Safe Grid Use of the 5000ES and transformer

[timselectric]

Are those wires from the inverter?

Autotransformer conductors.

 

[timselectric]

Playing with my new 3-phase transformer, primary reconfigured as WYE and just L1, L2, L3, N of source connected, I was surprised to see N current about the sum of three phases. I didn't like that, so I opened the neutral. That'll teach it! Now, zero current on the neutral (of course) and reduced no-load phase current by a factor of four.

3 phase changes the math a bit.
I've never done a primary WYE setup before. (Never had any reason to in the field)
I would assume that it could help balance the individual phases on the secondary.
But the losses would be high.
I could see where it would be beneficial if sensitive equipment was connected to the secondary.

 

[Hedges]

Autotransformer conductors.

Current measurements as running.
N = 4.35a
L1 = 1.86a
L2 = 2.65a

What are the currents with no load?

With the above load, you could also measure (L1 and N?) of load, (L1 and L2, there is no neutral) of source.

N is only 0.16A less than sum. I don't understand why L1 and L2 are so different.

3 phase changes the math a bit.
I've never done a primary WYE setup before. (Never had any reason to in the field)
I would assume that it could help balance the individual phases on the secondary.
But the losses would be high.
I could see where it would be beneficial if sensitive equipment was connected to the secondary.

My thinking is that that the three 120 degree vectors should draw current balancing each other, so I expected zero amps on WYE neutral (so long as not loaded imbalanced.) Applied voltages were unequal, 125V from grid through Sunny Island master to L1, 120V from Slaves (says so on display, although SMA manuals now use more sensitive language) which generate the two missing phases.

I didn't think it would impose balance. Although, this transformer is three (primary + secondary) windings on an E core, is not three separate transformers.

I also have a "Zig-Zag" transformer I bought to establish neutral of a delta secondary.

Why high losses? I haven't figured out where current goes or power is dissipated. Neutral ~ sum of phase currents indicates currents in phase, not out of phase. Oh, stuffed back into source? (just like your auto-transformer.)

A delta auto-transformer would also be possible. Taps connected as originally intended would do that, probably not the voltages I'm looking for. Took me a while to find this delta I could convert to WYE and do what I wanted.

 

[esposcar]

What are the currents with no load?

With the above load, you could also measure (L1 and N?) of load, (L1 and L2, there is no neutral) of source.

N is only 0.16A less than sum. I don't understand why L1 and L2 are so different.



My thinking is that that the three 120 degree vectors should draw current balancing each other, so I expected zero amps on WYE neutral (so long as not loaded imbalanced.) Applied voltages were unequal, 125V from grid through Sunny Island master to L1, 120V from Slaves (says so on display, although SMA manuals now use more sensitive language) which generate the two missing phases.

I didn't think it would impose balance. Although, this transformer is three (primary + secondary) windings on an E core, is not three separate transformers.

I also have a "Zig-Zag" transformer I bought to establish neutral of a delta secondary.

Why high losses? I haven't figured out where current goes or power is dissipated. Neutral ~ sum of phase currents indicates currents in phase, not out of phase. Oh, stuffed back into source? (just like your auto-transformer.)

A delta auto-transformer would also be possible. Taps connected as originally intended would do that, probably not the voltages I'm looking for. Took me a while to find this delta I could convert to WYE and do what I wanted.

It sounds like you are describing a three-phase power system with a WYE connection, where the three 120 degree vectors should ideally draw current that balances each other. However, in practice, there may be imbalances in the system due to unequal voltages or loads.

If the voltages applied to the WYE system are unequal, then there may be a current imbalance that can cause high losses. This is because the phases will not draw equal amounts of current, which can lead to an increase in power losses and potentially cause damage to the system.

One possible solution to establish a neutral for a delta secondary is to use a Zig-Zag transformer. This type of transformer can provide a neutral connection by using two coils wound in opposite directions. However, it's important to ensure that the transformer is properly rated for the voltage and current requirements of the system.

As for the high losses, it's possible that some of the current is being dissipated as heat in the transformer, which can cause losses and reduce efficiency. It's also possible that some of the current is being reflected back into the source, which can cause problems with the power quality of the system.

 

[Hedges]

It sounds like you are describing a three-phase power system with a WYE connection, where the three 120 degree vectors should ideally draw current that balances each other. However, in practice, there may be imbalances in the system due to unequal voltages or loads.

If the voltages applied to the WYE system are unequal, then there may be a current imbalance that can cause high losses. This is because the phases will not draw equal amounts of current, which can lead to an increase in power losses and potentially cause damage to the system.

Yes, my source is 3x 120V inverters, 120 degrees apart.
I initially drove 200V delta primary with the 208V, but that has higher idle current than I like (OK on-grid, to be avoided in an off-grid system.) With a typical transformer, I prefer to operate around half of rated voltage.

There was an imbalance in voltage, 125V vs. 120V.

One possible solution to establish a neutral for a delta secondary is to use a Zig-Zag transformer. This type of transformer can provide a neutral connection by using two coils wound in opposite directions. However, it's important to ensure that the transformer is properly rated for the voltage and current requirements of the system.

That was the path I started down earlier. Wanted to step 208 up to 480. Used a 480 delta to 120/208Y transformer backwards. Didn't need neutral for my "load" (a 3-phase inverter), but I did need to ensure L1/L2/L3 didn't go significantly more than 277V above ground (which could happen with one line shorted to ground, and no current so no breaker tripping.)

I got a 600V 16A zig-zag transformer. Even at 480V, it drew more current than I preferred. Maybe due to that same phase voltage imbalance? (I forget if I even connected neutral. Probably not.)

I expected a zig-zag transformer to attempt balancing phase voltages. I did not expect a WYE transformer to do that.

The one I got was big enough to provide some 277V imbalanced loads, but not the full 30 kVA available. My intention was to run its neutral through a breaker, and gang it to breaker feeding primary.

My SPICE of zig-zag shows it works, but thus far I don't have a model that carries anywhere near the 16A this one is rated for. Just a dummied model, not actually based on measurements of this transformer. My attempts at measuring larger commercial transformers to make saturable hysteretic models have not gone well. Small toroids and chokes, I've been more successful.

As for the high losses, it's possible that some of the current is being dissipated as heat in the transformer, which can cause losses and reduce efficiency. It's also possible that some of the current is being reflected back into the source, which can cause problems with the power quality of the system.

Some combination of those in my case. My inverter is able to do four quadrants. I wouldn't be surprised if HF inverters have more difficulty recovering energy from complex loads.

 

[esposcar]

Yes, my source is 3x 120V inverters, 120 degrees apart.
I initially drove 200V delta primary with the 208V, but that has higher idle current than I like (OK on-grid, to be avoided in an off-grid system.) With a typical transformer, I prefer to operate around half of rated voltage.

There was an imbalance in voltage, 125V vs. 120V.



That was the path I started down earlier. Wanted to step 208 up to 480. Used a 480 delta to 120/208Y transformer backwards. Didn't need neutral for my "load" (a 3-phase inverter), but I did need to ensure L1/L2/L3 didn't go significantly more than 277V above ground (which could happen with one line shorted to ground, and no current so no breaker tripping.)

I got a 600V 16A zig-zag transformer. Even at 480V, it drew more current than I preferred. Maybe due to that same phase voltage imbalance? (I forget if I even connected neutral. Probably not.)

I expected a zig-zag transformer to attempt balancing phase voltages. I did not expect a WYE transformer to do that.

The one I got was big enough to provide some 277V imbalanced loads, but not the full 30 kVA available. My intention was to run its neutral through a breaker, and gang it to breaker feeding primary.

My SPICE of zig-zag shows it works, but thus far I don't have a model that carries anywhere near the 16A this one is rated for. Just a dummied model, not actually based on measurements of this transformer. My attempts at measuring larger commercial transformers to make saturable hysteretic models have not gone well. Small toroids and chokes, I've been more successful.



Some combination of those in my case. My inverter is able to do four quadrants. I wouldn't be surprised if HF inverters have more difficulty recovering energy from complex loads.

You see we can begin building a more friendly environment Lol

 

[timselectric]

What are the currents with no load?

Can't shut down right now.
Unless I want to get yelled at. lol

With the above load, you could also measure (L1 and N?) of load, (L1 and L2, there is no neutral) of source.

Only 3 wires. L1 and L2 are Line. N is load.

N is only 0.16A less than sum. I don't understand why L1 and L2 are so different.

Because it's an autotransformer. And the load is unbalanced. Heavy towards L1. So L2 is picking up the difference.

Oh, stuffed back into source? (just like your auto-transformer.)

That's my assumption.
Due to the common core.

A delta auto-transformer would also be possible. Taps connected as originally intended would do that, probably not the voltages I'm looking for. Took me a while to find this delta I could convert to WYE and do what I wanted.

Can't quite wrap my head around that one, at the moment. Might have to think on it a bit.

 

[12VoltInstalls]

Well now I’m not learning. You lost me lol
You guys are smoking the same grade as Elon Musk.

 

[timselectric]

Well now I’m not learning. You lost me lol
You guys are smoking the same grade as Elon Musk.

@Hedges is thinking so far out of the box, that he's going to need a map to find his way back. lol

 

[esposcar]

Yes, my source is 3x 120V inverters, 120 degrees apart.
I initially drove 200V delta primary with the 208V, but that has higher idle current than I like (OK on-grid, to be avoided in an off-grid system.) With a typical transformer, I prefer to operate around half of rated voltage.

There was an imbalance in voltage, 125V vs. 120V.



That was the path I started down earlier. Wanted to step 208 up to 480. Used a 480 delta to 120/208Y transformer backwards. Didn't need neutral for my "load" (a 3-phase inverter), but I did need to ensure L1/L2/L3 didn't go significantly more than 277V above ground (which could happen with one line shorted to ground, and no current so no breaker tripping.)

I got a 600V 16A zig-zag transformer. Even at 480V, it drew more current than I preferred. Maybe due to that same phase voltage imbalance? (I forget if I even connected neutral. Probably not.)

I expected a zig-zag transformer to attempt balancing phase voltages. I did not expect a WYE transformer to do that.

The one I got was big enough to provide some 277V imbalanced loads, but not the full 30 kVA available. My intention was to run its neutral through a breaker, and gang it to breaker feeding primary.

My SPICE of zig-zag shows it works, but thus far I don't have a model that carries anywhere near the 16A this one is rated for. Just a dummied model, not actually based on measurements of this transformer. My attempts at measuring larger commercial transformers to make saturable hysteretic models have not gone well. Small toroids and chokes, I've been more successful.



Some combination of those in my case. My inverter is able to do four quadrants. I wouldn't be surprised if HF inverters have more difficulty recovering energy from complex loads.

Maybe one possible approach to improving your model is to try and obtain more accurate measurements of the transformer's behavior. This could involve using specialized equipment to measure the transformer's electrical characteristics, or conducting more extensive testing to observe how the transformer behaves under different conditions.

If you're struggling with larger commercial transformers, you may want to consider to continue working with smaller toroids and chokes first to refine your measurement and modeling techniques. Once you have a more solid foundation, you can gradually scale up to larger and more complex components.

 

[Hedges]

That sounds like exactly the sort of advice ChatGPT would give.
In fact, you did show up just after https://diysolarforum.com/threads/lets-find-out-what-chatgpt-ai-thinks.54361/
(I think I'm gonna start me a new Conspiracy Theory!)

I think I'll start a separate thread on transformer parameter measurement and simulation, rather than keeping something so important buried around post #1000 of this 5000es thread.

 

[msllarson2]

So, I found free access to NEC (NFPA 70) 2014 and have been trying to design a system that complies to it.
Signature Solar assured me that if I paid a licensed electrician they could setup the system with the auto-transformer according to code.

(challenge accepted)
I bought an extra auto-transformer(40a imbalance) to make sure I had capacity on my 3 inverter setup(15kw), and resolved to never plug the grid into my inverters(only my generator as needed).


However, while I was detailing the AC side of system schematic I stumbled across NEC 2014 Section 215.11.

It seems to me that NEC 2014 Section 215.11 is stating that feeders(anything after the supply and before a circuit breaker) shall not be derived from an auto-transformer unless the supplied circuit(load) has a grounded conductor(neutral) electrically connected to the grounded conductor(neutral) supplying the auto-transformer(inverter).

1. Is this requirement ever possible, unless the auto-transformer is directly being supplied by the grid/generator since the inverter does not have a grounded conductor(neutral) to supply to the auto-transformer?
2. If this is the case, can this system ever be run without being connected to grid or a generator and still meet fire code?

 

[timselectric]

So, I found free access to NEC (NFPA 70) 2014 and have been trying to design a system that complies to it.
Signature Solar assured me that if I paid a licensed electrician they could setup the system with the auto-transformer according to code.

(challenge accepted)
I bought an extra auto-transformer(40a imbalance) to make sure I had capacity on my 3 inverter setup(15kw), and resolved to never plug the grid into my inverters(only my generator as needed).


However, while I was detailing the AC side of system schematic I stumbled across NEC 2014 Section 215.11.

It seems to me that NEC 2014 Section 215.11 is stating that feeders(anything after the supply and before a circuit breaker) shall not be derived from an auto-transformer unless the supplied circuit(load) has a grounded conductor(neutral) electrically connected to the grounded conductor(neutral) supplying the auto-transformer(inverter).

1. Is this requirement ever possible, unless the auto-transformer is directly being supplied by the grid/generator since the inverter does not have a grounded conductor(neutral) to supply to the auto-transformer?
2. If this is the case, can this system ever be run without being connected to grid or a generator and still meet fire code?

This doesn't apply if you don't have the grid connected to your system.

 

[timselectric]

And you will have a grounded conductor (neutral).
You should have a N/G bond at your loads (main) panel.

 

[Hedges]

It seems to me that NEC 2014 Section 215.11 is stating that feeders(anything after the supply and before a circuit breaker) shall not be derived from an auto-transformer unless the supplied circuit(load) has a grounded conductor(neutral) electrically connected to the grounded conductor(neutral) supplying the auto-transformer(inverter).

I knew I read that somewhere, couldn't find it again.

I take it to mean you can't use center-tap of neutral to supply neutral of loads, such that the only path to grid neutral/ground is through windings of auto-transformer.
You can tie the center tap to grid neutral (will now try to rebalance grid)
You can tie load's neutral to grid neutral and not use auto-transformer center-tap while fed from grid. If you disconnect from grid, then connect load neutral to center-tap.

In other words, I think one should not switch neutral-ground bond. For a fixed not mobile system, connecting to grid neutral provides neutral-ground bond. Instead, switch center-tap to neutral connection, connecting it when disconnecting from grid. That is NOT what Victron auto-transformer does. I would consider modifying it.

I've been playing with this for 3-phase. I need to step up 120/208Y to 480 delta, or 277/480Y.
If I use three separate transformers in "Y" arrangement, no problem.
If I use an E-core "Y" configured auto-transformer, if mid point is wired to neutral of 120/208Y source (which happens to be imbalanced in voltage), high idle current of 4.2A
If I don't connect neutral, idle current 0.6A

The 480V load I'm trying to drive (TriPower GT PV inverter) seems to need the neutral connection, but I'm still confirming that.

 

[timselectric]

When I was connected to grid.
My autotransformer neutral was connected with the grid neutral.
I had a relay that disconnected the line conductors, in bypass mode.
Either way, the main N/G bond was used for all modes. (Common neutral)

 

[Hedges]

Either way (disconnect N, or disconnect L1 & L2), need to avoid operating loads with 240V from inverter and neutral undefined.
Was your autotransformer hardwired, no OCP? Or was there thermal/OCP protection that disconnected loads from inverter?

 

[msllarson2]

When I was connected to grid.
My autotransformer neutral was connected with the grid neutral.
I had a relay that disconnected the line conductors, in bypass mode.
Either way, the main N/G bond was used for all modes. (Common neutral)

Ok, So passing the generator ground-bonded neutral through to the panel that is feeding the transformer and the loads would satisfy 215.11 even when the generator is turned off (not actively suplying the auto-transformer)?

 

[timselectric]

Either way (disconnect N, or disconnect L1 & L2), need to avoid operating loads with 240V from inverter and neutral undefined.
Was your autotransformer hardwired, no OCP? Or was there thermal/OCP protection that disconnected loads from inverter?

Hard wired.

 

[timselectric]

Ok, So passing the generator ground-bonded neutral through to the panel that is feeding the transformer and the loads would satisfy 215.11 even when the generator is turned off (not actively suplying the auto-transformer)?

If your generator has a N/G bond. It needs to be removed. The generator manual should have instructions for doing this.