Transformer question

[LarryFine]

I've seen diagrams of 3 phase transformers wound on a figure 8 (3 leg) core, presumably with each leg having a primary and secondary coil for one phase. That would suggest that the core does not force the different legs to all be in phase.

Correct. They behave like three independent cores.

I've also seen a diagram of a single phase transformer wound on a loop (2 leg) core, where one leg has the primary coil and one the secondary coil. [I gather this is not done in practice, but was illustrated that way for educational purposes.] That would suggest that the core forces the two different legs to be in phase.

Correct again. They behave like one single core.

So what's the difference? Is it a matter of degree? Or does the 3 leg 3 phase transformer rely on a cancellation property that you get with a 3 phase input? I.e. if you took that transformer and instead energized it with 3 out of 4 phases of a 2 phase system, there'd be a problem?

I imagine that each phase would behave independently.

 

[wwhitney]

Correct. They behave like three independent cores.
. . .
Correct again. They behave like one single core.

OK, but why does the figure 8 behave like three independent cores, while the loop behaves like one single core? That doesn't make any sense to me.

Speaking of core geometry and the effect on the transformer behavior, say you have two bare single phase transformers, each consisting of just a bar with single primary and secondary coils cowound along the full length of the bar. [I gather a loop core perform better, but I don't understand the details.] Supply each primary with a different phase.

If the two transformers are arranged so that the cores are coaxial but a foot apart, I assume everything works as expected. What happens if I move the transformers so that the cores are only 1 cm apart? What happens if I then swap out the two separate bar cores for one long bar core?

Cheers, Wayne

 

[jim dungar]

Transformers work based on the magnetic flux flowing through the core. It is somewhat like the current in a conductor, a closed loop is required. However the magnetic flux can flow through air, so while inefficient, a transformer core could be a straight bar. You need to get into some detailed engineering to really understand transformer core construction, but effectively the three phase are wound on the different core legs in a manner that allows the magnetic flux to flow in the same direction (which is what I think you are calling 'in phase').

 

[winnie]

All coils which 'link' the same magnetic flux are in phase.

The steel is a magnetic conductor. In a 2 leg core there is only one magnetic path, so the coils on both legs link the same flux.

On a three leg core there are multiple magnetic paths, so that the flux in one leg is equal to the sum of the flux in the other two legs. This let's a three leg core work for balanced 3 phase use, but it isn't 3 completely independent phases.

This approximation falls apart because air is also a magnetic conductor and there are no magnetic insulators. A text on magnetic circuit design described this as trying to design electrical circuits using bare copper sitting in salt water.

Jon

 

[wwhitney]

On a three leg core there are multiple magnetic paths, so that the flux in one leg is equal to the sum of the flux in the other two legs. This let's a three leg core work for balanced 3 phase use, but it isn't 3 completely independent phases.

That means you could have 2 completely independent phases on a figure 8 core, right? So for my last question with the bar cores (where the magnetic path is completed through the air), if instead loop cores were used, so that the part about joining them would create a figure 8 loop, there'd be no problem with any of the arrangements?

But if I understand correctly, as originally posed, bringing the bar cores in proximity coaxially or connecting them would be a problem? As the lower reluctance of the cores would suggest a single magnetic path through both cores and looping back through air, rather than the two separate loops required to allow the two transformers to function on separate phases?

Thanks,
Wayne

 

[winnie]

You could have 2 completely independent phases on a figure 8 core where the 'common' leg has twice the cross section of the independent legs.

'Bar' cores would be quite unlikely in power distribution; air core transformers are pretty well reserved for signal work at high frequencies. But they would be independent if far enough apart. In proximity they would interact with each other.

-Jon

 

[LarryFine]

There are too many to choose some to post, but look at these Google-search images and go to some of the sites that they come from:

flux in 3 phase transformer core - Google Search

 

[Mr.jjrooney]

Sorry. The transformer size is 112.5 kva - primary 480v, secondary 120/208v.

Seeing the size of the transformer and the requirements of the vessel I will presume that the transformer serves multiple vessels.
1) the vessel in question load nameplate requirements may very well be a maximum voltage requirement.
2) if it is not a maximum requirement, then the loads that require "240v" should be verified by their individual nameplates.
3) if they are compressors or motors labeled for 240v they will not run a thier rated efficiency and will draw more amps creating more heat and their life expectancy will be shortened.
4) a 120/208v system is connected as a wye system, 120v to neutral from each phase/line. A 120/240v is usually either single phase or a open or closed 3 phase delta system with a center tap grounded conductor
5) Buck-boost transformers should not be used to create a 240v/120v single phase system fed by a 208Y/120V 3-phase, 4-wire source. If this is done , 2 neutrals would exist on the same circuit. This wiring creates unbalanced line to neutral voltages where one line is 120v and the other line line greater than 130v.
6) If the 240v loads on the vessel are not 208v-240v, then a separate 120v/240v-16v/32v buck-boost transformer should be connected after the source and directly before the loads requiring 240v only.
Not taking into consideration voltage drop from the source this should give you 236v.
7) according to USA C.F.R. Title 46 Part 111 section 111.01-17
Voltage and frequency variations.
Unless otherwise stated, electrical equipment must function at variations of at least ±5 percent of rated frequency and +6 percent to −10 percent of rated voltage. This limitation does not address transient conditions.

To answer your question if this will damage the transformer the answer is no if it is not overloaded from other loads. It very well may damage equipment on the vessel by not supplying proper voltage.

Hope this clearly answers your question.