Single Phase Inverters on 208 3 Phase

[wwhitney]

Makes little difference really, but I was considering 60 degrees on each leg, for a total of 120.

It's still 1/6 of the total time. If you are going to add the 60 degrees together when the current in a coil is in the wrong direction relative to the coil voltage, then for comparison you need to add the 300 degrees together when its in the right direction. So that's 120:600, or 1:5, or 1/6.

Cheers, Wayne

 

[wwhitney]

The primary coils only see the results of actual current flow. When the voltage waveforms opposed each other, the resultant current flow would be reduced and induce a different resultant voltage waveform into the primary. So the primary would not see those smaller opposing voltages, it would only see something less in the greater amplitude wave as it increased toward each peak.

I think you are stating the above as a rebuttal of my statement about transformers, but none of the above is correct.

In a transformer, if you are going to look at it at the level of primary coils and secondary coils, then for the typical case the coils are going to be in 1:1 correspondence, one primary coil for each secondary coil, and one secondary coil for each primary coil. Within each such pair, there will be a turns ratio, say 4:1 for a 480V delta : 208Y/120V transformer. And then further within that pair, the primary coil voltage will be exactly 4 times the secondary coil voltage (and of the same phase), while the primary coil current will be exactly 1/4 of the secondary coil current (and of the same phase).

So if there's a phase difference between the secondary coil voltage and the secondary coil current, the identical phase difference will be seen in the primary coil.

Cheers, Wayne

 

[bellington]

It's still 1/6 of the total time. If you are going to add the 60 degrees together when the current in a coil is in the wrong direction relative to the coil voltage, then for comparison you need to add the 300 degrees together when its in the right direction. So that's 120:600, or 1:5, or 1/6.

I am getting 60 degree on the positive peak and 60 degrees on the negative peak out of 360 degrees. Is my graph off?

Cheers, Wayne

 

[wwhitney]

I am getting 60 degree on the positive peak and 60 degrees on the negative peak out of 360 degrees. Is my graph off?

You're comparing the dashed red line (N-B voltage) with the blue line (A-N voltage), which are 60 degrees out of phase.

But for power transfer in a coil, you want to compare current through that coil with voltage across that coil. The current from our A-B inverter or A-B resistive load is in phase with the A-B voltage, the purple line; it will be scaled differently, but the zero crossings and regions where it's positive or negative will be same.

So if you compare the purple line to either the dashed red line, or to the blue line (the two different coil voltages), then those are only 30 degrees out of phase, which means that they have mismatched sign 30 degrees each half-cycle, or 60 degrees per total cycle.

[For an inverter, the current would actually be (a positive multiple of) the negative of the purple line, but the desired power backfeed would occur when the two sinewaves have opposite sign, so the mismatch is still occurring when the graphed purple sinewave and the dashed red or solid blue sinewaves have opposite sign.]

Cheers, Wayne

 

[ruxton.stanislaw]

To make this discussion even more interesting, I have seen inverters with field adjustable power factor. E.g. +/-0.2 to 1.

 

[ggunn]

A closely related topic that I got thinking about while skimming over this thread: I got thinking about the KW of inverters when run on 208 vs 240. My recollection was that inverters have a "hard" current limit, thus the KW will be reduced at 208 operation. This is indeed the case for SMA sunny boy inverters. Interesting though Solis and fronius inverters provide for more current and the same KW at 208V operation. So it depends.

The data sheet for SMA Sunny Boy inverters is a bit confusing on this matter, and it is inconsistent inverter to inverter.

 

[Carultch]

I have a Geogebra resource that shows visuals for the math behind this concept.

Interphase Load

Interphase Load

www.geogebra.org

Thank you. I appreciate the waveform view better.

To make this discussion even more interesting, I have seen inverters with field adjustable power factor. E.g. +/-0.2 to 1.

I've added a user-adjustable power factor, so you can also see what happens in both views, when an interphase load has a power factor other than unity.

 

[electrofelon]

The data sheet for SMA Sunny Boy inverters is a bit confusing on this matter, and it is inconsistent inverter to inverter.

The spec sheet I looked at seemed clear. What did you find confusing?

 

[ggunn]

The spec sheet I looked at seemed clear. What did you find confusing?

It's been a while since I've looked at the combined data sheet for all the SMA single phase inverters, but the last time I did some inverters showed the same power at both 208V and 240V while others showed the same current at both voltages. Obviously, it cannot be both.

 

[Carultch]

It's been a while since I've looked at the combined data sheet for all the SMA single phase inverters, but the last time I did some inverters showed the same power at both 208V and 240V while others showed the same current at both voltages. Obviously, it cannot be both.

If you are looking at this one:
https://files.sma.de/downloads/SBxx...3.1290147605.1708279129-2105309908.1708279129

Half the models standardize on the same output current, between grid configurations, while the other half has a current limit that is different for each grid configuration. Likely, they build it for a certain AC capacity for each pair of "neighboring" models, but don't use all of it at 240V for the model with less power. It probably has to do with the amount of hardware needed on the DC side to handle the power, and adapting the amount of it, based on what is needed. Like the same circuit board, but populated with fewer capacitors. The power in the part number is only valid for the 240V configurations, since it is either slightly less, or significantly less, for the 208V configuration of the same model. There are two models where it is exactly the same power, and a current limit that is a function of the grid voltage to compensate.

Older sunny boys (early 2010's), had integrated transformers, and user-selectable taps. So given an internal transformer as was standard, it was much easier to build the unit to maintain the same power, among the configuration options. They even had 3 options, 208V, 240V, and 277V, so you could triplex them on either kind of 3-phase grid.

 

[electrofelon]

,some inverters showed the same power at both 208V and 240V while others showed the same current at both voltages. Obviously, it cannot be both.

Sure it can. I understand not liking the inconsistency, but you'd have to take that up with SMA's consistency department

 

[ggunn]

Sure it can. I understand not liking the inconsistency, but you'd have to take that up with SMA's consistency department

What I mean is that an inverter cannot operate at both the same current and the same power at both 240V and 208V. P = VI, and if V changes, either P or I must change. I cannae change the laws of physics, Cap'n!

 

[electrofelon]

What I mean is that an inverter cannot operate at both the same current and the same power at both 240V and 208V. P = VI, and if V changes, either P or I must change. I cannae change the laws of physics, Cap'n!

I don't see anything erroneous in the spec sheet, it just appears that some models are constant AC power and some are constant current (with less power). I thought that's what you are referring to as an inconsistency.

 

[ggunn]

If you are looking at this one:
https://files.sma.de/downloads/SBxx...3.1290147605.1708279129-2105309908.1708279129

Half the models standardize on the same output current, between grid configurations, while the other half has a current limit that is different for each grid configuration. Likely, they build it for a certain AC capacity for each pair of "neighboring" models, but don't use all of it at 240V for the model with less power. It probably has to do with the amount of hardware needed on the DC side to handle the power, and adapting the amount of it, based on what is needed. Like the same circuit board, but populated with fewer capacitors. The power in the part number is only valid for the 240V configurations, since it is either slightly less, or significantly less, for the 208V configuration of the same model. There are two models where it is exactly the same power, and a current limit that is a function of the grid voltage to compensate.

Older sunny boys (early 2010's), had integrated transformers, and user-selectable taps. So given an internal transformer as was standard, it was much easier to build the unit to maintain the same power, among the configuration options. They even had 3 options, 208V, 240V, and 277V, so you could triplex them on either kind of 3-phase grid.

I am hip. My first commercial project was about 640kW on two buildings, one of which was supplied at 480/277V and the other at 208/120V. I used two banks of transformer coupled single phase SB 7000's, phase to neutral at 277V on one building and phase to phase at 208V on the other. That project would have been a helluva lot simpler with today's inverters.

 

[ggunn]

I don't see anything erroneous in the spec sheet, it just appears that some models are constant AC power and some are constant current (with less power). I thought that's what you are referring to as an inconsistency.

Yes, that's what I meant; some are rated one way and some are the other. I didn't mean that I thought the numbers were wrong.

 

[bellington]

I am hip. My first commercial project was about 640kW on two buildings, one of which was supplied at 480/277V and the other at 208/120V. I used two banks of transformer coupled single phase SB 7000's, phase to neutral at 277V on one building and phase to phase at 208V on the other. That project would have been a helluva lot simpler with today's inverters.

With today's inverters, would you prefer to use several banks of single phase (7.6 kW on 240, 6.6 kW on 208) inverters for a 134 kW solar roof coupled across each two legs of 208 Y transformer, or a smaller number of 3 phase inverters coupled to each leg of 208, or even directly coupled to the local 480 grid? Would one arrangement possibly be more efficient than the other?

 

[electrofelon]

With today's inverters, would you prefer to use several banks of single phase (7.6 kW on 240, 6.6 kW on 208) inverters for a 134 kW solar roof coupled across each two legs of 208 Y transformer, or a smaller number of 3 phase inverters coupled to each leg of 208, or even directly coupled to the local 480 grid? Would one arrangement possibly be more efficient than the other?

My choice would be 480 three phase inverters if possible. Three phase 208 inverters are limited in size with the biggest one available being a 25 kw IIRC. Now for 134 KW that would only be say 5 inverters which isnt too bad. 480 would be smaller wire on the AC side which is a big plus.

 

[Bill_F]

Just a completely different thought, but I myself find 3 phase power systems complex.
I want to thank all the contributors for their extreme time; I know I've learned & I'm sure, so have many others.
Bill

 

[BackCountry]

Just a completely different thought, but I myself find 3 phase power systems complex.
I want to thank all the contributors for their extreme time; I know I've learned & I'm sure, so have many others.
Bill

It may seem tricky but it’s just an exponential increase in capacity.

480v all day long, with no neutral to deal with. My preference is SMA.

 

[ggunn]

With today's inverters, would you prefer to use several banks of single phase (7.6 kW on 240, 6.6 kW on 208) inverters for a 134 kW solar roof coupled across each two legs of 208 Y transformer, or a smaller number of 3 phase inverters coupled to each leg of 208, or even directly coupled to the local 480 grid? Would one arrangement possibly be more efficient than the other?

Three phase inverters, absolutely, no question. For a large PV system connected at 208/120V, 480/277 inverter(s) and a transformer is often a better choice than 208/120V inverters since 208/120V native PV inverters are limited in size. I will only connect single phase inverters to three phase services when I have no alternative.