Prue Sine wave 3000 watt RMS inverter

[pwd666]

Hi Everyone!

I'm trying to build a power inverter to convert my DC solar cells and DC wind generator to 120 VAC. I have tried to find ideas of how to start online but I didn't find much as in circuits. This is what I want to do:


-have multiple inputs (wind, solar)
-have single out (120VAC)
-grid tied (same phase as the grid 120VAC)
-low THD
-Microcontroller controlled

I have figured out how to do the grid tied aspect of it but now I'm stuck on the pure sine wave part. I was thinking of using a couple boost regulators in parallel with the gate of the FET attached to a microcontrollers A/D. I would have the 120 VAC timed with the microcontroller and the generation of the sine wave would be phase locked. As I understand the boost regulator is used when you output voltage is greater then your input.

Also in pure sine wave intverters are large transformers used? If so how do you properly calculate the timing of the waves with changing RLC loads?

Thanks,

 

[Styx]

Well as you said you have figured out how to lock onto the mains freqeucy. so that is good.

Problem is it will be very noisy and distorted since everyone haning off you Point-of-common-coupling (PCC) will be distorting the mains waveform.

I would get a small voltage XFMR to take the mains voltage and step the 120Vrms to say 10Vrms (the level will be based upon what voltaqe you ADC can take).

You now have a copy of the AC at a lower voltage BUT noisy as hell. You will need to band-pass this signal to extract the fundemental 60Hz (it is 60Hz for the US?). a band-pass of abt 20Hz should suffice.


Now digitise it as your reference waveform.

Basically you need a closed-loop control system.
You have a reference waveform to follow, you need a feedback signal also.

With these two signals you need to take the difference between them (demand - feedback). This gives you the error signal. You will definitly need a PROPORTIONAL CONTROL (gain of which will take some tuning), but it might be worth concidering INTERGRAL CONTROL as well as PROPORTIONAL control to remove and steady-state error.

This error signal you then feed into a comparator input pin. The other be a triangle wave. This triangle wave definds your switching frequency. To have low THD you want this as high as posiible. I am working on a VF-CF converter, a 20Kw one and the switching frequency used is 20kW, you could get away with higher since it is only a 3kW inverter. With this triangle wave going into the other input of teh comparator.
The output will be your PWM.

Obvoiusly this can be done in your micro-controller, i am jut and analogue engineer.


So you have a demand (digities, filtered verion of the mains), you need a feedback to also digitise to then make the comparison. Use the output of your inverter, this will be at a low voltage already (you say it will run off of solar Cell)


Solar cells are low-voltage generators ~48V I have seen, Now your wind generator might be higher or lower. Whichever one prodices the lowest output will define your inverter operation.


Say the solar generate 48V and teh wind generates 30, you have to ensure it works at 30V.

 

[Styx]

Now you also say you want it to operate from teh 2 sources. IF bose sources are the same voltage output, just use 2 diodes to OR them together (into as big of a DC-link capacitor you can get!!!).

IF they dont, like I just said you are only going to be able to run off one at a time (without alot more complexity).


So lets say you now have a 30V DC-link (from whatever means), you need to get this upto 120V AC (which is 170V)

Thus you will need out ouput transformer to upp your inverter voltage to 120Vrms.
However, from your DC-link to your transformed voltage will be losses.

Thus the aim will be to make a 20V peak, this provides your voltage headroom needed. With a higher link, your control will still cope and still make a 20V peak waveform.

So the inverter. You will need MOSFETS (since low voltage) But ones that can handle 150A or greater (assuming a link of 30V, a higher link will cause less amps)

All this inverter needs to be is two FETS one on top of each other, the star point will be teh AC output. This needs to be connected to a L-C filter to remove the switching frequency. After the filter you will take the feedback signal as well as feeding it into teh trasformer.


OK I know this is alot, IF you give me abt 15min I will quickly draw up a conceptual drawing, it will be analogue and no specific IC/parts, but from that you should be able to isolate what needs to be done in the microprossor and also find the relevant parts

 

[Styx]

Here is an overview, details (transformer windoing ratios...) are needed to be set by you and alot of other details as well.

But this should be a start

 

[Synapse]

what happened with your project?

 

[asif_malik76]

i need a circuit diagram of purse sinewave inverter upto 1500 watts.
if anyone have this please send me.