Is this OK for a Grid tied inverter?

Hi,
Is the attached OK to take forward as a Grid Tied Inverter design?
As you know, it still needs all the protections to be added.
It shovels a sinusoid current into the mains.
LTspice sim and pdf schem attached.

No.
The first thing you need for grid tie is phase and amplitude synchronisation, before the power circuit connection to the grid is made during the startup sequence.
The output must be sinusoidal, or very close to it, to avoid noise and harmonics.

Then advance the drive phase to achieve the required current out. I do not believe you can do it without a fast MCU directly controlling the PWM.

As it is, without power circuit isolation, your power bridge will work in reverse as a bridge rectifier and power the DC side from the AC side - and it has no external sync or amplitude matching control.

This includes a sine wave PWM system, but not the phase sync; it's a starting point:

(Look up the sequence for bringing a power station alternator online - that's in essence what you have to emulate with a grid tie inverter).

As it is, without power circuit isolation, your power bridge will work in reverse as a bridge rectifier and power the DC side from the AC side

Click to expand...

Thanks, i dont know if i labelled it properly...because the DC side is isolated 400V, and the AC cant get through to it, as its only 339Vpk.
The simulation in the top post confirms this.

What happens at night / dead batteries?
You still need isolation and synchronisation.

Plus it is a basic regulatory requirement that the solar power side has _total_ isolation (eg. transformer) from the grid side.

And, it must shut down if the grid power fails, so it does not keep anything outside the property live.

This is a basic functions block diagram (from Analog.com)

What happens at night / dead batteries?
You still need isolation and synchronisation.

Plus it is a basic regulatory requirement that the solar power side has _total_ isolation (eg. transformer) from the grid side.

And, it must shut down if the grid power fails, so it does not keep anything outside the property live.

Click to expand...

..thanks yes, i agree with all that you say, but i mustnt have explained well enough,.......the 400V of my diagram of the top post is indeed an isolated 400V...it doesnt come from a battery......your concerns are good ones though, but the top post is a general schematic, not the finished thing, as you would know.

But the stark thing is, that it shows the basis design of a grid tied inverter, using noddy level electronics......which i believe shows that these things arent as rocket-science as the tech-world would make them out to be.

I agree with the need for synchronisation...the sim of the top post generates a sine in sync with the mains...so thats already there, yes.
So yes i agree with all you say, and all you say , is in keeping with what i show in the top post.
It shows the basis for a grid tied inverter, being sucessfully done, with what is bascially 14 year old kid electronics.
And yet it is obviously a "goody".....yes it needs the extra protections to be added, its not the finished article yet, but a very good basis, you would agree?

You need that isolation transformer. Period....
Your design is nowhere near complete, it will not work properly.

Yes i agree ...i havent shown the isolation transformer...ive just shown its 400V output. The emphasis is to demonstrate the inverter, rather than the isolating power supply adjacent to it.

it will not work properly

Click to expand...

...when the simulation is run...it is seen to work very well. Puts a sine into the mains. Its by no means finished, but is anything wrong with the principle?

Flyback said:

...when the simulation is run...it is seen to work very well. Puts a sine into the mains. Its by no means finished, but is anything wrong with the principle?

Click to expand...

It is an inverter. It is not in any way a "grid tied" inverter; the only reason it appears to work is that the frequencies are locked by the sim.
Change the mains frequency and the result is chaos, rather than adaption.



Frequency, phase and voltage matching are the first essentials, before the power side connection is made.

It is an inverter. It is not in any way a "grid tied" inverter; the only reason it appears to work is that the frequencies are locked by the sim.
Change the mains frequency and the result is chaos, rather than adaption.

Click to expand...

Thanks yes, i agree with your statement.
It was just to demonstrate, that the method shown , is worth taking forward, for developement, towards becoming a grid tied inverter.
My apologies if i misled that this was a "completed" grid tied inverter.
The method shown, shows the basis of a successful way of injecting a sinusoidal current into the mains and thats all.
I was wondering, if viewers, may spot something about this injection method, which suggests it may not be useable?

An inverter can create its own sinusoidal wave, with whatever phase shift or frequency stability it wants. Any load will run reasonably with an inverter.
A grid tied inverter has to synchronize with the existing grid phase and frequency, and monitor/adjust the output voltage to be able to inject to the grid. The output voltage has to be a "bit" higher than the existing grid voltage to feed back to the grid. Your circuit does not have those components. What you have is a simple inverter plan so far.

Flyback said:

I was wondering, if viewers, may spot something about this injection method, which suggests it may not be useable?

Click to expand...

Trying to shoehorn a basic PWM IC into an application such as this, rather than looking at _all_ the requirements for such a system and allowing for them from the start.

Grid tied systems are very different thing to a basic "dumb" inverter. Your diagram is also missing such as current feedback from the legs of the bridge and dead-time control.

Start with an MCU based system. Many have power control PWMs with current feedback, dead time and other safety and monitoring related functions, before you even get to the rest of the peripherals.


The system has to be fully controllable with serious built in "intelligence". The requirements are similar to those for bringing a conventional alternator generator online, but whereas they inherently give a decent sine wave and will self-synchronise tiny errors, you have to do _everything_ in software for a grid tie unit.


Once it is synced and brought online, the power delivery is largely controlled by phase advance, not much voltage increase; the grid voltage is supposed to be at a constant voltage, but a slight phase advance means loads can draw power from that source fractionally before others, so the output current increases.

Get the .pdf from the top link on this page:

That shows an outline design for a low power grid tie inverter, less the input DC-DC section and output isolation control, but a reasonable example of the monitoring, control and bridge drive in between those parts.