Unraveling the Mystery of Electrical Phases

[cabraham]

You can do it that way. If the secondary is wye with neutral grounded, then a hot wire & a neutral wire can be delivered to a home w/o a problem. Or a single phase xfmr can be used. The 2 hot wires from the wye (or hot & neutral) can be fed into a 1 phase xfmr. The 1 phase secondary can be grounded at either end of the winding, or center tapped like here in North America.

Different locales have their own methods of distribution and grounding. I was only pointing out that wye connected windings do not necessarily require a 4th wire to support unbalanced loads. Power reference manuals detail this along with the math. Does this help?

Claude

 

[sophiecentaur]

I see some of what you mean but I don't see why one should want three extra single phase transformers, co-sited with your three phase transformer, just to allow the use of a delta secondary. What they do on extended supply chain, with separate transformers on poles would clearly be very different. I assume that you do get what I'm saying about UK supplies.
I accept that you know what you're talking about on the unbalanced current thing. I shall have to read up on it. Power engineering is such a different world.

 

[cabraham]

The reason for the 3 single phase xfmrs, is not to allow for a delta, but for a center tapped 240V residential power. In N America it is common to have a 3 phase xfmr step down from 13 kV or so, to 2400V line-neutral, which is 4157V line-line if a Y is used on the secondary.

Then, 2 wires from the 2400/4157V 3 phase, 2 hots, or 1 hot plus 1 neutral, is fed into the primary of a 1 phase xfmr. The secondary is 240V center tapped. This center tap configuration means that each of the 2 outer hot wires is at a potential wrt Earth of just 120V rms.

We could, as you suggest, step the 13kV 3 phase down to 240V 3 phase and distribute to houses. But we don't have that center tap any more, which to me is a desirable thing. My understanding is that in Europe and other places, the house service is around 230V, with one side grounded, not center tapped. The hot wire has a potential of 230V rms wrt earth. This is IMHO less safe than the center tap system in N America.

I don't wish to start a war about which system is better, so let's just say they differ and let it go at that. BR.

Claude

 

[sophiecentaur]

The reason for the 3 single phase xfmrs, is not to allow for a delta, but for a center tapped 240V residential power. In N America it is common to have a 3 phase xfmr step down from 13 kV or so, to 2400V line-neutral, which is 4157V line-line if a Y is used on the secondary.

Then, 2 wires from the 2400/4157V 3 phase, 2 hots, or 1 hot plus 1 neutral, is fed into the primary of a 1 phase xfmr. The secondary is 240V center tapped. This center tap configuration means that each of the 2 outer hot wires is at a potential wrt Earth of just 120V rms.

We could, as you suggest, step the 13kV 3 phase down to 240V 3 phase and distribute to houses. But we don't have that center tap any more, which to me is a desirable thing. My understanding is that in Europe and other places, the house service is around 230V, with one side grounded, not center tapped. The hot wire has a potential of 230V rms wrt earth. This is IMHO less safe than the center tap system in N America.

I don't wish to start a war about which system is better, so let's just say they differ and let it go at that. BR.

Claude

Precisely. I get the impression that many (/most) houses in the US have their own transformer. Quite the opposite in the UK; virtually no one has their own transformer. Because we never used the 100V(ish) standard, we stuck with the same single (about twice) value of supply volts as cables can supply all domestic needs at 230V without being too thick. The original decisions were based on very different population densities in UK and US, I think.
Problem with using a centre tapped system in the UK model would be that you wouldn't be able to supply premises with three phase off the same transformer (not an uncommon requirement) as single phase domestic supplies - you'd get contention between the neutrals, each of which would be grounded somewhere. You are certainly right about the better inherent safety, though.

It's too late to change now, in any case. The only aggro these days seems to be between two sets of forum members who are unaware of the differences and make correct assertions about their own system that the other guys think are wrong because it's not the same on their side of the pond.

 

[mheslep]

Precisely. I get the impression that many (/most) houses in the US have their own transformer. ...

The US layout has at least several houses per xfmr. Lightning strike to a xfmr often takes out the street.

 

[sophiecentaur]

Interesting. So they're all on the the same phase then? It would make economic sense. These practical details about other peoples' lives are fascinating. We all live in different worlds.

 

[zgozvrm]

One signal, which cannot have a phase difference with itself.

Correct!

There is a 180° phase difference between the two sine waves.

Ummm, not quite…

There is only ONE primary winding and ONE secondary winding. We are applying a single voltage, to the primary side which is represented by one specific sine wave. A copy of that sine wave is then produced (induced) on the secondary winding, differing only in amplitude, assuming that there are fewer turns of wire in one winding than in the other. Let’s assume there are one-half as many turns of wire in the secondary winding as in the primary winding; that would mean that the voltage measured across the secondary winding would be one-half of that applied to the primary winding, at the same frequency. That is, 480 volts at 60 Hz applied to the primary would induce 240 volts at 60 Hz on the secondary.

Now, we connect a wire to the exact center of the secondary winding. This does not change the sine wave on the secondary (or on the primary). All it does, in effect, is give us 2 secondary coils that are in series with each other (and in phase). Each half of the coil has half as many turns of wire as the entire secondary winding, therefore, only half of the secondary voltage would be measured across either half of the secondary (in this case, 120 volts).

Note that by placing a tap at the center of the secondary winding, we haven’t changed either half of the coil. Had we actually separated the 2 halves, reversed one of them and re-connected them, THEN we would have two 120 volt sine waves that are 180 degrees out of phase with each other. But, when we measure across the entire secondary, the voltages would cancel out giving 0 volts, not 240 volts.

The reason people tend to think that one half of the secondary is 180 degrees out of phase with the other, is that they usually place the negative lead of an oscilloscope to the center tap (which is generally grounded) and measure one end or the other with the probe. This does show 2 sine waves that are out of phase, but you have to remember that by leaving the ground lead at the center tap, you are basically reversing the leads when you measure one side as compared to the other.

There is NO WAY that the 2 halves could truly be out of phase with each other, it's just an issue of perspective.

An example for the doubters:
Suppose you stand beside a train track and there is a train traveling from left to right, as you look at the track. If you look to the left, it will appear that the train is coming toward you. BUT, if you look to the right, it will appear that the train is going away from you. Obviously, the train is only going one way ... it's a matter of perspective.

 

[sophiecentaur]

What does "truly out of phase" mean? Use the centre tap as a reference (reasonable?). Those other two connections will be in antiphase as much as any other pair of antiphase signals.

 

[zgozvrm]

What does "truly out of phase" mean? Use the centre tap as a reference (reasonable?). Those other two connections will be in antiphase as much as any other pair of antiphase signals.

To state it another way, I could reword the sentence as:

"Truly, there is NO WAY that the 2 halves could be out of phase with each other, it's just an issue of perspective."
or
"There is NO WAY that the 2 halves could actually be out of phase with each other, it's just an issue of perspective."
or
"There is NO WAY that the 2 halves could really be out of phase with each other, it's just an issue of perspective."



As for using the center tap as a reference, that is my point...
It makes it appear that the sine waves are out of phase, when in all actuality, they aren't. (Just as the train appears to be going toward or away from you, when it is actually only going in one direction.)

The point being that simply tapping off of a transformer winding does NOT change the phase of any part of the sine wave. It merely reduces the amplitude of the individual sine waves that are now accessible due to the new access point (the tap).

 

[zgozvrm]

It's why we call it "split-phase" and not "2-phase"

 

[Averagesupernova]

I have gone round and round with someone on this forum about whether something is 180 degrees out of phase or not. I don't recall who. BUT, if we are going say that a center tapped secondary does NOT have each half 180 degrees out of phase with the other half then can we EVER say that ANYTHING is 180 degrees out of phase?

 

[vk6kro]

"There is NO WAY that the 2 halves could actually be out of phase with each other, it's just an issue of perspective."

There shouldn't be any doubt about this.

One wire can't have a voltage, a phase or a frequency.

Two wires can have an instantaneous voltage between them as long as you define a direction, like FROM wire A, TO wire B. Two wires cannot have a phase difference.

Four wires can have two instantaneous voltages and these may have a phase relationship with each other if they have the same frequency.

In the case of a tapped transformer winding, if you define the directions as AWAY from the center tap, then the phases of the waveforms on each end of the winding are 180 degrees out of phase.

 

[sophiecentaur]

'Split Phase' refers to.how the two signals happen to have been produced.
'Antiphase' because the two voltages which appear on the terminals (PDs referred to the earthed centre tap) are equal in magnitude and opposite in sign at all times. Those two varying voltages are indistinguishable from another pair of 'antiphase' signals, produced from two phase locked generators or whatever.

 

[Averagesupernova]

in the case of a tapped transformer winding, if you define the directions as away from the center tap, then the phases of the waveforms on each end of the winding are 180 degrees out of phase.

Amen!

 

[zgozvrm]

"There is NO WAY that the 2 halves could actually be out of phase with each other, it's just an issue of perspective."

There shouldn't be any doubt about this.

In the case of a tapped transformer winding, if you define the directions as AWAY from the center tap, then the phases of the waveforms on each end of the winding are 180 degrees out of phase.

I can't tell if you're agreeing with me or arguing with me.



Looking at something from a different perspective doesn't change its direction (or magnitude). A north-bound train goes north, regardless of which side of the tracks you look at it from ... even though it appears to travel from left-to-right when you stand on one side but appears to travel from right-to-left when you stand on the other side.

Likewise, we choose (for voltage selection and load balancing) to look at one half of the secondary winding of a split-phase transformer "from the other side." That doesn't change the waveform; it still rises and falls at the same rate and at the same period in time as the other half. It just changes our perspective of it.

And, to re-iterate, if the 2 "halves" were 180 degrees out of phase with each other, they would not combine to form 240 volts; they would cancel each other out to 0 volts (assuming of course that the average value of the sine wave is 0 volts).

Also, if they were 180 degrees out of phase with each other, we would have 2-phase power and could utilize that phase difference in the starting of electric motors. However, because they are NOT 180 degrees out of phase with each other, we need to use starting capacitors to create a phase difference.


It all comes down to wording...
The 2 halves of the secondary winding of a split-phase transformer are NOT 180 degrees out of phase, but (due to the reference point we choose), they appear to be 180 degrees out of phase with each other.

 

[sophiecentaur]

This is daft. How can the PD between two points 'have a phase difference'? It is only when you compare two signals referenced to a common point that you can assign a phase difference between them. Are the live and neutral wires in any particular phase relationship? No. They just have an alternating PD between them.

 

[jim hardy]

indeed wording is important , as is the mental picture we carry in our head

we forget that voltage is a potential difference
between two points

so to speak of a difference of phase between two voltages, we must choose some common point of reference

Sophie grew up using for reference one end of his electric company's transformer winding
i grew up using the middle
for both of us our reference point is earthedhere's a thought experiment
let there be two 50 hz generators,
one on the Earth and one on the moon so they cannot both be earthed
each is producing voltage 230√2sin(100∏t + θ) at its terminals

i assert an observer would find them in phase only if he were observing from a point equidistant between them, anyplace else he'd see two different θ's due to difference in transit times to point of measurement. Over the distance involved that time difference could amount to somewhat over a second, 50 whole cycles.

so my common point of reference is time.
is my thinking straight?

old jim

 

[zgozvrm]

Sophie and Jim: I'm not sure if your last post was directed to me, but for the most part, it sounds like we are all in agreement.



I don't agree with Jim's "thought experiment" though...

A single phase voltage, by itself, has no "direction" and therefore no phase angle. So, the generators can both supply 230 volts (that is, they each have a potential difference of 230 volts between their individual terminals), but each voltage waveform has no angle. Not until there is some physical connection between the 2 voltages, can you have a phase difference (phase angle) between them.

It's as if you took 2 arrows with you into space... If you let one arrow float in space, which way is it pointing? You can't answer that, since we haven't defined directions in space. And, certainly there is no phase difference when you consider just the one single arrow. Now, suppose you glued the 2 arrows at their tail ends at some fixed angle. You still cannot say which direction either arrow is pointing, but now that they're connected, you can measure the angle between them.

The same goes for voltages.

 

[sophiecentaur]

Ref the use of a two phase (180 degree phase diff) for motors. How could that work? Which way would the motor turn if windings were in exact anti phase? The whole point about an induction motor is that there is a quadrature component in the system so that there will be a torque. The way to achieve this when using a single phase supply is to use a shaded pole or capacity start system etc.. In a simple split phase system, the same 'frig' would be required.

 

[jim hardy]

thanks, zgo
i just wanted to keep the discussion going because it seems headed in a thoughtful direction.

you're right on about '...some physical connection...' though not necessarily by a copper conductor
and that was the point of my thought experiment
my observer ties them together at his oscilloscope, or at his clock, or whatever is his measuring device


your two arrows could be both referred to horizon and an angle between them measured, i think;
difference in their angles to horizon is analogous to phase difference ?
i never used a mariner's sextant but i think that's what it does for stars...

just as voltage refers to difference of potential, let me for just an instant call it potential displacement, phase refers to angular displacement . It requires two values to measure between.

......

Proof of Sophie's statement that you can feel:
take a single phase motor and disable its start winding, perhaps by lifting a wire from start capacitor.
Energize it and it will hum but not start.
Grab the shaft and you can turn in either way with your fingers, but move it slowly and don't grip it tight..
Give it a spin either way and it will accelerate and run that direction; that's why you don't grip it tight...

dont let it hum for very long or it will overheat..

single phase can be thought of as two phasors(vectors?) rotating opposite directions. Rotor current, once rotation starts, cancels out one of them.

 

[russ_watters]

And, to re-iterate, if the 2 "halves" were 180 degrees out of phase with each other, they would not combine to form 240 volts; they would cancel each other out to 0 volts (assuming of course that the average value of the sine wave is 0 volts).

You're misunderstanding what is going on. If you tried to combine the two signals on one wire, they'd cancel, but the two signals aren't traveling down the same wire, they are traveling down separate wires. Just draw yourself a graph!

A single phase voltage, by itself, has no "direction" and therefore no phase angle. So, the generators can both supply 230 volts (that is, they each have a potential difference of 230 volts between their individual terminals), but each voltage waveform has no angle. Not until there is some physical connection between the 2 voltages, can you have a phase difference (phase angle) between them.

This is correct...and it is why when you feed both wires into the same device, you are now using each as the reference for the other.

When you measure the voltage between the two wires, you get 240V because when one is +120V, the other is -120V.

 

[Averagesupernova]

I can't tell if you're agreeing with me or arguing with me.



Looking at something from a different perspective doesn't change its direction (or magnitude). A north-bound train goes north, regardless of which side of the tracks you look at it from ... even though it appears to travel from left-to-right when you stand on one side but appears to travel from right-to-left when you stand on the other side.

Likewise, we choose (for voltage selection and load balancing) to look at one half of the secondary winding of a split-phase transformer "from the other side." That doesn't change the waveform; it still rises and falls at the same rate and at the same period in time as the other half. It just changes our perspective of it.

And, to re-iterate, if the 2 "halves" were 180 degrees out of phase with each other, they would not combine to form 240 volts; they would cancel each other out to 0 volts (assuming of course that the average value of the sine wave is 0 volts).

Also, if they were 180 degrees out of phase with each other, we would have 2-phase power and could utilize that phase difference in the starting of electric motors. However, because they are NOT 180 degrees out of phase with each other, we need to use starting capacitors to create a phase difference.


It all comes down to wording...
The 2 halves of the secondary winding of a split-phase transformer are NOT 180 degrees out of phase, but (due to the reference point we choose), they appear to be 180 degrees out of phase with each other.

Tell me this then: Suppose we had two wire-pairs with each pair having 120 VAC on them. They are 180 degrees out of phase with each other by your definintion (whatever that actually is because I cannot tell from your posting). The pairs are electrically isolated from each other using isolation transformers or the method of your choice UNTIL we connect one lead from each pair together. At this point, what will the voltage be between any two of the three nodes?

 

[zgozvrm]

Ref the use of a two phase (180 degree phase diff) for motors. How could that work? ...

Of course, you are correct ... I got ahead of myself a bit there!

For 2-phase to be useful, the phase difference would have to be something other than 0 or 180 degrees; the vectors couldn't be "in line" with each other. The ideal offset would be 90 degrees which would generate the most starting torque in a motor.

In fact 90 degree, 2-phase was the standard in the early 1900's in the U.S. I believe it is still used in a few locations.

 

[zgozvrm]

your two arrows could be both referred to horizon and an angle between them measured, i think;
difference in their angles to horizon is analogous to phase difference ?
i never used a mariner's sextant but i think that's what it does for stars...

No, the phase angle would be the angle between the arrows, regardless of any other chosen reference point. To measure the angle between any 2 vectors (my arrows, for instance), you must connect their tails, then measure the angle between the vectors.

 

[zgozvrm]

you're right on about '...some physical connection...' though not necessarily by a copper conductor
and that was the point of my thought experiment
my observer ties them together at his oscilloscope, or at his clock, or whatever is his measuring device

If you connect your "measuring device" to the terminals of each generator at the same time (in order to make a comparative measurement), you are, in fact, making a physical connection.

 

[zgozvrm]

You're misunderstanding what is going on. If you tried to combine the two signals on one wire, they'd cancel, but the two signals aren't traveling down the same wire, they are traveling down separate wires. Just draw yourself a graph!

There has to be 2 wires to be useful, otherwise you have no circuit.
If we take one wire from one of the hot bus bars in a 120/240 split-phase panel, and another wire from the other hot bus bar, we are taking the full voltage across the SINGLE transformer winding and would have 240 volts. See my next post for a clearer example.

They'd only cancel if they were 180 degrees out of phase and of the same magnitude.

In other words, if you combined a 183 volt, 60 Hz signal with a another 183 volt, 60 Hz signal that was 180 degrees out of phase, the result would be 0 volts.

However, if you combined a 183 volt, 60 Hz signal with a 47 volt, 60 Hz signal that was 180 degrees out of phase, the result would be a 136 volt, 60 Hz signal.

If you combine two 183 volt, 60 Hz signals that were 90 degrees out of phase, the result would be 258.8 volts at 60 Hz.

If the two 183 volt signals were exactly in phase with each other, they would combine to form 366 volts.


Just draw yourself a graph!

See http://www.acs.psu.edu/drussell/demos/superposition/superposition.html for a reference if you need.
Also, look up vector addition.

 

[zgozvrm]

Tell me this then: Suppose we had two wire-pairs with each pair having 120 VAC on them. They are 180 degrees out of phase with each other by your definintion (whatever that actually is because I cannot tell from your posting). The pairs are electrically isolated from each other using isolation transformers or the method of your choice UNTIL we connect one lead from each pair together. At this point, what will the voltage be between any two of the three nodes?

First of all, they would only be 180 degrees out of phase depending on your point of view.
They would have to be completely in synch with each other, which (unless they came from the same source) would be difficult without other equipment to adjust for minor offsets in speed.

To answer your question: It would depend on how you connected them...
If you connected them one way, the 2 signals would be 180 degrees out of phase, in which the voltage across the combination would be 0 volts, whereas if you connected them another way, they would be in phase with each other and their voltages would add to 240 volts.


Take for instance 2 D-Cell batteries. Each measures 1.5 volts. When we combine them (as in a flashlight, for instance), the total voltage across the pair is 3 volts. But, if you reverse one battery (making it 180 degrees out of phase with the other), the combined voltage is 0.

If we combine those batteries the "correct" way (where the total voltage across the pair is 3 volts), we can still measure each battery independently: Place the black probe of a voltmeter on the negative terminal of the combined pair and place the red probe at the junction between the 2 batteries ... you'll get 1.5 volts. Next, place the black probe at the junction between the 2 batteries and the red probe at the positive terminal of the combined pair ... you'll read 1.5 volts from the other battery. Now, if you leave the black probe at the junction and move the red probe to the negative terminal of the combined pair, the meter will read NEGATIVE 1.5 volts! All we've done is changed our perspective of that battery (how we look at it), but we have not changed its orientation with respect to the other battery ... they are still in phase.

 

[Averagesupernova]

No, they'd only cancel if they were 180 degrees out of phase and of the same magnitude.

In other words, if you combined a 183 volt, 60 Hz signal with a another 183 volt, 60 Hz signal that was 180 degrees out of phase, the result would be 0 volts.

However, if you combined a 183 volt, 60 Hz signal with a 47 volt, 60 Hz signal that was 180 degrees out of phase, the result would be a 136 volt, 60 Hz signal.

If you combine two 183 volt, 60 Hz signals that were 90 degrees out of phase, the result would be 258.8 volts at 60 Hz.

If the two 183 volt signals were exactly in phase with each other, they would combine to form 366 volts.


Just draw yourself a graph!

See http://www.acs.psu.edu/drussell/demos/superposition/superposition.html for a reference if you need.
Also, look up vector addition.

You are barking up the wrong tree. What you describe can be accomplished through summing. Of course if you sum 2 signals of equal magnitude and opposite phase they will cancel but what is being discussed here about opposite ends of a transformer with a center tap does not apply here.

 

[zgozvrm]

You are barking up the wrong tree. What you describe can be accomplished through summing. Of course if you sum 2 signals of equal magnitude and opposite phase they will cancel but what is being discussed here about opposite ends of a transformer with a center tap does not apply here.

It absolutely DOES apply...
See my example in post #57 which compares the AC situation we are talking about with its DC equivalent.

 

[zgozvrm]

Suppose you had 2 separate, identical single phase transformers, each having a secondary voltage of 120 volts. Now, let's assume that a single 480 volt source is connected to the primary of each transformer (identically connected)...

The primaries are obviously in phase with each other since they are supplied by the same voltage source (any signal is in phase with itself).
The signal on the secondary of any transformer is in phase with the signal on its primary (due to the way transformers work; how voltage is induced).
Therefore the voltages on the secondaries of each transformer are in phase with each other.
Agreed?


If we connect the X2 terminals of each transformer together, one will be reversed with respect to the other (their secondary voltages will be 180 degrees out of phase) such that the sum total of the voltage across the combination (as measured from the X1 terminals) will cancel out to 0 volts.
But, if we connect X2 of xfmr #1 to X1 of xfmr #2, the total voltage across the 2 transformers (as measured from X1 of xfmr #1 to X2 of xfmr #2), we would see 240 volts. That is because the signals add, and they are in phase.
Now if we attach a wire to the connection between the 2 transformers, we have in effect, created a center tap. Note that the secondary voltages are still 120 volts each and they are still in phase with each other.

The center tap does not change the phase (or polarity) of any part of the transformer's windings. It merely gives you a point to tap off a lesser voltage. Granted, in reference to the center tap, the 2 signals look to be 180 degrees out of phase. In fact, they are not. The presence of that tap didn't change the orientation of one half of the winding!