pure sinewave feedback

Thread Starter

myil

Joined May 2, 2020
142
Hi Everyone,

I have a question for you. I have been working on designing an inverter. The idea is to boost up 12VDC to 340VDC by using push-pull topology.
After that, this 340VDC will be chopped by SPWM signals and will be filtered to obtain pure sine wave. My question is I have a feedback circuit for DC to DC converter part which seems to be working okay and keeps 340 volt stable. But, for the pure sinewave output, do I need a feedback circuit to keep the voltage and sinewave stable. If so, how can I manage it? Thanks
 

Ian0

Joined Aug 7, 2020
6,738
If your 340V DC is stable, then you'll probably get away with running the sinewave stage open-loop, provided that your output spec isn't too stringent.
What THD and voltage spec do you have to meet?
 

Thread Starter

myil

Joined May 2, 2020
142
If your 340V DC is stable, then you'll probably get away with running the sinewave stage open-loop, provided that your output spec isn't too stringent.
What THD and voltage spec do you have to meet?
THD isn't important. I would like to get 220VAC as an output.
 

Ian0

Joined Aug 7, 2020
6,738
220V AC +/-1%, 220V AC +/-5%, 220V AC +/-10% ?
And if THD isn't important, then why not use the "modified sine wave" approach, which is really a modified square wave that's been to the marketing department.
 

Thread Starter

myil

Joined May 2, 2020
142
220V AC +/-1%, 220V AC +/-5%, 220V AC +/-10% ?
And if THD isn't important, then why not use the "modified sine wave" approach, which is really a modified square wave that's been to the marketing department.
As long as it's pure sinewave, no problem. it's necessary to be a pure sinewave for my project.
 

Ian0

Joined Aug 7, 2020
6,738
You didn't say how accurate you needed the voltage. As an example an emergency power supply requires regulation of +/- 6% and distortion of <5%.
How good's your knowledge of Bode Plots, Nyquist stability criterion etc.?
If you're good at that, add feedback, otherwise keep it open loop.
I'm assuming that R8 is actually your load resistor and the power is 160 Watts.
You have a 50Hz side and a PWM side of a full bridge. The PWM should NOT be 50% when the sinewave is at zero-crossing.
It should vary from all zeros at zero-crossing to all ones at peak then back again to all zeros at zero-crossing. Then it should swap immediately to all ones, then vary to all zeros at negative peak then back to all ones at zero crossing.
Otherwise your output is equal to a sinewave plus a squarewave, or, worse, a sinewave minus a squarewave.
Also, as each pair of comparators has the same input connections (but reversed) you could use half as many comparators and drive LIN with an inverter (74HC04). There are some half-bridge drivers out there (FAN73932, for instance) that only require one input.
 
Top