Single Phase Inverters on 208 3 Phase

[jaggedben]

I'm not sure precisely what you are driving at, but if you connect a delta source, e.g., an inverter without a neutral, to a a wye service or transformer secondary with a grounded neutral, the neutral provides the zero V reference for the phase voltages, and the inverter needs to reference it as well, but obviously a delta source has no neutral. In practice a three phase inverter can either have a neutral connection back to the service or transformer, or there can be a grounded "neutral" at the inverter, but obviously not both; running a neutral to a neutral agnostic inverter would be a waste of money, so I don't do it unless a customer wants it for some reason.

I was talking about connecting a delta voltage source to the wye side of a transformer. (Forget about grid tied inverters and current sources.) Do I get a true neutral point?

 

[ggunn]

Well I could stand to be further educated on this. But I think we agree that you don't get the neutral point if you have only one single phase inverter feeding that transformer with no other source present.

You won't connect to the neutral but the neutral of the wye is grounded at the service or the transformer secondary. The phase voltages are referenced to ground at that point, so it doesn't matter than the inverter isn't connected to the neutral as long as the inverter doesn't need it.

 

[jaggedben]

You won't connect to the neutral but the neutral of the wye is grounded at the service or the transformer secondary. The phase voltages are referenced to ground at that point, so it doesn't matter than the inverter isn't connected to the neutral as long as the inverter doesn't need it.

No utility service. I don't care if anything is grounded.

Say I have a delta voltage source (generator, inverter, doesn't matter). It's my only power source. I connect three lines to three windings in a wye configuration, and I don't connect anything to anything on the other side of those windings. Do I get line-neutral voltages that are 1/sqrt(3) of my source line-line voltages?

 

[ggunn]

I was talking about connecting a delta voltage source to the wye side of a transformer. (Forget about grid tied inverters and current sources.) Do I get a true neutral point?

Well, yes. The center of the wye is the grounded neutral. You say "voltage source"; do you mean that as opposed to a current source, as in a grid tied PV inverter?

 

[jaggedben]

I think that if you had two single phase grid-forming inverters connected A-B and B-C, and they had the firmware and connection to talk to each other and sync their voltage waveforms to be 120 degree apart, effectively acting as an open delta source, then depending on the answer to the previous question, you could get a neutral point.

Sorry what's the previous question? For what it's worth I did some math on this a few days ago and I struggled to see how this is possible. It seemed to me the line-neutral voltages would be sqrt(3/4) of the line-line voltages. But I'm not so good at the math as you.

 

[jaggedben]

Well, yes. The center of the wye is the grounded neutral. You say "voltage source"; do you mean that as opposed to a current source, as in a grid tied PV inverter?

Grounding has nothing to do with the question.

Yes, a voltage source is not a grid-tied inverter. I mean a generator or an off-grid inverter or something.
I guess if your answer remains the same I'll take your word for it?

 

[wwhitney]

Say I have a delta voltage source (generator, inverter, doesn't matter). It's my only power source. I connect three lines to three windings in a wye configuration, and I don't connect anything to anything on the other side of those windings. Do I get line-neutral voltages that are 1/sqrt(3) of my source line-line voltages?

You certainly do, because (a) that wye-center point has some voltage and (b) by symmetry the only voltage it can be is the neutral point.

The question I have is if you then connect a load L-N, would the "N" voltage stay at that center point, or would it be shifted by the presence of the load.

The analogous single phase question is if you just have 2 conductors L-L (say 240V) and you connect them to the outer terminals of a center-tapped autotransformer, can use use the center tap as a neutral point and supply L-N loads with it. Here the answer is yes, as all the windings are on one 2-leg core, so the 240V applied L-L defines a fixed volts/turn on that core, and the center tap must be at midpoint voltage.

That doesn't seem to carry over directly to the 3-phase case at hand, as we'd have a 3 or 5 leg core, with each winding on a different leg.

Cheers, Wayne

 

[wwhitney]

Sorry what's the previous question?

Whether supplying one side of a wye transformer configuration with a 3W delta source gives you a neutral with a fixed voltage.

My point was that two communicating single phase 2-wire inverters connected A-B and B-C are theoretically capable of making an open delta source.

 

[bellington]

Let me try this. If we use a 3-phase inverter connected to the same 208/120 Y transformer on legs A, B, C and neutral, then the A phase inverter is injecting a 120 volt sine wave in phase with A-N. The B phase inverter is injecting a 120 volt sine wave 120 degrees behind A, and in phase with B-N. The C inverter is injecting a 120 volt sine wave 240 degrees behind A, and in phase with C-N. There is current flow on N, but the sum of all currents is 0.

I'm not sure such an inverter system would work on a delta connection.

Am I still missing something in the above configuration?

 

[wwhitney]

Whether supplying one side of a wye transformer configuration with a 3W delta source gives you a neutral with a fixed voltage.

My current thoughts:

Again, this is for a 3-wire 3 phase delta source, hooked up to H1, H2, H3 of a transformer with a wye primary configuration (secondary configuration unspecified), and the question is if H0 will give you a stable neutral point you can use to supply loads.

My current thinking is sometimes yes. If the secondary configuration is a delta, then that configuration of coils imposes the requirement that the the 3 secondary voltages (as phasors) sum to zero. Which requires that the 3 primary coil voltages sum to zero. Which forces H0 to have voltage the average of H1, H2, and H3. [Literally the coordinate by coordinate average if we fix a zero voltage point and write H1, H2 and H3 in coordinates.]

If the secondary configuration is a wye, then we don't have that constraint from the secondary. But if the transformer is a 3 phase transformer with a 3 leg core, then the core configuration imposes the constraint that the sum of the coil voltages is zero. So again H0 is fixed at the average voltage of H1, H2, and H3.

But if the secondary configuration is a wye, and we are using 3 separate single phase transformers, or a 3 phase transformer with a 5 leg core, then we don't have any such constraint. So I believe adding a load H0 to one leg would shift the voltage of H0, based on the load impedance and the coil impedances.

Cheers, Wayne

 

[jim dungar]

No utility service. I don't care if anything is grounded.

Say I have a delta voltage source (generator, inverter, doesn't matter). It's my only power source. I connect three lines to three windings in a wye configuration, and I don't connect anything to anything on the other side of those windings. Do I get line-neutral voltages that are 1/sqrt(3) of my source line-line voltages?

Yes.
This is the premise for most ground detection schemes utilizing "light bulbs".

 

[wwhitney]

Let me try this. If we use a 3-phase inverter connected to the same 208/120 Y transformer on legs A, B, C and neutral, then the A phase inverter is injecting a 120 volt sine wave in phase with A-N. The B phase inverter is injecting a 120 volt sine wave 120 degrees behind A, and in phase with B-N. The C inverter is injecting a 120 volt sine wave 240 degrees behind A, and in phase with C-N.

[Side note: the phrase "3-phase inverter" does not imply a wye-connected inverter, but your description is of a wye-connected inverter. ]

You need to clarify whether you are talking about the "grid up" condition, when the other side of the 208Y/120V (notice the naming convention) transformer is connected to a 3 phase utility or a 3 phase grid forming inverter, and your inverter is grid-following, or the "grid down" condition, when your 3 phase inverter is grid forming itself.

For the "grid up" condition, your grid-following inverter does not inject voltage at all! It follows the voltages it sees, and it injects currents.

For the "grid down" condition, your description is basically correct for a 3 phase grid forming inverter. The part about "in phase with A-N," etc is true by definition, as the inverter is the only voltage source present, and a voltage is always in phase with itself.

I'm not sure such an inverter system would work on a delta connection.

For the "grid up" condition, your 3 phase inverter could be delta connected and work equally well as being wye connected. Maybe simpler, as you only need 3 wires.

For the "grid down" condition, this is the question we've been discussing recently, whether connecting a delta source to a transformer wye would give you a stable neutral point voltage you can use for loads.

Cheers, Wayne

 

[jaggedben]

Let me try this. If we use a 3-phase inverter connected to the same 208/120 Y transformer on legs A, B, C and neutral, then the A phase inverter is injecting a 120 volt sine wave in phase with A-N. The B phase inverter is injecting a 120 volt sine wave 120 degrees behind A, and in phase with B-N. The C inverter is injecting a 120 volt sine wave 240 degrees behind A, and in phase with C-N. There is current flow on N, but the sum of all currents is 0.

Assuming the three-phase inverter injects balanced current (as a grid-tied inverter would) it causes no neutral current flow. But a resistive load would draw that current in phase with the inverter.

I'm not sure such an inverter system would work on a delta connection.

Grid-tied it would work.
Off-grid you would not be able to draw neutral current to a load if you didn't hook up the neutral. Obviously. ;-)

 

[jaggedben]

Off-grid you would not be able to draw neutral current to a load if you didn't hook up the neutral. Obviously. ;-)

However, if you connected your off-grid wye inverter to a delta-wye transformer, L-N loads on the wye side would have no problem. That's done all the time with a wye utility service.

 

[wwhitney]

However, if you connected your off-grid wye inverter to a delta-wye transformer, L-N loads on the wye side would have no problem. That's done all the time with a wye utility service.

Assume you meant "off-grid delta inverter" to the "delta side" of a delta-wye transformer.

Cheers, Wayne

 

[jaggedben]

Assume you meant "off-grid delta inverter" to the "delta side" of a delta-wye transformer.

Cheers, Wayne

Yes that's what I meant. (But wait, according to your post #212 it should work okay the other way too. ;-) )

 

[BackCountry]

Yes that's what I meant. (But wait, according to your post #212 it should work okay the other way too. ;-) )

This may hold a record for the most active post ever, only second to the debate of pre twisting when installing a wire nut.

 

[winnie]

This is an aside from the OP's scenario: say you have an (ungrounded for simplicity) 480V delta 3W service from the utility, and you have a transformer with a 480Y/277V primary, so you hook up your 480V 3W delta to H1-H3, and leave the secondary disconnected. Can we use H0 as a neutral point and power a 277V load using it? And does the answer depend on whether the secondary is a wye or delta?

Something I need to think about, but not relevant for the OP.

Cheers, Wayne

A stand alone Wye does not provide a low impedance neutral.

A 'zig-zag' type Wye does derive a low impedance neutral.

My understanding is that you can use a Wyeelta transformer to derive a low impedance neutral; any imbalance on the Wye legs shows up as circulating currents on the Delta side, even if nothing is connected.

I believe that Wyeelta:Wye transformers (with the tertiary delta not connected to anything external) are used in some situations.

-Jon

 

[wwhitney]

Wyeelta transformer

Apparently you need some white space if you don't want the forum software to render the : D (without white space) as a smiley face. So Wye : Delta transformer.

Your comments are compatible with my thoughts in post #210, but that doesn't mean necessarily post #210 is correct. Comments on post #210?

Cheers, Wayne

 

[jaggedben]

This may hold a record for the most active post ever, only second to the debate of pre twisting when installing a wire nut.

Nah, that would be this thread. Which, actually, contains references to another, possibly longer thread on the same subject that was hidden.
There have already been references to that thread in this one. Thank god we are discussing three-phase power in this one so (almost) all the phase angles are other than 180deg.