# DC-AC converters, Solar inverters

1. May 24, 2015

### 2sin54

(I am a big newbie when it comes to electrical circuits)

So I am searching the Internet for how the Solar inverters work. Most schemes that I come across have an H-bridge with some sort of transistors. Now how I understand the transistors are controlled by an oscillator and bridge itself then converts DC current to AC current (though a square wave one).
Is this correct so far? Also, how do they get a pure sine wave from a square wave? By using inductors?

2. May 25, 2015

### Hesch

Yes, you are right.

The output from the transistor bridge is a rectangular wave, where the relation between high-voltage-time and low-voltage-time is varied. It is called a pwm-signal ( pulse-width-modulated signal ):

In this way the mean voltage in the output signal can be controlled, and filtering the ouput current through an inductor, it will become sinusoidal. If the net-frequency is 50Hz, the rectangular signal could be clocked at say 16kHz.

The net is regarded as a synchronous motor to be driven. The inverter will simply follow the net/motor frequency and will try to accelerate the motor thereby delivering power to the net/motor. This is done by making a phaseshift ( 90 degrees which yields max torque as for the motor ) and the relation between active/reactive power can be controlled by the amplitude of the output voltage. If a negative phaseshift is made ( -90 degrees ) you will brake the motor/consume power from the net. So the inverter can easy be made an AC-DC converter.

Last edited: May 25, 2015
3. May 26, 2015

### 2sin54

I came across the scheme I am sharing a link to below. I have a hard time understanding how does the first H-bridge (Q2-Q5) NOT convert the DC to AC. Apparently, this conversion is only achieved in the second H-bridge (Q6-Q9). What other function could the first bridge serve?
http://solar.smps.us/grid-tie-inverter-schematic.png

4. May 26, 2015

### Hesch

Well, first of all the L1-Q1 circuit acts as a DC-Voltage step up circuit ( don't remember its name ). So C2 will be charged with a higher voltage than C1.

In principle C2 could be charged to as high voltage as needed (voltage crossing C3) with respect to the grid-voltage. But of some reason "they" have chosen to make an "inter-inverter" that drives the transformer T1, probably at some high frequency ( say 1kHz sinusoidal ). Then of course you are missing the (safety)transformer, T1, that separates the solar common voltage from the grid-voltage, but Q2-Q5 and T1 can be used to further step-up the voltage.

Missing the safety-transformer, T1, you could instead replace L3-A and L3-B with a transformer with a core wherein is made an air-gap. In this way the surrogat transformer will act as a transformer and a reactance at the same time.

Last edited: May 26, 2015
5. Jun 10, 2015

### jim hardy

Q2 -Q5 indeed does make AC from the stepped up DC voltage at C2.
That AC is immediately made back into DC by bridge D2-D5 and smoothed by L2-C3.

The first bridge and T1 together raise the low solar panel DC voltage to a higher DC voltage suitable for the second bridge to chop into a sinewave , as Hesch showed in post #2. Observe that to chop DC into 120 volts AC you need to start with 120√2 volts .of DC..

While they didn't say what is voltage at C2 i'd bet a steak dinner it's a lot less less than at C3 .
If you have one of these gizmos take a look ay C2 and C3. Their working voltages will be written on their sides.

6. Jun 11, 2015

### 2sin54

Thanks guys for the replies and clearing things up.