Transformer Questions on Core Type, Reactive Currents

[wwhitney]

1) Say I give you a black box that performs as a 240V 2-wire to 120/240V 3-wire isolation transformer. Inside could be a single core transformer with a split secondary, or two separate 240V:120V transformers, with the primaries in parallel but the secondaries in series in the proper way. Is there any external testing you can do to determine which transformer arrangement is inside?

2) Suppose I connect 10A constant current loads L1-N and L2-N, with the L1 load 60 degrees lagging power factor, while the L2 load is 60 degrees leading power factor. So the component of the current in phase with the voltage is 5A for each load, but the reactive currents are in opposing directions. Is the primary current (as measured outside the black box) simply 5A in phase with the voltage?

Thanks,
Wayne

 

[LarryFine]

1) Say I give you a black box that performs as a 240V 2-wire to 120/240V 3-wire isolation transformer. Inside could be a single core transformer with a split secondary, or two separate 240V:120V transformers, with the primaries in parallel but the secondaries in series in the proper way. Is there any external testing you can do to determine which transformer arrangement is inside?

I don't think so.

I thought about energizing one of the 120v windings and checking for voltage on the other one, but the paralleled primaries would cause that to happen anyway.

I then thought about a DC resistance check, but the common shared secondary output wire would make it behave the same way also.

2) Suppose I connect 10A constant current loads L1-N and L2-N, with the L1 load 60 degrees lagging power factor, while the L2 load is 60 degrees leading power factor. So the component of the current in phase with the voltage is 5A for each load, but the reactive currents are in opposing directions. Is the primary current (as measured outside the black box) simply 5A in phase with the voltage?

Tougher, but I'm thinking again that the paralleled primaries will make two units behave as one.

 

[jaggedben]

1) Say I give you a black box that performs as a 240V 2-wire to 120/240V 3-wire isolation transformer. Inside could be a single core transformer with a split secondary, or two separate 240V:120V transformers, with the primaries in parallel but the secondaries in series in the proper way. ...

Maybe I'm wrong but can't it also be two 120V:120V isolation transformers, with just a neutral brought out from the middle on one side and not the other?

 

[LarryFine]

Maybe I'm wrong but can't it also be two 120V:120V isolation transformers, with just a neutral brought out from the middle on one side and not the other?

If you mean with the primaries in series, that would allow each unit to behave independently.

Then my suggestions ought to work.

 

[wwhitney]

Maybe I'm wrong but can't it also be two 120V:120V isolation transformers, with just a neutral brought out from the middle on one side and not the other?

Then if there's no primary neutral, the voltage of the midpoint of the two transformer primaries is not fixed, and from the point of view of the primary supply you just have two loads in series, the two primary transformer coils. Or if you like, the two reflected impedances from the secondary loads, which with a 1:1 turns ratio would be identical to the actual secondary impedances.

E.g. if on the secondary you had 5 ohms L1-N and 10 ohms L2-N, the primary would see 15 ohms across the 240V supply, giving a current of 16A. The floating internal primary midpoint would be at 160V / 80V from the two primary wires, and the secondary would see 80V L1-N and 160V L2-N.

Cheers, Wayne

 

[jaggedben]

Here's an actual transformer which, if I'm understanding the instructions and pictures correctly, has two cores thay can be wired either as in your second scenario or as I suggested (with 480V primary, of course). Am I misunderstanding something?

15kVA Split-Phase Step-Down Transformer - 480/240V to 120/240V | Current Connected

Experience reliable and durable power conversion with the 15kVA Split-Phase Step-Down Transformer. Reliable performance in demanding conditions.

www.currentconnected.com

 

[wwhitney]

Here's an actual transformer which, if I'm understanding the instructions and pictures correctly, has two cores thay can be wired either as in your second scenario or as I suggested (with 480V primary, of course). Am I misunderstanding something?

The pictures of the two sets of coils confused me at first, but if you look closely they are on a single iron core. You can see the core laminations above the two sets of coils and below the H1/H2/H3 bus bars.

So that is not what you proposed in post #3. The transformer in post #6 is just a single core transformer with a split primary coil, allowing it to be energized by either 240V two wire by wiring the two halves of the primary in parallel, or 480V two wire by wiring the two halves of the primary in series.

When the two series primary coils are on a common core, then they just look like a single coil; the magnetic side of things (which I don't understand as well as I would like) enforces a single value of volts/turn for both coils. Whereas if you have two series primaries each on its own core, then each primary can have a different volts/turn, and the situation I described in post #5 applies.

Cheers, Wayne

P.S. If the core in the above transformer were a figure 8 core, with the coils on just 2 of the 3 legs, then I believe you are correct that it would behave like two separate transformers as in post #3. But the core is just a single loop with 2 legs.

 

[synchro]

1) Say I give you a black box that performs as a 240V 2-wire to 120/240V 3-wire isolation transformer. Inside could be a single core transformer with a split secondary, or two separate 240V:120V transformers, with the primaries in parallel but the secondaries in series in the proper way. Is there any external testing you can do to determine which transformer arrangement is inside?

If the transformers were perfectly linear it would be hard to distinguish one case from the other.
But say you applied some DC current across the L1-N on one side of the split secondary so that in one direction of the AC waveform the transformer goes into partial saturation. Assume that the primary is driven with a low impedance source so that the current peaks drawn during saturation do not distort the voltage waveform on the primary side.

If there were two transformers, the L1-N output voltage would show some waveform distortion on an oscilloscope, but the L2-N output would not be distorted because its transformer would not be saturating.
But if there was a single transformer then both the L1-N and L2-N windings would be on the same core, and so they would both have a distorted output voltage waveform across them.

2) Suppose I connect 10A constant current loads L1-N and L2-N, with the L1 load 60 degrees lagging power factor, while the L2 load is 60 degrees leading power factor. So the component of the current in phase with the voltage is 5A for each load, but the reactive currents are in opposing directions. Is the primary current (as measured outside the black box) simply 5A in phase with the voltage?

Yes, I believe the primary current should be 5A in phase with the voltage.

The situation you described could be created by having an inductor across the L1-N winding and a capacitor across the L2-N winding, and where the inductive reactance is the same value as the capacitive reactance. Stored energy would be exchanged through the transformer between the capacitor and inductor during each AC cycle, and ideally it would not affect the primary current. However, there will be some relatively small losses in the transformer from this circulating reactive current, and it will contribute a small current draw on the primary side.
As you say, the reactive currents developed will be in opposing directions, and they would not be transformed over to the primary side.

 

[wwhitney]

Yes, I believe the primary current should be 5A in phase with the voltage. . . . As you say, the reactive currents developed will be in opposing directions, and they would not be transformed over to the primary side.

Well in the case of two separate 240V : 120V transformers in parallel, I believe the reactive current is reflected on the primary side of each transformer, and that the currents cancel when they add in the joint primary conductors. So not reflected outside our black box.

But in the case of a single 240V : 120/240V transformer, and ignoring the black box aspect for the moment, is the cancellation in the secondary coil itself? I.e. is the magnetic side of things identical for the 10A 60 degree case I specified and the corresponding 5A resistive case?

Thanks,
Wayne

 

[wwhitney]

P.S. If the core in the above transformer were a figure 8 core, with the coils on just 2 of the 3 legs, then I believe you are correct that it would behave like two separate transformers as in post #3. But the core is just a single loop with 2 legs.

An amendment to the above: I didn't look closely at the pictures to determine how the coils are wound around that 2 leg core. If each leg has one primary coil and one secondary coil, then my above answer applies, although I'm not 100% confident in that.

But if one leg has both primary coils, and one leg has both secondary coils, then by being on the same leg the two primary coils will have to have the same volts/turn. I'm unclear on what the presence of the 3rd leg would do in that case.

Cheers, Wayne

 

[jim dungar]

An amendment to the above: I didn't look closely at the pictures to determine how the coils are wound around that 2 leg core. If each leg has one primary coil and one secondary coil, then my above answer applies, although I'm not 100% confident in that.

But if one leg has both primary coils, and one leg has both secondary coils, then by being on the same leg the two primary coils will have to have the same volts/turn. I'm unclear on what the presence of the 3rd leg would do in that case.

Cheers, Wayne

Transformers are usually wound with the secondary (LV) windings closest to the core and the primary (HV) windings over the top of the secondary windings.

It would be rare for primary and secondary windings to be on separate legs.

 

[synchro]

But in the case of a single 240V : 120/240V transformer, and ignoring the black box aspect for the moment, is the cancellation in the secondary coil itself? I.e. is the magnetic side of things identical for the 10A 60 degree case I specified and the corresponding 5A resistive case?

I would say that the cancellation is happening within the transformer core that is conveying magnetic flux though all of the windings, and is coupling the windings together via Faraday's law of induction.

 

[synchro]

Transformers are usually wound with the secondary (LV) windings closest to the core and the primary (HV) windings over the top of the secondary windings.

It would be rare for primary and secondary windings to be on separate legs.

This is certainly true for transformers used in the distribution of electrical power, which is the application that we are primarily addressing on this forum. For such applications the transformer impedance is relatively low for good voltage regulation, but high enough to keep the short circuit current acceptable.

There are applications outside of power distribution where a relatively high impedance transformer with a high leakage inductance is desired. Examples include neon sign transformers, welders, constant voltage transformers, previous generations of microwave ovens, etc. In addition to having the primary and secondary windings on separate legs, such transformers often include magnetic "shunts" in the core to even further increase the leakage inductance and therefore the transformer's impedance.

 

[wwhitney]

I would say that the cancellation is happening within the transformer core that is conveying magnetic flux though all of the windings, and is coupling the windings together via Faraday's law of induction.

So I take that to mean that if the transformer core has all windings on one leg, and physically the windings along the length of the leg are secondaries S1 followed by S2 (spatially disjoint), with the single primary winding P overlaid on both, then: S1 and S2 both see 10A (5A in phase, 5*sqrt(3) A reactive); P sees 5A (in phase); and the magnetic flux through the core at a cross-section underneath S1 differs from the magnetic flux at a cross-section underneath S2. Is there a standard way to draw this out as a magnetic circuit? (Referrals to standard references appreciated).

Conversely, if the mixed 10A 60 degrees leading/lagging loads are replaced with 5A resistive loads, then S1, S2, and P all see 5A, and the magnetic flux at any cross section of the core is the same.

Is that all correct? And what changes in the first scenario if the windings S1 and S2 each extend the full length of the core, i.e. are intertwined? Do we then find that the magnetic flux at any cross section of the core is same, as the effects of the reactive current are cancelling spatially infinitesimally?

Thanks,
Wayne