DC/AC conversion topologies. Share your experience...

[happy2005]

Hi All,
I want to design an DC/AC power inverter. input voltage is 24V DC, output voltage is 220V/50Hz AC and output power will be about 1500VA. I want to keep overall efficiency at maximum. I want to use high frequency transformer, not 50Hz transformer. Switching frequency is >20kHz beyond audible frequency.

I want to use push-pull DC-DC converter for 24-to-311Vdc conversion. (220rms*√2=311Vpeak). Output voltage will be controlled in this stage at 311Vdc for all load conditions by a microcontroller, So inverter output voltage will be fixed at 220Vrms.

Later a full bridge DC/AC inverter from 311Vdc to 220Vrms. In this stage a fixed PWM pulses will be used at modulation index m=1. Maybe harmonic elimination method can be used.

I think, i should use double conversion method. At first DC-DC conversion, later DC-AC conversion. So, which topology is suitable for each conversion stage for this power level? half bridge, push-pull or full bridge? advantages and disadvantages? power limitations?

Please share your comments, experiences, design examples, documents, internet links etc. I think this topic will be very useful for the power electronic designers in this forum.

Regards!

 

[Davood Amerion]

I think in first and second stage is better to use full bridge.

 

[scottieman]

Hi All,
I want to design an DC/AC power inverter. input voltage is 24V DC, output voltage is 220V/50Hz AC and output power will be about 1500VA. I want to keep overall efficiency at maximum. I want to use high frequency transformer, not 50Hz transformer. Switching frequency is >20kHz beyond audible frequency.

I want to use push-pull DC-DC converter for 24-to-311Vdc conversion. (220rms*√2=311Vpeak). Output voltage will be controlled in this stage at 311Vdc for all load conditions by a microcontroller, So inverter output voltage will be fixed at 220Vrms.

Later a full bridge DC/AC inverter from 311Vdc to 220Vrms. In this stage a fixed PWM pulses will be used at modulation index m=1. Maybe harmonic elimination method can be used.

I think, i should use double conversion method. At first DC-DC conversion, later DC-AC conversion. So, which topology is suitable for each conversion stage for this power level? half bridge, push-pull or full bridge? advantages and disadvantages? power limitations?

Please share your comments, experiences, design examples, documents, internet links etc. I think this topic will be very useful for the power electronic designers in this forum.

Regards!

It is very difficult to mention all stuff here. Therefore, let me start asking some question for the discussion

As mentioned by you, you would like to controlled the VDC to be 311V (e.g. by feedback) to make sure the sine wave will be 220Vrms. My question is, will you use your DC-AC for any inductive loading or capacitive loading. If the answer is yes, your Vrms MAY NO LONGER be 311V/sqrt(2). Even your loading is pure resistive, your sine wave may not be exactly equal to 220Vrms, because there is always power loss in power transistor that causes voltage drop.

Hope we can share more here

Scottie

 

[happy2005]

Hi Scottie,
Thank you for your reply. I think the voltage drop at switching transistors can be compensated easily by increasing 311Vdc to 312Vdc or more.
How the capacitive and inductive loads effects the output ac rms voltage? I know it makes only phase difference between voltage and current waveforms. Additionally, may be, it can cause some stability problems in controlling 311Vdc. What is your opinion?

Regards.

 

[scottieman]

Hi Scottie,
Thank you for your reply. I think the voltage drop at switching transistors can be compensated easily by increasing 311Vdc to 312Vdc or more.
How the capacitive and inductive loads effects the output ac rms voltage? I know it makes only phase difference between voltage and current waveforms. Additionally, may be, it can cause some stability problems in controlling 311Vdc. What is your opinion?

Regards.

For the 311Vdc r 312Vdc problem, it is not as simple as raising the DC voltage, because the voltage drop acorss the power transistor depends on loading current. In case you raise the VDC to 313Vdc to compensate the IR drop acorss the power transistor in 2A loading current to get 220Vrms, you may have 225Vrms in 0.5A loading current. It is a problem.

For driving capacitive or inductive load, the problem is the load will change the LC filter characteristic used in the DC-AC. Therefore, the output wave form may not be pure sine wave anymore. In worse case, the waveform can be triangular.

Scottie

 

[happy2005]

i think, i can solve the I.R voltage drop problem by using output rms voltage as a feedback signal. But, may be, inductive and capacitive loads effects the stability of the control loop as you sad. If it is occurs, i can try classical PWM technics.
what about mosfet driver circuit? IR2113 is a little bit expensive but it is simple solution. main disadvantage is no isolation. can anybody advice different driver circuit? especially for high side mosfet.

thanks.