bridgeless AC-DC converter

[tonyctsiu]

I am developing a wind mill powered bright LED.
The windmill generates power from 0-30vAC depending on windspeed.
The target operating input voltage of the AC-DC circuit is 2-40vAC.
The LED operates at 3V or 6VDC (1-2LEDs)

I have found the papers below which discuss how to convert AC to DC efficiently without using bridge rectifier.

Paper 1: https://www.ijser.in/archives/v2i1/SjIwMTM4MQ==.pdf
Paper 2: https://digitalcommons.calpoly.edu/cgi/viewcontent.cgi?article=1219&context=eeng_fac

My question is, which circuit will yield more brightness from the LED under the same windspeed?
Are there still any AC-DC works better then those ones?

Thank you and best regards.

Tony Chun Tung Siu

 

[chuckey]

I do not believe that these circuits are suitable, because when you read the papers, they talk about 50 KHZ and then work out the storage components (4.7 * 10^-6 H + 500MFD). Your windmill will be producing a few hundred Hertz (at the most) when its is producing 3V. Get the frequency up as high as possible by using more pick up coils in the wind turbine and a transformer to step up the AC before turning it into DC. Use a constant current generator to power the LED.
Frank

 

[FvM]

Two limitations should be considered:
At low wind speeds, the load current can slow down or effectively stall the wind generator. At high wind speeds, the output power may exceed the LED power. To get an idea of the possible converter operation range, you would determine a load characteristic for different wind speeds. As long as you don't need to limit the power delivered to the LED, there will be a maximum power point for each wind speed.

You have to find out, if boost operation is of any practical interest, or if the output power is too low for output voltages that might need boosting.

I agree with Chuckey that a simpler AC/DC converter possibly can work for the application.

- - - Updated - - -

I do not believe that these circuits are suitable, because when you read the papers, they talk about 50 KHZ and then work out the storage components (4.7 * 10^-6 H + 500MFD). Your windmill will be producing a few hundred Hertz (at the most) when its is producing 3V. Get the frequency up as high as possible by using more pick up coils in the wind turbine and a transformer to step up the AC before turning it into DC. Use a constant current generator to power the LED.

50 kHz is the PWM frequency and isn't related to the AC input. The circuits in the two papers are primarly intended for 50/60 Hz input. The component values should be however scaled to the actual voltage and power range. In addition, you'll better add an input filter capacitor to prevent switching frequent losses in the generator impedance.

I'm not sure if the considerable circuit effort pays for the windmill application. But if you want it just as a demonstration of the converter principle without a circuit complexity versus point x percent efficiency improvement trade-off, it can still make sense.

 

[tonyctsiu]

Thank you for you guys replying.

The wind mill is producing 500 hz at 15VAC actually when I blow it at wind at 40km/h.. at 10km/h the wind mill produce 3VAC at 100Hz.
I think if the wind mill is not generating high enough voltage, we can some up boost it up to produce enough voltage for lighting up the LED even the current is smaller
When the wind mill is operating at a speed which generate too high voltage, the MOSFET will switch on at very low duty cycle to buck (step down) the voltage for the LEDs.
That's my idea.

 

[FvM]

Load characteristic involves voltage versus current. You are only talking about voltage yet.

 

[Externet]

I would do it this way:

(AC)--------|>|----------------Creg-----------A(led)K-----------(AC)
(AC)------K(led)A-------------Creg--------------|<|-------------(AC) <--- optional for second LED

Creg are current regulators, can be LM317L based;
Diodes like 1N400x or Schottky. Uses 2 LEDs, high efficiency ones at 15000mcd @ 20mA are available.
A, K are anode, kathode. AC are your generator output leads.

The leds will be as bright as the current is set for them, on a wide span of voltage generation.
-No bridges- Some filtering would be good at the current regulators inputs.