Little confused about LV/MV/HV

[Tyrindor]

I'm not sure I understand it all. I threw in 10 overclocked into a couple of my machines that were connected up to a fully charged MFE. However, they struggled to maintain full power until I went down to 5 overclockers. I searched around to figure out why this is, and basically it's because MFE can only provide a certain amount of power to my understanding.

So I connected up a MFSU, but then I had to convert it to MV then LV because other machines on the line require it. It's still happening and I believe that is because when you convert to MV/LV, you basically apply the same low voltage?

So then I built a couple transformer upgrades, threw them in so they could at least accept 128EU/T, removed the LV transformer and gave them pure MV power. However, same thing, they struggle to keep powered up even with a fully charged MFSU providing MV power. I have machines on the line that simply don't accept transformer upgrades and can't accept HV power, so throwing in a few more transformer upgrades and removing the MV transformer isn't really an option.

Can someone please explain to me whats going on and why I can't throw more than ~5 or so overclocker in my machines?

 

[ILoveGregTech]

When you put in overclockers the power usage increases exponentially.
I don't know the exact numbers but 10 OCs is hundreds of EU/t

 

[slay_mithos]

I think what you need is the energy storage upgrade you disregarded.

Basically, with too many overclockers, they eat through all the internal storage at once, and need to wait fo the machine to fill enough to run again for an other tick.

No matter how much you feed it, it will remain limited to the internal energy storage.

I would say to put transformers (1 or 2, depending on MV or HV), 1 slot worth of energy storage, and fill the rest with overclockers.

 

[Bigglesworth]

Also, transformers arnt required. Just make more LV behind the machine.

2 LV = 2x the 'bandwidth' of incoming EUt.

 

[Chocorate]

EU/Voltage would actually be a pretty cool numbers system, if it weren't for machines exploding.

 

[Siro]

It's not the packet size that's the problem, it's the throughput. You can get more throughput by having two mv transformers connected to 8 lv transformers to give you 32 lv packets per tick instead of only 4 (which is why it keeps running out of power). If you have a lot of power generation, you might want to install a bypass of your storage so you can flood a larger array of transformers with packets (or skip the transformers entirely if you're using lv power generation). Just try to avoid creating a loop in your cabling.

 

[Bigglesworth]

EU/Voltage would actually be a pretty cool numbers system, if it weren't for machines exploding.

EUp = volts
EUt = amps
EUCableLength = Ohms
Cobble = Bananas

Simple

 

[Neirin]

As an example: a recycler with 10 overclockers requires 109 EU/t. However, the recycler can only store 45 EU. This pretty much breaks the machine.

For the record: a recycler can handle 8 overclockers before needing an energy upgrade. An electric furnace can handle 10 overclockers, and a macerator/compressor/extractor can handle 12. However at those levels the recycler will pull 42 EU/t, the furnace will pull 329 EU/t, and the rest will pull 562 EU/t. With 10 overclockers (your situation) a furnace will pull 329 EU/t and macerator/compressor/extractor will pull 219 EU/t

It's also worth mentioning that LV/MV/HV only apply to the packet size of EU being sent, not the total amount of EU being sent. So if you feed the output of an MFE (128 EU/t) through a LV transformer you end up with 4 packets of 32 EU each. It's still the same total energy, but more spread out. Keep in mind, though, that distance-based EU loss is applied to every packet, so if you ran the power through enough cable to lose 1 EU per packet the MFE would effectively deliver 127 EU/t whereas the transformed output would give 124 EU/t. It's not a big deal if you're using glass fiber everywhere, but it can make a huge difference with some of the more loss-y cables.

Also, transformers only put out 4 lower-voltage packets a tick, so you'll only ever get 128 EU/t out of a LV transformer, even if you feed it 512 EU/t through a MV transformer.

EDIT: added the overclocker numbers as an attachment before refreshing the page to see if it had already been posted

 

[Bigglesworth]

It's not the packet size that's the problem, it's the throughput. You can get more throughput by having two mv transformers connected to 8 lv transformers to give you 32 lv packets per tick instead of only 4 (which is why it keeps running out of power). If you have a lot of power generation, you might want to install a bypass of your storage so you can flood a larger array of transformers with packets (or skip the transformers entirely if you're using lv power generation). Just try to avoid creating a loop in your cabling.

Lag issues aside; Yeah I liked having an array of LV, that let me ditch transformer upgrades entirely. Then issues of internal storeagle being able to keep up ect become a problem. But thats a oversight in the mod itself (internal storage should be at least enough to cover the current EU/t requirement of an overclocked machine. So mostly youre stuck with 12 overclockers and a storage upgrade in the last slot.... but I havent done this in months so maybe it changed?

 

[Tyrindor]

Thanks guys, I figured it out. It wasn't the internal storage, I put 4 into each and still had the issue.

I put transformer upgrades in them, then removed the LV transformer so it was just 2x MFSU > MV > Machines. This seems to supply 1024EU/t instead of 256EU/t according to a EU meter. Those LV transformers really hurt!

 

[Berserkenstein]

As an example: a recycler with 10 overclockers requires 109 EU/t. However, the recycler can only store 45 EU. This pretty much breaks the machine.

For the record: a recycler can handle 8 overclockers before needing an energy upgrade. An electric furnace can handle 10 overclockers, and a macerator/compressor/extractor can handle 12. However at those levels the recycler will pull 42 EU/t, the furnace will pull 329 EU/t, and the rest will pull 562 EU/t. With 10 overclockers (your situation) a furnace will pull 329 EU/t and macerator/compressor/extractor will pull 219 EU/t

If you use transformer upgrades in the machines and feed it a large enough packet, energy upgrades are not needed. I run 10 overclockers with 2 transformer upgrades in my recyclers and feed them HV.

 

[Bigglesworth]

If you use transformer upgrades in the machines and feed it a large enough packet, energy upgrades are not needed. I run 10 overclockers with 2 transformer upgrades in my recyclers and feed them HV.

Perhaps the changed the internal storage on recyclers, or how that works? What version are you using? For me if I used to try and use that setup, it wouldn't work because the internal storage wasn't enough to fill a 1 tick storage requirement even if incoming packets were big enough.

 

[Berserkenstein]

Perhaps the changed the internal storage on recyclers, or how that works? What version are you using? For me if I used to try and use that setup, it wouldn't work because the internal storage wasn't enough to fill a 1 tick storage requirement even if incoming packets were big enough.

Ultimate 1.0.1 with updated gregtech, MFR, and AE.
Here's a video of the gui during operation:

youtu.be/DIMOUn7nCWw?t=8m52s

Cut and paste, the media link isn't working with timestamps for some reason.

 

[khorozm]

EUp = volts
EUt = amps
EUCableLength = Ohms
Cobble = Bananas

Simple

EU/p voltage
p/t is current (amps)
EU/t is power (energy per time)

 

[Omicron]

EUp = volts
EUt = amps
EUCableLength = Ohms

EU/p voltage
p/t is current (amps)
EU/t is power (energy per time)

Bad Biggle, that's not how it goes! khorozm almost got it, but only almost

EU is energy -> realworld joule
EU/t is power (energy/time) -> realworld watt
EU/p is charge -> realworld coulomb (very tenuously so...)
p/t is current (charge/time) -> realworld ampere

There is no equivalent to realworld voltage in the IC2 energy net. As far as electrical resistance goes, the comparison also can't be made because it doesn't equally affect all packet sizes. IC2 basically avoids the subject of both voltage and electrical resistance entirely by means of their EU packet size mechanic coupled with static loss per cable length, which creates a roughly adequate representation of "transform up to have less loss", but doesn't have anything to do with how that actually works in physics.

The advantage is that you don't need to use math... the disadvantage is that you can't use math.

 

[Jess887cp]

But...what are bannas then?

I can never remember the different EU/p for the different voltages. If someone posts them I'll add then to that nifty chart up there, and repost the image.
Or I could do the research myself...hmm.

EDIT: Bad photoshopping is bad. Oh well.

 

[PonyKuu]

Bananas are the fruits growing on palms

ULV <= 8
LV  <= 32
MV  <= 128
HV  <= 512
EV  <= 2048

Basically each next tier is the previous one multiplied by 4

 

[Omicron]

*looks at hands*

But there are no fruit growing on my palms at all!

(by the way, ULV is <= 5)

 

[Jess887cp]

Thanks guy, I edited my old post, but yours is much cleaner. I would edit the image again, but I'm way too lazy *shrug*

 

[PonyKuu]

(by the way, ULV is <= 5)

Those IC devs are always breaking nice concepts...