The discussed idea of a simple grid-tie inverter circuit was suggested by one of the interested readers of this blog, Mr. RTO.
The images sent by him are shown below. In the first image we find the simple circuit diagram comprising a step down transformer for translating the grid data, a mosfet triggering circuit which accepts the grid data and a corresponding inverter transformer which is used to amplify the DC conversion of the grid data from the mosfet network.
The idea looks pretty simple, and indeed very smart:
The left side step down transformer feeds the half wave rectified voltage to the corresponding mosfets which begin conducting in-sync with the grid input and convert the DC source into a corresponding AC across the inverter transformer at the right hand side. The output from the inverter transformer which is now a grid synchronized AC feeds the grid with the intended GTI results.
The idea has been tested by Mr RTO, but he complains about lower efficiency from the unit.
This could be because of one major issue in the design, that is the absence of a "neutral" wire across the output of the inverter transformer.
With the shown set-up, the output would respond with a push-pull action across the secondary of the right hand transformer, meaning both the ends would become "HOT" or "LIVE" alternately during the operations.
The grid will take this as a "short" for every inverted half cycle from the transformer, because the grid voltage always has one wire as the neutral which is never a "LIVE" terminal.
We don't want this to happen.
A simple solution is to use a center tap winding for the secondary of the inverter transformer. This would render the center as the "dead" or "neutral" wire relative to the outer taps of the trafo. The upper tap may be configured with the grid while the lower tap to a balancing load or more effectively fed back to the primary side for charging the battery or reinforcing the DC source itself.
The test set-up of the above design can be witnessed here:
Another issue which could remotely transpire is the conduction from the mosfet which wouldn't be exponential, rather an "awkward" and unrecognizable sinewave.
The mosfets could be replaced with SCRS, as shown below. This would allow a perfect sine wave to be induced across the inverter transformer and the grid.
A much improved grid-tie inverter circuit using the above concept and SCRs is shown below. The idea looks greatly simplified, and quite impressive.
The output of the right and transformer could be seen converted to a center tap topology, wherein one half winding is integrated with the grid, while the other half is subjected to a balancing load so that the center tap is appropriately conditioned to be the neutral for the system.
The balancing load could be replaced with a charger circuit for charging the inverter battery itself, this would reinforce the input with additional power and more backup time.
At first glance it appears that the SCRs would get latched since a DC is being used across its anode/cathode, however according to me it won't happen, because the gate of the SCR is subjected with an alternately reversing AC which would prevent the SCR from getting latched everytime the gate AC feed changes its polarity