May I ask a question about the single vs two phase stuff

[Besoeker]

Define 'phase'.

Wave forms with a displacement in time or angle between the vector representation. Would that do?

 

[winnie]

N is grounded
it IS the reference or ground plane
not by choice

Small disagreement:

N is grounded. Using ground as the reference is generally the most sensible and common choice.

But it is still a choice.

I'd propose: if two terminals of a system show a phase difference when measured relative to a third terminal, and if that phase difference can be made to disappear by appropriate selection of reference terminal, then those are hemiphases, not true separate phases. Not saying this is right, just suggesting it.

-Jon

 

[winnie]

no one is saying this is a 2 phase power system
only that split phase can be resolved into 2 phases
one is a 90 deg shift, the other 180
if the shift was 170 is that 2 ph? 179? 181?

I agree; no one is trying to claim that split phase power is somehow the equivalent of a true '2 phase' system with 90 degree phase displacement.

I was trying to say that a 180 degree phase shift is qualitatively different from 90 degrees, 160 degrees, 179 degrees, etc.

If I had a supply consisting of three terminals; grounded G, 'hot' H1 and 'hot' H2, and there was a 160 degree phase difference between H1-G and H2-G, then with proper transformers I could generate a 90 degree phase difference.

But if the supply phase difference was 180 degrees then I could not do this.

In principal, with _perfect_ components I could do this starting with a 179 degree phase difference. Reality intrudes so there is going to be a continuum rather than a sudden discontinuity right at 180 degrees.

But the point is that a 180 degree phase difference is _different_ than another arbitrary phase di

 

[jaggedben]

N is grounded
it IS the reference or ground plane
not by choice

I simply don't agree. It's a choice. I can set up my scope to measure waveforms however I want. For the purposes of this discussion it doesn't matter if N is grounded, that has no effect on the waveforms or what the scope shows. We could ground L1 or L2 on a split phase system, it wouldn't be code but it wouldn't change the waveforms or voltages between the nodes either.

The bottom line for me is that, for the person trying to understand how the system works, I believe it is more illuminating to show how the graph changes when you flip the probes than to argue over which way to set the probes is the 'right' way.

 

[Besoeker]

Small disagreement:

N is grounded. Using ground as the reference is generally the most sensible and common choice.

Grounded or otherwise, it is the common point of a 120-0-120 system. That alone would make it the logical choice.
After all, if you are trouble shooting an electronic PCB you would use the zero volts as the common reference.

 

[jaggedben]

Wave forms with a displacement in time or angle between the vector representation. Would that do?

No, I think it is a combination of two wires in an AC system that are different colors. Remember, I'm an electrician not an engineer.

In all seriousness though, I think there's a flaw in your defining 'waveforms' (plural) as a 'phase' (singular).

 

[Ingenieur]

I agree; no one is trying to claim that split phase power is somehow the equivalent of a true '2 phase' system with 90 degree phase displacement.

I was trying to say that a 180 degree phase shift is qualitatively different from 90 degrees, 160 degrees, 179 degrees, etc.

If I had a supply consisting of three terminals; grounded G, 'hot' H1 and 'hot' H2, and there was a 160 degree phase difference between H1-G and H2-G, then with proper transformers I could generate a 90 degree phase difference.

But if the supply phase difference was 180 degrees then I could not do this.

In principal, with _perfect_ components I could do this starting with a 179 degree phase difference. Reality intrudes so there is going to be a continuum rather than a sudden discontinuity right at 180 degrees.

But the point is that a 180 degree phase difference is _different_ than another arbitrary phase di

it is arbitrary to say for a phase diff to exist it must spin a motor

 

[Besoeker]

it is arbitrary to say for a phase diff to exist it must spin a motor

I agree. That's simply an irrelevance.

 

[Ingenieur]

480:120/240 xfmr
are the sec coils using the dot convention wired 'in phase' or in opposition, ie out of phase?

 

[Besoeker]

480:120/240 xfmr
are the sec coils using the dot convention wired 'in phase' or in opposition, ie out of phase?

It's usually a centre tapped continuous winding with the centre tap as the common (N) and the opposite ends being L1 and L2 and this comes about the 180° displacement between them.
I think that's as simple as I can make it.

 

[wwhitney]

Using the same two terminals? A change in polarity.

[I understand the phrase "change in polarity" to mean "multiply by -1".]

OK, then that's all you have with L1-N versus L2-N in 120V/240V split phase.

I think we can all agree that L1-N is in phase with N-L2, no phase difference. Reverse the terminals on the second measurement, and you have just introduced a change in polarity.

Cheers, Wayne

 

[wwhitney]

I'd be willing to settle for "degenerate" phase difference. That is entirely related to Golddigger's and Winnie's points about independent vectors/spinning motors.

Cheers, Wayne

 

[Besoeker]

[I understand the phrase "change in polarity" to mean "multiply by -1".]

Polarity doesn't really pertain to alternating quantities.

 

[wwhitney]

Polarity doesn't really pertain to alternating quantities.

Sure it does. If it didn't, as I commented earlier, you could arbitrarily multiply by -1. That is what the phrase "polarity doesn't matter" means to me.

Cheers, Wayne

 

[jumper]

C.P. Steinmetz was one of the founding fathers of AC systems and a recognized genius.

He recognized that there were two phases present even if you could reduce the system of two smaller phases to a larger single phase system by taking one phase to be the negative of the other.

I don't recall the exact wording but I have his paper somewhere in my office.

General Lectures on Electrical Engineering?

 

[Ingenieur]

General Lectures on Electrical Engineering?

condensed version of relative portion
https://www.fsb.unizg.hr/usb_frontend/files/1508338227-0-elte_predavanje_03-dodatak.pdf

a split ph xfmr typ has 2 sec coils
they are connected in series (dot wise), low-dot-low-dot
one dot is a line hence the other dot is grounded
opposite phasing and hence polarity
it is not how you reference it
they are PHYSICALLY different
wound opposite hand relative to the primary and common ground reference

 

[GeorgeB]

It's usually a centre tapped continuous winding with the centre tap as the common (N) and the opposite ends being L1 and L2 and this comes about the 180° displacement between them.

Secondary is usually separate coils; in series for 240, in parallel for 120. If one gets the dots backwards in series, "zero" output. If backwards in parallel, effectively shorted output, much heat produced accidentally if the OCP doesn't take care of it.

Indeed, most of the time, there are 2 primary coils with the "power" transformers for 240/480:120/240.

 

[Besoeker]

Sure it does. If it didn't, as I commented earlier, you could arbitrarily multiply by -1. That is what the phrase "polarity doesn't matter" means to me.

Cheers, Wayne

I disagree. That's life.

 

[jaggedben]

condensed version of relative portion
https://www.fsb.unizg.hr/usb_frontend/files/1508338227-0-elte_predavanje_03-dodatak.pdf

a split ph xfmr typ has 2 sec coils
they are connected in series (dot wise), low-dot-low-dot
one dot is a line hence the other dot is grounded
opposite phasing and hence polarity
it is not how you reference it
they are PHYSICALLY different
wound opposite hand relative to, the primary and common ground reference

Wait, what? :happysad:

They are wound the same hand relative to the primary. Dot is in the same direction for both relative to primary. Otherwise you don't get 240V.

 

[GeorgeB]

Would you just try to think of what current actually is; it's clearly the work that's being performed.

I can have very high currents with a capacitor or inductor as load with substantially no power consumed, indeed, if an ideal device, zero power consumed.