Remember electricity is dangerous and can be fatal, you'd be qualified and competent to carry out any electrical work.
If you have seen other AVR schematics, you will notice that this one is quite similar in its fundamental operation. But there are some differences to the most typical designs. For example, I placed the 10µF capacitor, which filters the voltage sample, behind a large portion of the total resistance in the divider, unlike most designs that place it pretty much at the input, without the 47k resistors in front of the bridge rectifier. My approach is neither better nor worse than the standard one. It's simply different. The standard design will tend to regulate the peak voltage, while mine will tend to regulate the average voltage. The former is best for powering electronic equipment, while the latter is best for avoiding light flickering.
The 12 V winding feeds a voltage doubler, that's made with 4 diodes, rather than the more usual 2-diode design. The additional diodes basically provide protection against reverse polarity on the filter caps, in the event of any weird phenomenom occuring in the generator's stator, caused by loads with poor power factors. Most likely these extra diodes aren't really required, but they are cheap protection. And if you want me to be honest to the bones, I have to confess that mainly I have them in the circuit because I was too lazy to remove them, when I changed the rectifier configuration from a bridge to a voltage doubler!
The 1N4007 feeding supply voltage into the 10µF capacitor is another useless remnant. In my original AVR, powered by the 55V winding, this diode would become forward biased when the generator is loaded by a very crooked load (low power factor), that drives up the 55V and down the 220V. This diode would then make the AVR regulate the 55 V winding rather than the depressed main one, providing protection against burning out the AVR from excessive supply voltage. Now, with the AVR powered from about 25V, there is no risk of this, and at the same time this diode will never be forward biased, and is thus exemplarily useless! I didn't remove it because of sheer laziness (did I mention already that I have been becoming lazy lately?), but you might want to use this idea if you will adapt this AVR to some generator that does need a higher excitation supply voltage than mine.
The most special part in this AVR is the battery pack, which consists of four AA-size NiMH cells. It's provided with a switch, so the battery can be disconnected when the generator is not in use. When the switch is on, the battery powers the circuit through the diode, providing enough voltage to turn the MOSFET on, and puts enough current through the field winding to make the generator start up at once, when it begins rotating. During normal operation, the battery is trickle-charged through the 2k7 resistor. If you are making an AVR for a generator that has enough remanence to self-excite, then of course you don't need to include the battery circuit! But with such a generator you might want either a bipolar transistor, or a MOSFET or IGBT with a very low threshold voltage, instead of the standard MOSFET I used. Because that generator might generate enough voltage on remanence to overcome the rectifier diodes' drop, but not enough to bias a normal MOSFET into conduction!
The two small transistors and the power MOSFET constitute a pulse width modulation cell, that modulates cleanly all the way from zero to 100% duty cycle. The scope screenshot shows the waveform at the MOSFET. The frequency is low, around 100 Hz, leading to high switching efficiency despite the large gate resistor. A higher frequency isn't needed, because of the very high inductance of the field winding, which holds the field current highly constant while the field voltage switches between zero and about 25V at 100Hz.
The control loop response is that of a simple proportional function with a narrow action range (high gain). It holds the voltage "stable as a plank" during speed and load variations, at least when looking at it on my analog panel voltmeter! Actually the output voltage fluctuates by about 1% when the load changes from zero
Негізгі бет Designing an AVR Automatic Voltage Regulator use Mosfet for 1kW 1kVA Generator Alternator genset
Пікірлер: 10