Swapping between two inverters

[William_H]

The analogy might be 2x 12v batteries connected in parallel but erroneously connected pos to negative. Thus you have 24v short circuited. Disaster for batteries.
Now AC power swings from peak 325v positive to peak 325v negative. Described as 240v Root Mean Squared.(The effective power is equal to 240v DC)
Now the problem is that an inverter creates it's own timing. Not a problem until you parallel it to another AC source. Some inverters can be timed to an external source. My domestic PV inverter relies on incoming AC for timing so won't provide power without mains connection.
If you can get the positive cycle to coincide with the positive cycle of the other source you have no problems. Assuming frequency is identical.
The problem is that if cycles do not coincide there is a chance that positive cycle on one coincides with negative cycle of the other. Either fully or partially trying to negate the other. Potentially up to 650v short circuited. Most likely less but in any event putting major load on the 2 sources.
Now you might ask how power houses get syncronisation. In the case of a mechanical alternator once syncronised one tends to force the other by varying load to stay in sync. Actual connection can be achieved by finding a time when they are syncronised by measuring voltage difference. Old DC9 jets had 2 alternators 400hzt 115v 3 phase. They had 3 neons on the panel and you wait till the neons go out before switching the second alternator in parallel.
For OP I would stick with one inverter and wear the added losses at low power use. But if he goes for 2 inverters go for separate power outlets for the 2. ol'will

 

[Ian_Edwards]

You just made that up, it's utter garbage.

Paul, you give a lot of valuable advice on the forum, which I for one really appreciate.
But on this one you are wrong.
230 AC is a Root Mean Square. It transfers exactly the same power as a 230V DC power supply would. It allows you to use Ohms law on AC in exactly the same way as you can on DC.
Think of it this way:
If you averaged an AC voltage centred on zero, you would get zero. Equal voltage swings above and below zero.
Square
If you square the voltage, you effectively invert the negative half of the cycle, making it positive.
Mean
This can now be averaged, or the Mean calculated.
Root
The value you have for the voltage is now too big by the square root of two. You need to correct by multiplying it by the square root of 2 or approximately 1.414
This basic A level physics, just Google it if you don't believe me.
I essence its a "fudge" to make calculating AC circuits easy, using Ohms Law, by making RMS AC voltages equivalent to the same DC voltages.
The result is that the actual peak voltages on AC circuits are bigger than 230v by a factor of 1.414.

 

[geem]

We have a 3000w inverter that feeds the main AC circuit in the boat. The standing losses of this low frequently inverter are high. For small loads like the sewing machine (90w), makita 18v battery charger, etc, we have an 800w dedicated inverter that feeds a double socket in the nav station. The 800w inverter is located under the nav station so cable lengths are 600mm. It's simple

 

[PaulRainbow]

Paul, you give a lot of valuable advice on the forum, which I for one really appreciate.
But on this one you are wrong.
230 AC is a Root Mean Square. It transfers exactly the same power as a 230V DC power supply would. It allows you to use Ohms law on AC in exactly the same way as you can on DC.
Think of it this way:
If you averaged an AC voltage centred on zero, you would get zero. Equal voltage swings above and below zero.
Square
If you square the voltage, you effectively invert the negative half of the cycle, making it positive.
Mean
This can now be averaged, or the Mean calculated.
Root
The value you have for the voltage is now too big by the square root of two. You need to correct by multiplying it by the square root of 2 or approximately 1.414
This basic A level physics, just Google it if you don't believe me.
I essence its a "fudge" to make calculating AC circuits easy, using Ohms Law, by making RMS AC voltages equivalent to the same DC voltages.
The result is that the actual peak voltages on AC circuits are bigger than 230v by a factor of 1.414.

That's not the point i was trying to make Ian. Noelex claimed that with 2 out of phase inverters "The potential peak voltage is higher than 480V"

That's not correct. if they are out of phase they effectively create a short circuit and the circuit protection on the inverters will trip. Besides, the thread isn't, and never was, about connecting two inverters in parallel, it's about switching from one inverter to another, something that's simple and safe to do. A great many boats have 2 or 3 AC sources, inverter, generator and shore power and safely switch between them.

 

[geem]

...

 

[billskip]

That's not the point i was trying to make Ian. Noelex claimed that with 2 out of phase inverters "The potential peak voltage is higher than 480V"

That's not correct. if they are out of phase they effectively create a short circuit and the circuit protection on the inverters will trip. Besides, the thread isn't, and never was, about connecting two inverters in parallel, it's about switching from one inverter to another, something that's simple and safe to do. A great many boats have 2 or 3 AC sources, inverter, generator and shore power and safely switch between them.

I think there is a little misunderstanding here about the application
I don't think those quoting RMS/Peak are considering the both outputs are directly connected.
IIRC (i may not be) the old 17th edition stated that two separate outlets on different phases had to be a minimum of 2mtrs apart ,(and a warning label) .
The concern by some here is if the two inverters were on with two outlets there would be possible high voltage between if out of phase.

 

[Ian_Edwards]

The simplest way, least in theory, to run two or more inverters in parallel, is to a sign one as the mast the others as slaves.
Then link the master clock (which determines the output frequency) to all the other inverters.
My guess is that most modern inverters have this function. I know my Victron 3kVA/100amp 50Hz inverter charger does. It can be configured to run in parallel or with 2 others to form a 3phase system.
When connected to shore power and operating shore power reinforcement is enabled. It syncs to the shore power frequency, so if shore power or the generator fails, it will seamlessly either, add power, or take over to provide the power.
Someone who really understands the Victron units the OP has, should be able sort that quite easily.

 

[wonkywinch]

A nominal 230V AC supply is quoting root mean square voltage (RMS) . The peak voltage is given by multiplying the RMS by the square root of 2.

230 x 1.41 = 325V
Therefore, the peak voltage is 325V.

With two identical sine waves, in the worst case, the maximum voltage produced is the sum of their peak voltages.

Voltage peak 1 + voltage peak 2 = 325V+ 325V = 650V.

As I indicated the potential peak voltage is above 480V.

However, the exact peak voltage is unimportant; the message is that two out-of-phase 230V AC supplies will likely damage connected equipment.

I agree, but only if the neutrals were bonded else the two lives would be floating relative to each other.

 

[PaulRainbow]

The simplest way, least in theory, to run two or more inverters in parallel, is to a sign one as the mast the others as slaves.
Then link the master clock (which determines the output frequency) to all the other inverters.
My guess is that most modern inverters have this function. I know my Victron 3kVA/100amp 50Hz inverter charger does. It can be configured to run in parallel or with 2 others to form a 3phase system.

It is not the case that most modern inverters have this function, it's not that common at all.

When connected to shore power and operating shore power reinforcement is enabled. It syncs to the shore power frequency, so if shore power or the generator fails, it will seamlessly either, add power, or take over to provide the power.

Several inverters will act as a UPS, but not so many will allow shore power to provide the Power Assist feature.

Someone who really understands the Victron units the OP has, should be able sort that quite easily.

Your Multiplus is capable of being connected in parallel, as are the Quatros and some larger VE Bus inverters, but as i said previously, the OPs Pheonix cannot be connected to the Multiplus in parallel. With Victron inverters the two units have to be the same model, firmware, software etc etc , 100% identical in every way. I have the software and interface to be able to do this and have installed parallel systems for customers.

That said, it's totally irrelevant to this thread because the OP doesn't want to run both of his inverters in parallel, that totally defeats what he's trying to do.

 

[Dockhead]

Yes, I know 3000VA is equivalent to 3kVA.

I questioned whether you meant to use such a large power (P=IV) unit.

Clearly you did. Not a problem.

There is a common model of Victron Multiplus charger/inverter which is rated at 3000VA. Worth about 2500 watts depending on operating temperature. Most likely what the OP has.

It's what I have. Mine has the entire AC system of the boat running through it, as designed by Victron, and will power any item of AC gear. It is somewhat inefficient compared to a smaller inverter. I actually had the very same thought as the OP some years ago, and bought a small Phoenix inverter which I intended to wire to outlets supplying my bulkhead-mounted video screen, as that was the only item of low-powered gear on board which I was unable to arrange a direct DC power supply for.

I came to the conclusion that it was not worth the complexity and so never installed it.

My inverter typically runs 24/7 when I'm off grid. The advantage of having AC power available instantly outweighs the power it uses to keep it running all the time. In my case. YMMV.

 

[rotrax]

Running 2 inverters on a boat even if switched independently is highly dangerous. If they are out of phase which they almost certainly will be, you can have 480v between the two, an almost certain death shock, in the small space in a boat you cannot guarantee to achieve sufficient separation. Don't risk it!.

My limited experience with 'out of phase' AC was watching my mate Glyn who was the Diesel Power Station Engineer on Alderney spending considerable time adjusting the Generators governors to be just 'right', the two needles from the outputs flicking up and down, perfectly syncronised.

When he demonstrated what went wrong if there were no perfect syncronisation, the switch gear threw out with a real bang!!

So, unless it is synchronised, it wont connect.

If it DOES connect, voltage would be the same as the phasing is syncronised.

I MAY be wrong, but having seen two AC Gensets being connected more than once, I dont think so.

One was a 18th Century watermill configured to be a water powered generator near Moravany in East Bohemia, Czech. This used a 1910 grease filled German made hydraulic govenor. After cleaning the sluice screens of debris, the wheel was restarted and connected to the Czech National Grid. Once again, the two phase needles looked good to me, but the switchgear threw out big time until a small, but obviously vital, adjustment was made.

Once again, it was my late friend Glyn from Alderney who had the 'feel' for it, and was successful where the Czech manager was struggling.

 

[MapisM]

I came across this thread by coincidence.
The reason why the OP didn't follow up yet is bound to be that he didn't have enough headache pills to follow the useless out of phase debate.

Or I could just keep one AC circuit and swap between inverters.

I have a 20A rated 1-off-2 changeover switch which would be useful for this.

Do that, and call it a day.
Just check that the switch you've got is 2 poles, as PR mentioned in post #13, and wire it as he suggested - job done.

One even simpler alternative that I would consider is use just the Multiplus, and keep the small inverter as spare. That's because whenever your AC demand is low, the Multiplus will seamlessly draw less DC current from the battery, with no need to do anything at all from your part.
In fact, the overhead (for lack of better wording) absorption of the Multiplus under these operating conditions may be a tad higher than the smaller unit, but I very much doubt that the difference is large enough to bother switching between them.
Don't ask me to quantify that, though: I can't recall to have ever seen it in any Victron specs sheet.
OTOH, it's quite easy to check the DC absorption of the two units under identical low load conditions, and see if the difference (if any) is worth getting out of bed...

 

[PaulRainbow]

There is a common model of Victron Multiplus charger/inverter which is rated at 3000VA. Worth about 2500 watts depending on operating temperature. Most likely what the OP has.

It's what I have. Mine has the entire AC system of the boat running through it, as designed by Victron, and will power any item of AC gear. It is somewhat inefficient compared to a smaller inverter. I actually had the very same thought as the OP some years ago, and bought a small Phoenix inverter which I intended to wire to outlets supplying my bulkhead-mounted video screen, as that was the only item of low-powered gear on board which I was unable to arrange a direct DC power supply for.

I came to the conclusion that it was not worth the complexity and so never installed it.

My inverter typically runs 24/7 when I'm off grid. The advantage of having AC power available instantly outweighs the power it uses to keep it running all the time. In my case. YMMV.

Our boat has a Victron Multiplus 2 3000/24/70 inverter charger. We also have 900W of solar and are connected to shore power when on our home berth. The boat is 100% electric, no gas.

The inverter is programmed to ignore the shore power inlet, so it runs the whole boat from the LFP batteries.

If the inverter output exceeds 3000W it switches to shore power. If the shore power current draw is subsequently above 4000W the inverter kicks back in with the Power Assist function, giving us up to 7000W, seamlessly.

When in inverter mode, with the shore power ignored, if the battery SOC falls below 50% it enables the sore power and charges the batteries back to 60%, then reverts back to inverter mode. This stops the batteries from getting low (yes, i could go lower with LFP, but that seems pointless in this scenario) but also leaves plenty of capacity for solar yield. This works well in scenarios such as using a lot of power in the evening for cooking etc, the batteries are kept between 50% and 60% until the Sun comes up in the morning and solar fully charges the batteries.

 

[geem]

I came across this thread by coincidence.
The reason why the OP didn't follow up yet is bound to be that he didn't have enough headache pills to follow the useless out of phase debate.

Do that, and call it a day.
Just check that the switch you've got is 2 poles, as PR mentioned in post #13, and wire it as he suggested - job done.

One even simpler alternative that I would consider is use just the Multiplus, and keep the small inverter as spare. That's because whenever your AC demand is low, the Multiplus will seamlessly draw less DC current from the battery, with no need to do anything at all from your part.
In fact, the overhead (for lack of better wording) absorption of the Multiplus under these operating conditions may be a tad higher than the smaller unit, but I very much doubt that the difference is large enough to bother switching between them.
Don't ask me to quantify that, though: I can't recall to have ever seen it in any Victron specs sheet.
OTOH, it's quite easy to check the DC absorption of the two units under identical low load conditions, and see if the difference (if any) is worth getting out of bed...

It would be good if somebody could measure the actual Multiplus current at low load. I know my low frequently inverter, load load current is surprisingly high. My inverter has an efficient of only 88% compared to the Multiplus 94% for a 24v 3kw version

 

[rotrax]

It is basic electrical engineering.

Hopefully, we can at least agree that connecting two inverters without synchronising the phases of their waveforms is a bad idea.

I have seen, on two occasions, two AC Generators being synchronised. The voltage remained the same. As I said earlier, when the operator - of, admittedly, very old kit - failed to accurately syncronise the two output waves, they would NOT connect. When they did, no increase in voltage on the instruments.

The switches were massive, manually controlled and would not have been out of place in a Frankenstein Movie!! The dials were about 6 inches in diameter and could be clearly read.

 

[Beneteau381]

I have seen, on two occasions, two AC Generators being synchronised. The voltage remained the same. As I said earlier, when the operator - of, admittedly, very old kit - failed to accurately syncronise the two output waves, they would NOT connect. When they did, no increase in voltage on the instruments.

The switches were massive, manually controlled and would not have been out of place in a Frankenstein Movie!! The dials were about 6 inches in diameter and could be clearly read.

I used to work offshore Angola with two 12 cylinder cats running on produced gas. I used to synchronise with the hertz meters, get them both up to 60 then press the button for the synch breaker. A big bang as the breaker connected and all well. If it wasn’t right, the breaker wouldn’t go in.

 

[JOHNPEET]

I used to work offshore Angola with two 12 cylinder cats running on produced gas. I used to synchronise with the hertz meters, get them both up to 60 then press the button for the synch breaker. A big bang as the breaker connected and all well. If it wasn’t right, the breaker wouldn’t go in.

You can’t synchronise generator outputs on a frequency meter which measures cycles per second and basically relates to engine rotational speed. Where generators are synchronised manually, this would be done using a synchroscope meter and measures the differential in phase angle between the two generator outputs. Very often, a check sync relay will back up the manual or even on modern systems automatic synchronisation, to prevent the generator circuit breaker from closing when the differential is outside of the set window of acceptability.

 

[rotrax]

You can’t synchronise generator outputs on a frequency meter which measures cycles per second and basically relates to engine rotational speed. Where generators are synchronised manually, this would be done using a synchroscope meter and measures the differential in phase angle between the two generator outputs. Very often, a check sync relay will back up the manual or even on modern systems automatic synchronisation, to prevent the generator circuit breaker from closing when the differential is outside of the set window of acceptability.

Well, my late mate Glyn used two flicking needles, one for each genset, mounted next to each other. As they were rising and falling, he made minute adjustments to what I imagine was a govenor. When he was happy, the needles rising and falling in perfect unison, a very large antique switch was thrown, connecting the two outputs.
When it connected, a collective sigh of relief. When it did not, a bloody great bang and a sore hand. What the needles measured was never explained any further than "I've got to get 'em in phase together, or it wont connect. So, IF these needles were measuring the phasing/frequency, your first premise is clearly wrong. I saw the Alderney kit synchronised, plus the Moravany water wheel syncronised, both, 30 - 35 years ago, using antique kit. A pleasure to see master of what was clearly a 'black art' at work.

 

[JOHNPEET]

Well, my late mate Glyn used two flicking needles, one for each genset, mounted next to each other. As they were rising and falling, he made minute adjustments to what I imagine was a govenor. When he was happy, the needles rising and falling in perfect unison, a very large antique switch was thrown, connecting the two outputs.
When it connected, a collective sigh of relief. When it did not, a bloody great bang and a sore hand. What the needles measured was never explained any further than "I've got to get 'em in phase together, or it wont connect. So, IF these needles were measuring the phasing/frequency, your first premise is clearly wrong. I saw the Alderney kit synchronised, plus the Moravany water wheel syncronised, both, 30 - 35 years ago, using antique kit. A pleasure to see master of what was clearly a 'black art' at work.

Before anyone mentions - yes, I know this is massive thread drift, but it’s all interesting stuff - well to me anyway!

Phasing and frequency are two separate things!
Frequency is simply a measure of the rotational speed of a generator - a 4 pole AC generator needs to rotate at 1500RPM to produce 50 cycles per second or 50Htz. One engine/generator revolution produces 2 complete sine waves 1500/60 = 25 revolutions per second. 25 x 2 = 50
A 2 pole machine needs to run at 3000RPM to produce 50Htz
You can’t synchronise two generators using a frequency meter! This is what I stated in my initial comment.

Whether you use flicking needles, lamps or any other form of instrument such as a synchroscope, they’re all set up to measure whether typically the L1 output (sine wave) of the two generators or one generator and the grid if syncing to the grid are close enough to being “in phase” with each other. In phase means that the L1 sine waves of both machines if laid on top of each other would match within acceptable limits. If they matched exactly, the voltage between them at a specific point in time would be zero. If they matched exactly and the voltage between them is zero - when closing the generator circuit breaker, there would be no current flow between them. A good time to close that breaker!
Once the generator circuit breaker is closed, you can then increase the fuel supply via the governor. This won’t increase the engine speed as the generator is magnetically locked in sync, but will have the effect of sucking load onto that generator from the other generator or grid depending on the application.

 

[Sea Change]

What you’re really describing is load sharing logic, and Victron doesn’t natively hand off between mismatched inverter sizes. I’ve seen people solve it with an auto transfer relay and current sensing, so the small one runs everything by default and the large one cuts in only if draw spikes past a threshold, that way you don’t rely on memory or switches

That's what I would ideally like to set up but from everything I've read on this thread it sounds pretty complicated and the risks of getting it wrong could be very bad.

I'll probably just go for the simplest suggestion, having separate outlets. Stuff like phone chargers tend to start plugged in to the same places anyway.

How bad is it to overload an inverter? I've never actually done it. I've previously overloaded my batteries, causing a BMS trip, which was well before the inverter's own limit. That killed the inverter, so it's left me a bit wary.