3-phase AC Questions

[Fried Chicken]

I'm watching Mehdi's excellent explanation of 3-phase power:

https://youtube.com/watch?v=quABfe4Ev3s

At 0:19 he introduces 3-phase power, and at 1:50 he underscores the point that a wire back to the source isn't needed.

Then he explains 3-phase is what power grids use in Canada to transmit power

At 3:50 he shows an outlet where the receptacles are wired with two separate phases out of the three phases, so instead of reading 240v he reads 203volts

at ~7:40 he explains "A 3-phase AC motor is a balanced load, so with only 3 wires it can output 3 times the power compared to a single phase AC or DC motor at the same current draw from every wire".


This is all well and good, but... #1: Does the U.S. not do 3-phase AC, or should I sometime actually read the output across my breaker panel?  I thought we have 240v split phase?  Or are Canadian houses wired for 3-phase where U.S. houses are 2-phase?

More intriguingly is the advantage to 3-phase AC as he presents it: the sum of the phases equal zero; no return line is needed, and each wire can carry a load.  How's that different from split-phase AC?  If I have 120v on one leg, and -120v on the other leg, they all equal out just like 3-phase AC, no?  No return line is needed (nor is there one).  Am I missing something here?

What is the advantage of 3-phase AC?

[helius]

The electric power used in the US and Canada is exactly the same. There are different regulatory bodies and model codes, but they use the same standards and quantities. You should already have been able to guess that based on the electric outlet systems being exactly the same.

From eaton.com:

The Canadian Electrical Code (CEC) is largely the same as the NEC, although the two electrical standards are not identical in every detail.

[microwhat]

Take that video with a grain of salt.

The U.S. has 3 phase power everywhere. Almost all utility-scale distribution is done with 3 phase.

Unless you live in an apartment building large enough that the developer wanted to pay the utility for 3 phase service, or if you have very deep pockets, you will have "split-phase" service (NOT two-phase), where you will measure 240VAC nominal between L1 and L2.

The neutral ("wire back to the source") is not needed when all loads are connected phase to phase, or all loads are exactly balanced. This is very rare, so it's common for utility distribution to include the neutral. There do exist delta-connected transformers and services. These don't have a neutral point, and are only used for commercial and industrial systems. You won't find this in a house, and to my understanding, most utilities try to replace these with wye-connected services whenever they can.

3 phases is the smallest number that allows a rotating magnetic field without needing capacitors. Thus, you can build a motor that doesn't need help starting. More phases will certainly work, but cost more. I've heard that there may still be some real two-phase services left in some areas of the east coast. These had four phases present, which could run a motor without capacitors, but (I presume) one had to pay closer attention to phase rotation. With two-phase, switching two phase leads could cause the motor to not start, and probably catch fire, where with 3-phase, the motor just runs the other direction.

he explains "A 3-phase AC motor is a balanced load, so with only 3 wires it can output 3 times the power compared to a single phase AC or DC motor at the same current draw from every wire".

He's conflating a 120VAC circuit with two wires, with a 240VAC "split-phase" circuit with 3 wires, with a 3-phase circuit with 3 or 4 wires. Adding one extra wire does not carry 3 times the power with the same current in each wire.

[Renate]

The real advantage of 3 phase is that it has direction.

Your normal residential 240V/120V has three wires and you might say it has two phases as when one "phase" is positive, the other is negative.
But it's not at all like 3 phase.

3 phase came along with Tesla and the idea of a rotating magnetic field (although he was originally thinking of two phase, but that's another story).
3 phase motors are advantageous since the original turning of the generator translates to the turning of the motor.
If a power station reversed the turning of the 3 phase (for instance, by swapping two of the three wires) all hell would break out with motors in every factory going backwards.
With residential power nothing would happen because there's no "direction".

Nowadays many big motors are no longer using 3 phase directly but using variable frequency drives.
So the role of power companies devolves to just supplying power devoid of "direction".
So we have things like high voltage DC transmission lines.
But we've still got a long way to go before we're done with 3 phase.

[rteodor]

[...]
at ~7:40 he explains "A 3-phase AC motor is a balanced load, so with only 3 wires it can output 3 times the power compared to a single phase AC or DC motor at the same current draw from every wire".
[...]
What is the advantage of 3-phase AC?

There are many possible ways - read multi-phase configurations - that can transfer energy but 3-phase is the most optimal in economical and technical terms. This is what I was told in schooool.

[IanB]

This is all well and good, but... #1: Does the U.S. not do 3-phase AC, or should I sometime actually read the output across my breaker panel?  I thought we have 240v split phase?  Or are Canadian houses wired for 3-phase where U.S. houses are 2-phase?

You see may 240 V split phase in your house. But that does not mean there is not a three phase supply upstream where you can't directly see it. The local substation where your electricity comes from definitely has three phase power in it.

More intriguingly is the advantage to 3-phase AC as he presents it: the sum of the phases equal zero; no return line is needed, and each wire can carry a load.  How's that different from split-phase AC?  If I have 120v on one leg, and -120v on the other leg, they all equal out just like 3-phase AC, no?  No return line is needed (nor is there one).  Am I missing something here?

Yes, you are missing something. With single phase power, one wire carries the current out, and the other wire carries the current back. So there is no equaling out. Both out and return wires carry the full load current.

With a three phase supply, three wires carry the current out, and no wires carry the current back. There is no return wire. This results in much less copper (aluminum) required for the transmission cables, and much lower losses in transmission.

You normally think of a circuit needing out and return to complete the circuit. But with three phase circuits, the three outbound currents cancel each other out, so no return is needed. It's like magic, like having a circuit that only needs one wire instead of two (or in reality, it needs three wires instead of four).

What is the advantage of 3-phase AC?

Not only the efficiency noted above, but electric motors require three phase supplies to work efficiently. Any AC electric motor bigger than a fraction of a horsepower needs a three phase supply (at least traditionally, without modern electronics). The three phases are needed to transmit the rotation of the generator to the rotation of the motor. Single phase supplies do not carry rotation, so they cannot make motors turn without horrible hacks.

[Siwastaja]

All large-scale power transmission everywhere is 3-phase. Almost all small loads are 1-phase. You get 1-phase by simply branching off from the 3-phase supply. Where this branching exactly happens varies per country and tradition. Mehdi shows all three phases being transformed to low voltage and supplied to his house; same here. For you, branching to single phase is simply done further upstream, probably so that the transformer which steps down the HV to low voltage already takes single phase HV in.

It doesn't make a huge fundamental difference; just tradition / arbitrary choice.

[TizianoHV]

What is the advantage of 3-phase AC?

The three phases are needed to transmit the rotation of the generator to the rotation of the motor.

Two phases allows rotation too. I've been told that three phases is the minimum number of phases needed to transmit constant power. With one and two phases the transmitted power is oscillating (not good for motors/generators...).

[langwadt]

All large-scale power transmission everywhere is 3-phase. Almost all small loads are 1-phase. You get 1-phase by simply branching off from the 3-phase supply. Where this branching exactly happens varies per country and tradition. Mehdi shows all three phases being transformed to low voltage and supplied to his house; same here. For you, branching to single phase is simply done further upstream, probably so that the transformer which steps down the HV to low voltage already takes single phase HV in.

It doesn't make a huge fundamental difference; just tradition / arbitrary choice.

I seem remember  Mehdi showing that at his house he had two out of three phases, so 120V and 208V, not split phase 120V 240V

[Siwastaja]

Two phases allows rotation too. I've been told that three phases is the minimum number of phases needed to transmit constant power. With one and two phases the transmitted power is oscillating (not good for motors/generators...).

Yes, two phases (90 deg phase shift) is still used in stepper motors and cheap/small BLDC fan motors. That's fine. "Constant power" "oscillation" explanation is probably your misinterpretation of the actual problem of 2-phase: the neutral wire carrying a lot of current, so a thick conductor is needed. 3-phase delivery on balanced currents (like in motors) works with the same number of wires as 2-phase: still just 3 wires, so the third phase is "free", and all wires carry the same current. That's the advantage.

Control systems still regularly transform the 3-phase math (120deg phase shifts) into 2-phase math (90 deg phase shift), and after calculations back to 3-phase, because 2-phase math is intuitively so much easier to understand for humans: X and Y axis 90deg apart, sine and cosine, obvious rotation.

Split phase doesn't enter the picture anywhere. It's just single phase, with an earthed center tap.

[synccond]

There are advantages to three phase, then there is the reason for three phase.
Advantages are what was stated above: less wiring for a given VA load and guaranteed direction (forward/reverse).
But the actual driving force for WHY three phase has been firmly adopted around the world is that instantaneous power (therefore also torque) of a three phase induction motor is a constant.  For single phase and two phase motors, it is a pulsating torque.  Try running a 5 hp motor on single phase and you will hear and feel the pulses; long term the motor struggles to survive.  Now try single/two phase on a 500hp motor and watch it completely fail both in application as well as lifetime.

This goes all the way back to Tesla; once he produced a reliable three phase induction motor, the adoption of three phase was sealed for AC systems.  The victory over DC was related, but also due to other issues like transmission advantages.

One could just as well design 5 phase and 6 phase systems, which actually do exist.  But the premise is 3 phase solves the biggest issue, and is good enough.  Constant torque is the goal.

[Siwastaja]

WHY three phase has been firmly adopted around the world is that instantaneous power (therefore also torque) of a three phase induction motor is a constant.  For single phase and two phase motors, it is a pulsating torque.

Not true. You forget that the waveforms are sinusoidal; theoretical point source magnetic force (relative to current) can produce perfectly smooth torque from the minimum of 2 phases, 4 poles total; it's good old sine and cosine: one of the electromagnets is gradually reducing its force, while the other one is increasing. Between the poles, both electromagnets are contributing to the force at the same time.

So there's nothing pulsating or wrong in 2-phase, that's exactly why when we implement FOC on 3-phase motors, we mathematically transform the measurements to two phases, and then the controls back to 3 phases; nothing is lost. No pulsating torque anywhere; torque is exactly controlled by a constant setpoint in a rotating frame of reference.

Small gains in torque ripple can be had by adding phases, because real motors can't have perfect iron shape. 3 phases is a modest improvement from already very good and smooth baseline. 6-phase motors exist in large end of machines.

The evidence is the ubiquitous capacitor-run single phase (i.e., two-phase) motors, which don't even have the luxury of exact 90deg phase shift (a capacitor cannot do that), yet good motors provide very smooth operation, with 2 phases. With 1 phase, that's impossible, though - there is no rotating field, control is 1-dimensional, so torque pulsates between maximum and exactly zero, and nothing can be done about it, except give starting torque somehow and add inertia to make the rotor pass the deadzone of no torque.

But 3-phase, 2-phase or any-phase where n>1 and phases define a 2D plane, perfectly smooth, constant torque can be had (in theory).

Really, the reason for 3-phase is the clever balancing of currents, so that an extra return conductor (neutral) is not needed for a motor. It's not a small optimization, but pretty big one. Naturally increased pole count and modest torque ripple reduction is an added benefit; a 4-pole pair per phase (1500 / 1800 rpm) 3-phase motor has 12 physical poles in it, which seems to be some sort of sweet spot in medium-sized motors. A larger motor might have 36 physical winding slots in its stator, which means one pole shares 3 adjacent slots; which is slightly non-optimal for torque ripple; 9-phase motor would use one slot per one winding. No big deal.

[IanB]

I seem remember  Mehdi showing that at his house he had two out of three phases, so 120V and 208V, not split phase 120V 240V

Some buildings with a large number of occupants may have a three phase supply transformer. The occupants may then get two hot phases and the neutral at their breaker panel instead of the more traditional split phase arrangement. In that case, the voltage phase to neutral will still be 120 V, but the voltage phase to phase will be 120 x 1.73 = 208 V instead of 240 V.

Most items expecting 240 V will still work, but things like water heaters will run at reduced power due to the lower voltage.

[synccond]

Not true.

Absolutely true.

https://relcoelectricals.com/why-3-phase-systems-are-preferred-over-single-phase-in-electrical-engineering/

[johansen]

Not true.

Absolutely true.
https://relcoelectricals.com/why-3-phase-systems-are-preferred-over-single-phase-in-electrical-engineering/

not everythign on the internet is correct and that website doesn't even mention two phase.


two phase (90 degree phase shift, not 180) produces constant torque, its used in every "single" phase induction motors via a capacitor shifted aux winding located at 90 degrees, and most stepper motors (tormach likes 3 phase stepper motors). (the cogging is torque ripple due to imperfect geometry and reluctance and saturation)

but 2 phase still requires 3 wires, one of which carries 1.41 units of current.

or it requires 4 equal size wires.

who wants to deal with that?

[Siwastaja]

Not true.

Absolutely true.

https://relcoelectricals.com/why-3-phase-systems-are-preferred-over-single-phase-in-electrical-engineering/

So random links give your false claim credibility?

Why didn't you read it? No mention of 2-phase power at all. Completely irrelevant to your claim of 2-phase torque ripple.

Come back with a source that explains why 2-phase power causes significant torque ripple, and we'll talk again.

Or, admit your honest and simple mistake, accidentally writing 2-phase when you meant just 1-phase, and stop making yourself look like a fool. We all make typos.

It's easy to make proper comparison of 1-, 2- and 3-phase power delivery. 2-phase system fixes the pulsating power by being 2D instead of 1D, but adds large "pulsating" neutral current. It's the third phase which keeps the 2D feature - it doesn't add any degree of freedom, the third phase is redundant, but balances the current between the three conductors. Way better conductor utilization in motors, simple as that.

And, as we all know, split phase is 1-phase, and as such, 1D, hence pulsating power and zero current delivery in certain rotor position.

[Freesurfer]

In Norway there are many places where the utility will provide you with 3 live phase wires, no neutral and no earth. The earth is done locally in such a way that everything internally in the building, is bonded like normal to an earth bar in your breaker board etc., then for the reference to the real ground, there are earth rods and/or buried bare wire around the building. But none of this connects back to the utility in any way.

For the actual three phase wires, you will measure: L1-L2 = 230V, L2-L3 = 230V, L3-L1 = 230V. If you measure from any phase wire to your ground rod, it should read approximately 133V.


So why would we do it like this? Well, one advantage is that when you have 3 wires running to your property with 120 degrees spaced phases, it is possible to max out the current delivery capabilities of every wire at once.

Suppose you have a mains cable that is capable of delivering 50 amps on each wire.
If you use 2 of the 3 wires, that's simply 230V * 50A = 11.5kW
Now add in the third wire, the calculation will be 230V * 50A * square root of 3 = ~20kW.
So as you can see, adding just a third wire, increases the power delivery by ~73%, whilst only using 50% more copper in the cable.

Doing the same thing in a split phase doesn't "do" anything.
Imagine you have just the 2 live wires, and can only use 240V in your house. That's 240V * 50A = 12kW. Now, you add in the third wire for your neutral. What does the neutral do in terms of power? Nothing. It only allows you to split that 240V down to 2x 120V and take the imbalance of your loads. It doesn't add anything to your total power budget. So that's 50% more copper in your mains cable doing "nothing". Any wire can of course deliver the full 50A, but not all 3 wires at once. It will always be split like 50/25/25, 50/0/50, 50/50/0 etc. Never 50/50/50.


(Note that these calculations are for resistive loads. I am not ruling out the possibility that some phase shifted loads could divide the current differently across the 3 wires in a split phase. But that still doesn't add anything in terms of real power)

[jahonen]

Quadrature arrangement will produce well defined rotating field with constant power transfer thus constant torque, but it uses the same amount of wiring than a three phase system. Because three phase system can transfer more power with same amount of conductor metal, as all three line conductors are contributing to the power transfer and there is no need for any return conductor (for true three phase load without neutral), it is a very obvious choice over the quadrature arrangement. As shown in previous post.

Regards,
Janne