Pure Sine Wave Inverter

  • Thread starter dmaj
  • Start date
  • #1
Hi i am a 3rd year electrical student, i have a limited knowledge about inverters. i am trying to make a pure sinewave inverter, my intial spec was to create a 1) boost converter dc to dc 2) 3 stage pwm using a bubba oscillator etc 3) using a filter.

my supervisor for the project said i could make this spec @ 220volts however the person in charge of the lab said we are limited to 0.5A or 1A and about 30volts.

I am now confused and dont know what to do any advice or recomended text books or links will be greatly appriciated

thank you

Answers and Replies

  • #2
You can make an inverter from 12 v dc to ac, then use a transformer to increase the voltage to 220 V ac.
But the problem in the control signal of the switching devices(Mosfets, IGBTS..). In order to make a sine wave inverter you need to vary the duty cycle of the PWM signal applied to each switching device in a sinosiodal manner, so you can use a microcontroller like PIC 18f2331 to generate PWM signals, and you can test your program using a simulator which is Proteus, it is very good for simulating micro controllers.
After the generation of the control signal you have to make a driver circuit for the Mosfets, since the control signal voltage is about 5v and the Mosfet need a voltage greater than 5v to switch on, so you have to make a driver stage, you can use IR2110 integrated circuit for driving the mosfets, see the data sheet for more information about the connections.
The third stage is to filter the signal in order to get the 50 HZ output, LC filters are used for such applications.
  • #3
Oh wow thanks, I used a 555 timer to generate a pulse 2kHz but i am having problems getting the negative cycle of the squarewave. i tried inverting(not gate) the signal and putting into a H-bridge chip(L293), still no negative cycle. i got two pulses that are out of phase o_O. I designed and simulated a active filter which works perfectly using a 741 op-amp. I will looking into the microcontroller option thanks alot ^_^


  • 1_Nand squarewave.jpg
    1_Nand squarewave.jpg
    26.6 KB · Views: 747
  • 555 timer plus active filter response.jpg
    555 timer plus active filter response.jpg
    18.5 KB · Views: 695
  • #4
you don't really need a negative pulse to form the negative part of your AC output, what u should do is to get flip-flop chip (D-type) feed your control signal through the flip flop data input and take your complementary outputs from the the Q and ~Q pins. However, doing this will cut your output frequency in half, so you have to go back to the timing circuit, and increase the frequency(i.e if your ac frequency is f then the signal frequency input to the flip flop will be 2f i.e timing frequency of your 555 timer should be 2f) Note: I suggest that you try using sg3524, it has complementary outputs and the control circuitry is simplify. Refer to the datasheet to see details. see the attached file

Also, note that to have pure sine wave ac output, you need to use High frequency ferrite transformer because at high frequency iron core transformer will saturate easily and overheat causing damages to your circuit.


  • SG3524.pdf
    681 KB · Views: 546
  • #6
Science Advisor
Your problem is mainly political. Maybe you should mention to your supervisor the lab manager's objection to you producing voltages above 30 volts, presumably for safety reasons.

Then they could sort it out between them.

Depending on where you live, this may be a safety regulation that places your inverter into the same category as mains equipment, and for very good reasons. You could kill yourself and your relatives could sue the school.

You could contact the local Electrical authority and ask for the relevant regulations.

It seems strange to generate a square wave when you really want a sinewave. Why not use an opamp to generate a sinewave?

Or, use a binary counter to feed a digital to analog converter and generate an approximate sinewave like that.