Discussion on the use of 2 120vAC HF inverters to create 240vAC.

I have not put much thought into these setups since I have not needed to supply 240vAC in a split phase arrangement. With a LF (Low Frequency) style inverter you have a transformer on the output that you use a center tap to give you the normal US grid style of 240vAC/120vAC where you have two hots (L1 and L2) and a neutral. But with two HF (High Frequency) inverters (not operating in parallel since that would give you the same output voltage) connected so that each sine wave output is offset 180 degrees your doing something quite different. How each inverter communicates with each other to maintain this offset is likely PFM (Pure Fracking Magic) to me but the consequences of the arrangement is of interest.

Technically speaking is having two HF inverters in this setup really producing a 2 phase system? How does each inverter handle neutral when doing this?

In case you've unblocked me...

HF/LF isn't relevant. It's no different than 2X LF 120V inverters. My neighbor has 2X 120V TRACE SW4024 from the 1990s that can operate in split phase.

From a layman's perspective, they simply synchronize their frequency via a communication cable. Their neutrals are connected, and their hots become L1 and L2, respectively. Their transformers are effectively stacked.

My Victron Quattros work this way as well.

Is this identical to a center tapped 240V transformer? No. Does it behave exactly like it in every meaningful way? Yes.

Im under the impression that line and neutral are sort of just naming conventions more than they refer to any physical distinction between the 2 terminals of an inverter as a single phase ac source.. I think either terminal of a single phase ac output is really BOTH poles of the source anyway (ie both terminals are functionally the exact same thing) but depending on which 60th of a second you’re asking it may be hooked to one or the other side of the source at that instant. I dont know if that’s a helpful statement, maybe just more confusing or possibly even false.. ?but i dont think an inverter needs to ‘handle neutral’ any differently. It’s mainly an issue of whether the inverter has the functionality of being able to sync with another to be run in parallel in the first place. I think..

Short answer, no. The inverters need to be able to syncronize their output sine waves and if they can't sync up, your sine waves will be all over the place.

Mattb4 said:

Technically speaking is having two HF inverters in this setup really producing a 2 phase system? How does each inverter handle neutral when doing this?

Click to expand...

For a better understanding of how 2 inverters produce 240V, its necessary to follow the current path THROUGH both "H" bridges. I'm sorry I don't have a link to a YT video or tutorial that demonstrates this.
Neutral is shared by the 2 inverters and is located between the 2 "H" bridges. Very similar to how a center tap in a transformer is placed.

If you look up a pic of an H bridge it may make more sense of what i was saying.. In a simplified sense the way you get ac from dc is just switching the dc polarity back and forth rapidly. So both terminals, regardless of being called line or neutral, function identically. Each is hooked to one ‘pole’ of the source half the time, and the other pole of the source the the other half of the time. So both terminals connect to the exact same things. There is nothing remotely like the physical difference of construction, routing and function of what neutral means in a split phase system. So neutral is just a name we stick on one terminal of a single phase inverter output, but that terminal functions the EXACT SAME as the other terminal we call Line. The only difference is neutral could be intentionally shorted to the chassis (neutral-ground bond) while line never should be and if it is its a fault condition.

I understand that neutral is simply the point that the sine wave passes through zero volts. However the difference of a single, 2 or 3 phase output from a standard generator is based on how you physically locate the winding and set of poles in the generator. Using two or more HF inverters that can synchronize their sine wave outputs to be out of phase so as to match this is where it makes me ask about the combined neutral.

But as I mentioned it is most likely PFM to me.

It doesn't matter if HF or LF inverters but they need to have serial capability.

Not only is sync (180 degs out of phase lock) necessary, but also coordinated operation, like if one inverter shuts down for any reason, the other also shuts down. There are other coordination's for other features, like charging.

Calibration is critical. If one inverter float and/or absorb charging settings is off calibration the two inverters can fight each other on the same battery.

Two series inverters have an advantage on AC input neutral imbalance as the two inverters can independently match their respective L-N grid voltage. On LF inverters when one side of AC L-N has a glitch it will create a momentary 'swapping of current' between the two inverters via common battery connection as the inverter(s) adjust to their glitched or varied L-N voltage.

One tricky part is when AC input goes open circuit, like if you flip off AC input breaker. Zero AC input current is a legitimate condition that does not cause pass-through relay to disconnect. Unless inverter has an active UL1741 phase wobble test, the inverter may not realize the AC input has gone open if AC input current is low.

When this happens, the inverter will wander off in frequency since, when pass-through relay is engaged, the inverters track to their respective L-N AC input phase and frequency. Inverter will eventually wander outside of freq range limits causing pass-through to release, but this wreaks havoc on 240 vac output due to freq wander.

It is not until pass-through relay is released that the inverter also switches to its own frequency reference and then relies on sync with other series connected inverter.

Inverter phase sync has to make a hard decision when to sync to AC input or to other inverter. It must be one or the other, and it coincides with pass-through relay engagement. A safety check is inverters monitor phase difference between the two inverters and if it gets too deviant, it decides something must be wrong with AC input and triggers a simultaneous pass-through disconnect of both inverters, slewing back to master inverter's freq/phase reference. Most inverters have an 'early/late' zero crossing phase detector. If it does not get some pulses from the early/late phase detector in a few cycles it usually means something is not right with AC input.

Thanks RC. I suspected that running two inverters together to produce a simulation of a 240vAC/120vAC setup was not easily done. What I get from your explanation is there is a good chance of problems.

HF and LF differ if the LF is outputting 240vAC/120v ac from the transformer versus combining two of the LF at 120vAC and using some control program to keep the phase offset. So it does matter.

Thanks all who posted for your input.

ETA: One thing I see a lot of that bugs me is using the word "parallel" to describe using two inverters to simulate a split phase type of output. This is a misuse of two sources of power being in parallel arrangement. In parallel voltage stays the same and current adds. But there are many terms misused in the solar world.

Series stacking is debatably not as tough as paralleling inverters. Nastier things can happen when inverters are paralleled.

Another item when stacking inverters is to be sure all inverters have same revision of firmware. Many manufacturers do not provide complete instructions on updating firmware when you have them operating in a stacked arrangement. It is another fine detail that must be thought of by firmware designers, but often overlooked.

I would imagine that one inverter must act as the Master inverter in a stacked arrangement even if both have the circuitry to do it. But as I said in my first post I have no experience in these type of setups.

Yes, one inverter is master, but when AC input is present, the respective AC input phase/freq is the master for each inverter's phasing sync. Not until all units pass-through relays open does the master inverter set the phasing reference.

That is another criterion for stacked inverters, no inverter will engage its pass-through relay until all inverters have acquired their respective AC input sync and all agree they are ready to close their pass-through relays. This is also part of simultaneous control such that if one inverter releases from AC input, they all should release from AC input.

There are a lot of these fine details of operational conditions that must be handled.

AC pass-through operation and UPS style of switching when using two inverters is something I had not thought about. On the face of it it would seem to be simple enough as not requiring the stacked inverters to work together. But I had not thought about the fact these inverters have to pick up a "bypass" relay to create the output versus a passive pass through.

ETA: One thing I see a lot of that bugs me is using the word "parallel" to describe using two inverters to simulate a split phase type of output. This is a misuse of two sources of power being in parallel arrangement. In parallel voltage stays the same and current adds. But there are many terms misused in the solar world.

Click to expand...

Well, now that you mention it, i saw your thread before work, went to work, and replied to it after work, and something happened at 9:49am in between there and while i can't remember i feel like it's relevant to why i was talking about parallel and why i posted a video about whether things are in phase or out of phase in a transformer based split phase system. It has to do with using the centertap transformer you mentioned, vs NOT using a centertap transformer because that totally changes things. If there is no centertap transformer than the whole concept of a 'neutral' terminal or conductor is meaningless other than having to bond one side of things to some kind of grounding conductor system for fault detection. Maybe i'm wrong and if so someone should tell me but if you are using two single phase sources in series with no centertapped transformer then there is no 'neutral' in the usual sense, there is just the side that is bonded to the ground, and the side that isn't. Neutral is just a name, and while in reference to a transformer system it does refer to a specific conductor/terminal that is definitely NOT the exact same thing as the Line conductor/terminal, in the case of a single phase system of two sources stacked in series and no transformer, neutral is not 'its own thing'. It's just a colloquialism/shorthand way of referring to which side is to be bonded to ground. But it is not 'handled' any different than the Line side because other than the ground bond it IS no different than the Line side. Each side becomes the other side at least 60 times a second (in reality way more otherwise you'd have square wave, not sine wave)! In my mind my 'misuse' of the term parallel was trying to respond to what i saw as a misunderstanding of something from a transformer system being applied to a non-transformer system.

Anyway, i'm sure it seems like a tangent even if im right, and if im wrong im just wrong (and someone should tell me). But i'll leave it alone after this!

I will have to think about your point for a while Vigo. By afternoon/evening my aged mind does not process things all that well.

On a related note: I got my hands on a 3 phase panel and though it may be possible to run a second, smaller inverter on the 3rd leg. If the neutrals of both inverters are separated, might this work?

HighDesertOffgrid said:

On a related note: I got my hands on a 3 phase panel and though it may be possible to run a second, smaller inverter on the 3rd leg. If the neutrals of both inverters are separated, might this work?

Click to expand...

No.

Vigo said:

...

Anyway, i'm sure it seems like a tangent even if im right, and if im wrong im just wrong (and someone should tell me). But i'll leave it alone after this!

Click to expand...

Probably RCinFla could explain the electrical theory behind why neutral is more than just a theoretical exercise when operating from two separate AC produced sources operating in tandem to create a 240vAC.

Yes you short the neutral outputs of both inverters together so that they are the same potential when combined. But they, like two batteries of slightly different voltages wired in parallel can output the same voltage, but the current each delivers is not the same. Any imbalance on the neutrals is going to create unwanted effects. I do not know if this furthers any understanding. At my age of 64 my ability to handle electrical theory is not what it once was. Especially since it is a rare occasion that I call upon it.

So are you right are are you wrong? Heck if I know. I am lucky to be able to still feed myself.

Two 120v inverters (either HF or LF makes no difference) creating 120/240v split-phase, are placed in series. If placed in parallel, you just get twice the amperage at 120v. In either orientation, they must have communications between them to be in sync. Otherwise the phasing will be off.
The output is a single phase (with a center tap) synchronized sine wave.
The center tap must bonded to ground if you want a neutral. But it's a good idea to do it even if you don't.
There's a lot of confusion about split-phase. The two legs are NOT 180° out of sync with each other. If they were you couldn't get 240v between them. They would appose each other. And the result would be zero volts between them.
The reason people think that they are 180° out of phase is because of how most people connect an oscilloscope. They connect the leads 180° from each other, so that they can see each leg individually.

timselectric said:

There's a lot of confusion about split-phase. The two legs are NOT 180° out of sync with each other. If they were you couldn't get 240v between them. They would appose each other. And the result would be zero volts between them.

Click to expand...

Again its the terminology that confuses things. The 180 deg. out of phase is simply referring to the direction of current flow through the "H" bridges in each inverter with respect to which 2 of the 4 FET's are turned on. The actual sine waves are very much IN PHASE or they wouldn't be additive.

BentleyJ said:

Again its the terminology that confuses things. The 180 deg. out of phase is simply referring to the direction of current flow through the "H" bridges in each inverter with respect to which 2 of the 4 FET's are turned on. The actual sine waves are very much IN PHASE or they wouldn't be additive.

Click to expand...

It's alternating current. It flows in both directions, through both inverters. But in the same direction at any point in time.

timselectric said:

It's alternating current. It flows in both directions, through both inverters. But in the same direction at any point in time.

Click to expand...

Yes, same direction with respect to the wires but this requires 2 of the H bridge FET's in one of the inverters to close FET's #1 & #3 while the other inverter closes FET's #2 & #4 to complete the circuit loop and vice versa. Thus the poor use of terminology calling it 180 deg out of phase. They are not really out of phase its just that one inverter can be thought of as Pushing while the other is Pulling (neutral in the middle) so the current path through the H bridges is different between the 2 inverters.
In a parallel configuration both inverters push or pull at the same time which requires the same 2 FET's in both inverter H bridges to open or close simultaneously so the relative current path through both H bridges is the same.

RCinFLA said:

That is another criterion for stacked inverters, no inverter will engage its pass-through relay until all inverters have acquired their respective AC input sync and all agree they are ready to close their pass-through relays. This is also part of simultaneous control such that if one inverter releases from AC input, they all should release from AC input.

Click to expand...

This is not the case for my Sunny Island. For instance, I have a 3-phase 120/208Y setup, and only the master syncs and connects to 120/240V grid (one phase.) The other two remain disconnected (they aren't even wired to grid) and produce the missing phases. Same works for split phase.

If loads or AC coupled PV exceeds what one inverter can put through its relays, this can be a problem. Loads, they will supply from battery to limit AC current draw. PV backfeeding, not clear it will charge then disconnect when it can't take the current, and manual says not to connect too much PV. Two in parallel, I think I've tripped a breaker and/or caused inverter to throw faults. I'm contemplating 2x separate 2s inverters, rather than my present 2p2s system. Might add a relay to parallel their AC outputs when "off grid" status is shown.

HighDesertOffgrid said:

On a related note: I got my hands on a 3 phase panel and though it may be possible to run a second, smaller inverter on the 3rd leg. If the neutrals of both inverters are separated, might this work?

Click to expand...

sunshine_eggo said:

No.

Click to expand...

I would say yes and no.
You could use the 3rd leg for another inverter running asynchronously. I considered having it be UPS backed up while the other two were grid fed. But easy to screw up, plug in a 2-pole breaker and span UPS to grid with a load.
You also have to be careful not to overload neutral, which could have sum of current from two unsynchronized phases.

Better to use separate boxes.

Hedges said:

I would say yes and no.
You could use the 3rd leg for another inverter running asynchronously. I considered having it be UPS backed up while the other two were grid fed. But easy to screw up, plug in a 2-pole breaker and span UPS to grid with a load.
You also have to be careful not to overload neutral, which could have sum of current from two unsynchronized phases.

Better to use separate boxes.

Click to expand...

I agree it would be better to use a separate panel, but...
I am currently using the panel as a single phase with the clips for the third leg removed. I have considered this since I installed it. My thought was to use two legs for my AIMS 6500 and the third leg for a smaller 2-3Kw inverter to run necessary circuits. This way the AIMS could go into economy mode and reduce idle consumption considerably. The idea was to completely separate the neutrals of each inverter's circuits to their own lugs and isolate the breakers for the small inverter at the bottom of the panel. I'm 100% off-grid and no one will have their hands in my panel but me. My concern has been with neutral/ground bonding and back feeding the unsynchronized phases. Either way, I appreciate the theoretical exercise.

BentleyJ said:

Again its the terminology that confuses things. The 180 deg. out of phase is simply referring to the direction of current flow through the "H" bridges in each inverter with respect to which 2 of the 4 FET's are turned on. The actual sine waves are very much IN PHASE or they wouldn't be additive.

Click to expand...

I have 2 Trace sw4024 24v inverters same firmware revision but not installed yet. I do have an Onan QD 12.5kw generator with both legs fed to a 120v panel. I have a 50amp shore power cord connected through an isolator/switch/filter. I can get 120v to the panel from shore or the generator. But I want to get two 24v 200Ah LifePO4 batteries and install the inverters. Should I just parallel the batteries and stack the inverters? I downloaded a file that shows a connecting cable to stack the inverters. I really need some help such as a diagram. Thanks hopefully I need help.