. Then it consumes that energy until it falls under the threshold again (no energy is lost), at which it begins asking for energy again.
This is why a Forestry electric engine can work just fine on a batbox sending 32 EU/t and a MFSU sending 2048 EU/t. Its internal buffer size does not matter; it only needs a very small one. The big 2048 pulse will completely overshoot it, but then the engine will just not ask for more until it has consumed all the excess energy.
Looking at timestamps, I'm guessing this applies to IC2 experimental, so I have a question based on this diagram:
__B__
__+__
B+O+B
__+__
__B__
The O is a copper cable from an unspecified source carrying it's full load of 128. The + is a tin cable. The Bs are input sides of 4 batboxes.
By splitting my 128 four ways, am I getting A, B, or C?
A: 4 round-robbin- style time-delayed packets of 128 causing each batbox to explode
B: 4 packets of 32 that fit just right into the tin cables
C: Something else
Should be B. You can easily test this because if it is anything other than B, one or more batboxes will blow up.
However, that is true only as long as all four batboxes are requesting power. The moment one of them stops (because it is full), you no longer split by 4, but rather by 3. This will kill the others.
Display MoreLooking at timestamps, I'm guessing this applies to IC2 experimental, so I have a question based on this diagram:
__B__
__+__
B+O+B
__+__
__B__
The O is a copper cable from an unspecified source carrying it's full load of 128. The + is a tin cable. The Bs are input sides of 4 batboxes.
By splitting my 128 four ways, am I getting A, B, or C?
A: 4 round-robbin- style time-delayed packets of 128 causing each batbox to explode
B: 4 packets of 32 that fit just right into the tin cables
C: Something else
Should be B. You can easily test this because if it is anything other than B, one or more batboxes will blow up.
However, that is true only as long as all four batboxes are requesting power. The moment one of them stops (because it is full), you no longer split by 4, but rather by 3. This will kill the others.
Nope, it is A, also not only the batboxes, the tin wires would (if they werent broken) melt aswell.
IC² exp is roughly the same as pre-IC² exp, the difference is that anything you used to hear as "EU/p" is EU/t, which means a batbox and a solar panel can blow another batbox (33 EU/t exceeds the limit of 32 EU/t).
]Nope, it is A, also not only the batboxes, the tin wires would (if they werent broken) melt aswell.
*sigh* 
Protip: I tend to be able to back up my claims because I test before posting. 
*sigh*
Protip: I tend to be able to back up my claims because I test before posting.
Oh, forgot this : http://ic2api.player.to:8080/job/IC2_experimental/303/
"disable machine explosions until the energy net is capable of simulating the voltage properly"
However, the behavior mentioned in A is valid for GregTech machines, which WILL explode violently upon surpassing the maximum EU/t
I'm aware of the change, but my server is running build 298 and so is the client I am in this test world with. 
I urge you to try it out for yourself. EU is split evenly between all available destinations, so each tin cable and batbox only receives 32 so long as there are 4 of them accepting. That is how all multi-output configurations work and have always worked in IC2, including generators and transformers.
I'm aware of the change, but my server is running build 298 and so is the client I am in this test world with.
I urge you to try it out for yourself. EU is split evenly between all available destinations, so each tin cable and batbox only receives 32 so long as there are 4 of them accepting. That is how all multi-output configurations work and have always worked in IC2, including generators and transformers.
Even before IC² exp? eh... but that is why we always take in count that it will receive the maximum EU/t, as if one stop requesting, boom.
Yes, even before IC2 Experimental. It is especially easy to confirm in those versions, because the EU reader still works there. Just load any old 1.4.7 modpack from the launcher of your choice.
Granted, back then is was largely irrelevant/unnoticed because you really did not have to pay attention as much as you do now. Still, the effect was used in a few niche cases. CASUC reactors in 1.2.5, for example, generally required multiple HV transformers stuck onto them in order to move all the power they generated. And the reason why this worked is because the reactor, being a generator with multiple output faces, split its power output evenly between all faces that had something connected. Thus even a 4k EU/t reactor did not blow up its pair of 2k-limited HV transformers.
