Why can any two phases be connected together?

Averagesupernova

I like to explain things in terms that makes it easy to visualize. 3-phase delta is a pretty easy one to answer. There are three transformer windings (secondaries) that are hooked in a series. Drawn out they appear as a triangle, hence the reason we call it delta. Each 'phase' comes off of a node from two windings. So, grab any two phases and you can see they are directly across a transformer winding. I don't see how you could not see that you can source power from any two phases. Maybe I missed the point?

Phrak

I think a phase diagram is the only thing that can return sanity to this thread.

sophiecentaur

foo - your red and blue picture is 'unconventional' and I think it shows that you are confused about what is happening. Arrows, such as you have drawn are usually taken to mean current flow. In that diagram, there would be no net current flow because it doesn't have anywhere to go - the two currents are in opposite directions.

Two wires with potentials which are alternating and in anti-phase will be in the same situation as if one wire is at Earth and the other is at twice that potential - the difference in potential is the same in each case.
If they are supplied with enough current to maintain them at these potentials then you can get power out of the arrangement. "Back at the supply" you could have a transformer winding which is connected to the two wires. If you connect the centre turn of this winding to Earth, the two outputs will be 180 degrees out of phase ' about Earth'.
You could also achieve this with two transformers, one for each wire and with its other output connected to Earth. The outputs could be chosen to he in phase or in anti phase- depending on which way round you connect the wires.
In a three phase supply your generator will produce three outputs, each of which is 120 degrees out of phase to the others. The generator often have three windings, connected as a 'star' or "Y" with one end of each of its windings at the centre and the other ends will have the three phases. This centre point may be connected to Earth, keeping the three phases nicely symmetrical about 0V. The 'return path' for currents flowing through a load connected across two of the phases will be via two of the windings - that was one of the original questions.

You really need to browse through Wikkers (and all the rest) to get a better idea of what's going on.

zgozvrm

First of all, you would never want to connect 2 different phases together. I think what you mean to say is “connect a load to 2 different phases.”

I know, I know … it sounds like a minor detail. But then again, if someone owes me $1000.00 dollars but misplaces the decimal point and only writes me a check for $10.00, I’m going to be a little upset.

The point is, precision and accuracy are important (not only in calculations, but in what you say as well).

zgozvrm

I think you're under the misconception that, at any given time, there is a positive voltage on one leg and a negative voltage on another leg. That is not true. Take 120V, single phase for example: You have a hot wire and a neutral wire. The neutral wire is always at 0 volts, whereas the hot wire oscillates between positive 120V and -120V (60 times per second). So there is never a time where you have 120V on one leg and -120V on the other (or +10V on one and -10 on the other, etc.) since one is always at zero volts.

zgozvrm

Two different phases of a standard three-phase system are never in phase with each other. Each phase is at the same frequency (60 Hz in the U.S.), so they rise and fall at the same rate and one never "catches up" with another. Therefore, you can't have one leg (a "phase wire" as you call it) at 0V, 120 degrees and another leg at 100V, 120 degrees.


I'm not sure, but what you may be trying to say is something like this:

"I am now finding out that when one leg is at 120 degrees in its cycle, it is at 0V and, at the same time, another leg is at 100V."

But this isn't even possible, because when one phase is at 0V, the other 2 phases are at approximately 103.9V and -103.9V, respectively. Similarly, when one phase is at 100V, the other 2 phases are at approximately 7.4V and -107.4V respectively.

zgozvrm

You don't need to have one line positive and one line negative in order to get current to flow. All you need is a difference in potential.

For example, in a DC circuit, you could supply a load with 2 different positive voltages, say 10V and 8V, you would have a potential difference of 2V. Therefore the circuit would be the same as if it were supplied by a single 2V battery where the positive side of the 2V battery would be pointing in the same direction around the circuit as the 10V battery did. (see attachment)

Attached Thumbnails

 

zgozvrm

Also, if you have 2 equal voltages from different phases, you still have a potential difference, and therefore you will have current flow. For instance, say you have one leg measuring 100V at 0 degrees, and another leg measuring 100V at 120 degrees, there is a potential difference of 100V at 60 degrees. This can be shown using vector addition (see attachment).

Note that the only time the voltages between any 2 phases of a 3-phase system coincide is at plus or minus 60V.

Attached Thumbnails

 

Averagesupernova

Need to be a bit careful how this is worded. This is easily measured and understood in a 3-phase wye setup. But 3-phase delta, not so easy. 3-phase delta typically center-taps the one transformer winding that is between the A and B legs for a neutral connection. In this scenario A and B can now be used for 120 volt legs in reference to the neutral connection and connection between A and B can run a 240 volt appliance such as a residential water heater or something. With these three wires (leg A, leg B and neutral), you cannot differentiate between single and 3-phase. Viewed on a scope with the scope grounded to the neutral in this scenario A and B will appear 180 degrees out of phase and not 120 degrees. So your original quote is correct, but possibly misleading. Is everyone even more confused than ever now?

zgozvrm

Yes and no:
This is the similar to having a single phase transformer where the secondary is center-tapped (as far as those 3 wires are concerned). The fact of the matter is that a single secondary coil can produce only one phase (one sine wave). You don't create another phase by tapping off that coil, you create 2 voltages that, when added together give the total voltage produced by the coil (end-to-end). These voltages are shown by sine waves having amplitudes less than that of the full voltage; these amplitudes, when added together, will be equal to that of the full secondary voltage. If these voltages were indeed 180 degrees apart, they would add to be 0V, rather than 240V. The reason most people think they see them as being out of phase by 180 degrees is because they are reversing the o-scope leads with respect to each other.

To illustrate this point, reference the attached pic.

The secondary voltage, measured from point 1 to point 3 will either be in phase with the primary voltage, measured from point A to point B, or it will be 180 degrees out of phase from the primary voltage, depending on how the transformer was wound and/or how the secondary taps were labeled. Either way, the secondary coil produces a single voltage, from which several voltages can be tapped off. This gives you voltages that are less than the full voltage but in phase.



Let me clarify:

Suppose you have an o-scope capable of measuring 3 signals at once.
Place the channel 1 ground clip on lead #1 and the channel 1 probe on lead #3; you will get a 240V sine wave.

Now, place the Ch 2 ground clip on lead #1 and the Ch 2 probe on lead #2; you will see a 120V sine wave that is in phase with the 240V sine wave measured by Ch 1. That is, it will rise and fall at the same time as the Ch 1 waveform, but only by 1/2 the amplitude.

Next, place the Ch 3 ground clip on lead #2 and the Ch 3 probe on lead #3; you will see a 2nd 120V sine wave that is in phase with the other 2 waveforms.

Notice that the ground clip for each channel is to the left of its corresponding probe.

What most people do, is place the ground clips of both channels 2 and 3 to lead #2 (the center tap, which is generally grounded), then place one probe on lead #1 and the other on lead #2. This will give you 2 waveforms that read 120V each, but they will appear to be 180 degrees out of phase. This is because the order of the probes has been reversed for one channel in relation to the other.

This is the same as measuring 2 D-cell batteries in series. Each battery measures 1.5V by itself, but the combination of the two measures 3V. If you were to place the negative lead at the center point of the two series-connected batteries and measure the ends of the combination with the positive probe, you would get +1.5V for one battery, and -1.5V for the other (they would appear to be "out of phase" with each other.

So, in actuality, the two 120V sources obtained from such a set-up are in phase with each other, but are 180 degrees out of phase with each other with respect to the center tap.

Attached Thumbnails

 

Averagesupernova

Phase is always measured relative to something else. So in your attachment zgozvrm, what is relative to what concerning the 120 degrees? You can't just take 2 wires no matter what the source is and say they are 120 degrees out of phase, which is what you have done on the primary side. You need someplace else to ground your scope at besides one of the two wires you provided in the schematic, or the center tap. This is the point I am trying to make. You can't do this with just one transformer. You are still implying that a pair of wires from leg A and B, with a center tap, or run through a transformer as in your attachment with a secondary center tap, that somehow we maintain an ability to measure a 120 degree phase difference. It cannot be done. There is nothing I don't understand about this, however, I'm not always very clear about it.

sophiecentaur

If you show all three sinusoids on a scope (with the common neutral being at Earth) and trigger on one of them (natch - to make the trace stand still and to give you a reference) you will see three sine traces, equally spaced in time. What you are seeing is each individual waveform ('phase'). Each will have an peak value of, say, 100V. I am assuming that a Y connection is being used; a delta connection would not really be usable because there is no inherent Earth reference to hang your scope Earths on.

If you were to connect a load between two of the phases there would be 170V (peak) - actually 100√3 V across that load. This is more than the PD of one phase to Earth but less than you'd get if the two sinusoids were exactly in antiphase (that would be 200V).
Try drawing out the three squiggles and see how the spacing between them (the PD) varies - or look at "three phase" pictures everywhere on the web.

zgozvrm

I meant to remove the "120 degrees" from the notation.

Notice that I didn't refer to that at all in the text of my posting. (All that really shows is that there is really no change in phase from the primary to the secondary.)

I was merely showing that the two legs of 120 are not, in actuality, 180 degrees out of phase because they are actually parts of the same voltage.


I also said (as did you) ...
...that you cannot create a second phase with only one transformer. Which means that tapping off the secondary cannot produce another phase.



It is perfectly acceptable to take measurements with a scope in the way that I stated (you don't need "someplace else to ground your scope.") Granted, you must have different voltages out of phase from each other in order to say that voltage A is at angle X and voltage B is at angle Y. But you can also take several measurements and compare them to each other, as I did in my earlier post. The point I was making is that because there is only one secondary coil, therefore you cannot take a voltage from it that is at another angle unless you reverse the leads of the scope with respect to each other (in which case they will appear to be 180 deg out of phase).

zgozvrm

You don't need to "hang your scope 'Earths' on" a ground. A scope measures non-grounded signals, as well as grounded signals. So, to measure the secondary signals from a delta transformer (assuming the legs are labeled X1, X2 and X3, as is standard), you would connect the leads as such:

Channel 1
Ground to X1, Probe to X2

Channel 2
Ground to X2, Probe to X3

Channel 3
Ground to X3, Probe to X1

sophiecentaur

'Scopes, these days may have a switch to decouple the probe ground lead from earth but a BNC connector is inherently an unbalanced connector. I agree that, with so many volts involved, it doesn't really matter but you wouldn't want to have floating earths for sensitive work - would you?

zgozvrm

Agreed, but most scopes I've dealt with can handle a potential of 300V between the terminals of the BNC. If you're sure the voltage won't go over 300V, then this shouldn't be a problem.

Besides, whether one scope or another is suited to measure an ungrounded voltage source is not the issue here.

foo

Thanks all for the help. I think I see now what is happening. The three phases are shifted slightly so that there is a smoother power flow into a load. I guess the extra pulses reducing the lull in power between phases as opposed to the spacing between each wave in single phase acts like extra torque.



So, this shifting is represented in this image that shows how the three move back and forth.
Which explains how current is traveling, thanks for the wikki suggestion for research.



So, then I take it even the load doesn't need to have a neutral? That the power circulates inside the load in some way that completes the circuit. I can't imagine how this looks. I am guessing that it's not exactly how the image is depicting that the three phases are connected because that would be some hellish dead short. How does the load complete the circuit on the inside of it? I think that's the last part of trying to understand this.

Thanks so much!