[SOLVED] sine wave inverter given squre

Hi all,

am working on a sine wave inverter using pic16f877a microcontroller
the specification is as follows;

the switching frequency is 10khz.
output frequency is 50hz.
am using ccp1 and ccp2 module. ccp1 switches the first half cycle while ccp2 for the 2nd half cycle.
am using h-bridge with ir2110 for mosfet driver.
am using a step-up transformer [12v to 230v]
output power is 300watt

at the output of the pic16f877a, i got a modulated sine wave
when i filter it with R-C i got sine wave
at the output of my transformer without filter i got sine wave but not pure
but with my filter the waveform is squre.
no matter the value of inductor or capacitor i put it will still be squre wave.
am thinking maybe is my osciloscop dat is given me the error.

pls. i need ur help because for the past two weeks i have been on it and no matter what i do i keep getting thesame result.

pls i really appreciate any help or advice.


pls if you need any other information about the inverter, let me no.

You will not be able to make the full 230 vac sinewave with a 12v to 230 vac transformer but that's not the question.

What type of transformer are you using, a high frequency ferrite based or low frequency (50 hz) transformer.

If using high frequency you need a good L-C output filter design based on maximum peak 50 Hz current so filter coil does not saturate. The low frequency 50 Hz component looks like a DC bias to the filter coil core. If core saturates the inductance drops off and you don't get filtering.

If using a low frequency transformer you can use the leakage inductance of the transformer to represent a series coil and filter the output with just a capacitor across secondary of transformer. A normal 50 Hz 300 watt transformer may have well confined fields resulting in low leakage inductance (normally a desired parameter) but this may be too low effective series inductance to make a good high frequency filter with just a cap across secondary. In this case you can suppliment filtering with an additional LC filter on output.

Measure secondary winding inductance (230vac side) with primary shorted (12v side) to get effective leakage inductance of the transformer. Use this value for the output filter cap design.

Normally, low cost sinewave inverter use a high frequency transformer driven push pull to create about 350 vdc rectified secondary output. (able to hit sinewave peak voltage) then have H-bridge output that PWM's the 350vdc to create sinewave. Output has high frequency L-C filter.

Thanks for the reply RCinFLA, i really appreciate it.

i am not after the output voltage for now but the wave form,
because i no if my output voltage is higher or lower,
i need to adjust the windings. thank you for observing that.

i am using a low frequency transformer. i don't have an inductance meter
to measure the leakage. this coming week after this weekend, i will go to one of the labs close
to me to find out if i will be able to get one to measure it.
if i can't get it then i will enback on constructing one. do you have any other advice for me?
then if i use a low frequency transformer of rating 8v to 230v,
and its gives me sine wave will i be able to use this trasformer to charge the battery [12v] just like ups?
as soon as i get through i will let you know.

i appreciate your reply.

If you have an audio generator and AC voltmeter you can resonate the winding with a capacitor to get an idea of inductance. It's small signal so full run voltage inductance in lower.

High voltage winding magnetizing inductance is probably in range of a Henry.
Secondary winding is fraction of Henry.
Leakage inductance with low voltage winding shorted is a few milliHenry.

Select an approximate cap value based on resonating near 50 Hz. Put a series cap in audio gen fed that is less then 1/10 the resonating parallel cap value to keep loading on winding low. Place high Z AC voltmeter across winding and adjust audio gen freq to peak AC voltage across winding. Back calculate inductance based on actual resonate freq with cap value used to resonate. If resonance is less then 30 Hz or greater then 70 Hz readjust the parallel cap value selected to get closer to 50 Hz resonance.

Measure series resistance of primary, series resistance of secondary, magnetizing inductance (no load on other winding), and finally leakage inductance. In your case, the high voltage secondary leakage inductance is important (shorting low voltage side) as it will represent the series inductor in the output high frequency L-C low pass filter with just a parallel cap on secondary.

On the final cap selection for filtering, make sure you keep the parallel resonance with secondary magnetizing inductance much higher then 50 Hz. Not sure what a good target is but I am guessing at least 4 or 5 times 50 Hz. Getting close to resonance can cause high voltage to develope on output during light loading on inverter. Oil filled AC moter run caps are what you can use.

Commercial units like Outback FX series, or Xantrex XW series DC-AC inverters work with low freq transformer with high freq PWM MOSFET H-bridge input. They design transformer with compromised laminate core coupling to get a target leakage inductance.

Normally this is not what you do for a normal power transformer as leakage inductance is considered a bad thing.

Thanks,
i discovered yesterday that the oscilloscope am using is given squre wave with a known sine wave source
which means i have to look for another one to test

really, i am using a software on my computer as my oscilloscope which take its input from my computer soundcard
let me go on and look for real oscilloscope to test then i will post my result here
then about my last post what can you say about the charger?

thanks,
is my oscilloscope that is malfunction
i have tested it with another and is giving sine wave