Synchronizing inverters with generators

Explained in detail here:

IDK what they do, but inverters for solar power have a problem identical to what you're describing. They need to sync the inverter output to the AC mains.

simple PLL?

It's called "AC couplting" in the off-grid world or on-grid PV with backup using batteries scenario. A decent grid-tie only PV inverter could work with that Honda inverter generator because it's frequency is steady. That is, if the generator is large enough that the PV inverter's anti-islanding doesn't trip because that inverter-generator is fairly small compared to the grid impedance I think.

Some battery powered inverters for off-grid or backup have a feature called "Gen-Support" where they synchronize with the incoming 60 Hz generator to supply more output power than the generator can supply by itself. The trick is to follow the wandering frequency of the generator but of course an inverter generator like that Honda is already very stable. Those inverter generators could do the same thing if they had the smarts.

No communications necessary although that can make things easier. Then it's called stacking inverters for more power.

You want to use an inverter with batteries to sync with it ? Try Schneider or Outback Power or maybe Victron Energy ?

Fun stuff ! PLLs are involved but usually in software.

Apert from the usual question 'Who's on first?'; if there is a measurable difference between two voltages (amplitude and phase), it can be used to govern the connection. The tie point switch or governor is not likely to be visible at the terminal block or physical tie.

The inverter 'clock' is not directly responsible for the output, and may have no relationship to the power line frequency.

Low power oscillators, disconnected and decoupled, are highly unlikely to synchronize, no matter what you thought you were looking at. There are infinite methods of coupling that are not obvious at first glance.

RL

Aye, there's the rub, a little too much detail 8-) My fluency in Patentese isn't quite up to the task. Still, there are references to "master" and "slave", implying some sort of active communication between the inverters. What I can't understand is how it can be done without any extra connections beyond the power output lines.

Thanks for writing!

bob prohaska

That's a good point which I overlooked entirely.

Thank you!

bob prohaska

Grid tie only PV inverters are comparitively easy to synchronize with the grid since they just put out a current that is proportional to the grid voltage.... Much like a PFC circuit draws an AC grid sinewave shape of current from the grid. Just the other way.

A voltage source inverter like the Honda is harder but is done all the time.

I don't think so. The diagram on page 71 of the manual shows the parallel o utput jack as simply a tap off the power receptacles. The inverter is prett y much a state-of-the-art switcher topology, using high power IGBTs, and pr obably switching at something like 30-50 kHz. The rotational speed of actua l magnetics/ windings generator portion is kept high, has nothing to do wi th output signal frequency, doesn't vary all that much with loading, and i s rectified to DC internally to power the inverter. It's the electronics i n the inverter that determines output frequency, phase, amplitude, and, for emost, spectral purity of the signal , which Honda brags about quite a bit. My ***guess*** is the inverter is sampling and monitoring the possibly bid irectional power flow at the output receptacle and adjusting the signal par ameters to maximize the delivered output. This would mean the two parallel generators end up splitting the load 50-50 exactly. You'll notice the gener ators want that parallel cable installed while they're off, plugging it in to a hot generator might damage something. And the people mixing and matching di fferent generator brands or models within the same brand, have no idea what they're doing. The Honda's are pricey but they pay for themselves in opera ting cost reduction and longevity. They're a good buy.

I should add that the parallel cable is mainly there to put the combined power on a single AC output receptacle.

The two 555 oscillators were on one breadboard, with the usual power decoupling applied. Perhaps it simply wasn't sufficient.....

Putting the question differently: Is there a way to smoothly transition loads from a conventional synchronous generator to a lower power and more efficient inverter unit? The idea would be to start wtih the synchronous generator, once the starting surge has damped out allow the inverter unit to pick up the load and shut the synchronous machine off. The motive is fuel efficiency. The starting load is too much for the inverter, but the running load well within capacity.

I already have the synchronous generator and wondered, based on the lack of explicit sync wiring, if an inverter unit capable of parallel operation might be able to synchronize itself. I picked the Honda EU1000i as an example simply because it's well-known and the manual is on-line. I wouldn't expect Honda to admit their generators can be paraleled with anybody else's even if the idea is workable.

An island grid tie inverter probably would do what I want, but that's overkill. I'm looking for a cheap solution to a small scale problem.

Thanks for everyone's insights!

bob prohaska

Consider the simplest inverter, the Royer oscillator, (which is a kind of blocking oscillator) connect two of them in parallel and they'll self-synchronise.

I don't know what they actually use inside the generator

It sounds plausable. but it's not stated how the generators synchronise with each other, if the second one come up in some sort of "slave mode" when you're using the link cables it might not work well when the master disappears.

The manual says not to disconnect the link cables when the generator is running, but doesn't say why.