Most dimmers use pulse width modulation (PWM) to control the amount of
power that is delivered to the lamp. Those that come bundled with a
switch faceplate control the firing angle of a Triac on the 240V mains
side. These work fine with resistive loads but may not be suitable for
inductive loads such as low-voltage halogen lamp transformers. This
circuit also employs PWM but it switches at a high frequency (22kHz) on
the low-voltage side of the lamp transformer. This high frequency also
simplifies EMI filtering. Furthermore, because this circuit is isolated
from the mains by the transformer, it is relatively safe to build and
install.
IC1 is a standard 555 astable oscillator with a high duty cycle. It produces a narrow negative-going pulse at its pin 3 output approximately every 45µs (ie, the frequency of oscillation is about 22kHz). These pulses trigger IC2, another 555 timer, this time wired as a variable monostable. IC2's pin 3 output is normally low which means that its internal discharge transistor is on and the 1nF capacitor on pins 6 & 7 is discharged. However, when the monostable is triggered (by IC1), its output goes high, the internal discharge transistor turns off and the 1nF capacitor charges via VR1 & VR2 until it reaches 2/3Vcc.
At this point, the output at pin 3 switches low again. Each time pin 3 of IC2 goes high, it turns on power Mosfet transistor Q1 which in turn switches on the lamp. Potentiometer VR2 is used to control the time it takes the 1nF capacitor to charge to the threshold voltage and thus sets the width of the output pulses. At maximum resistance, the pulse width is 55ms. This is longer that the 45ms period of oscillator IC1, and so IC2's pin 3 output is high for 100% of the time and the lamp operates with maximum brightness. Now consider what happens if the monostable's period is shorter than the astable's.
IC1 is a standard 555 astable oscillator with a high duty cycle. It produces a narrow negative-going pulse at its pin 3 output approximately every 45µs (ie, the frequency of oscillation is about 22kHz). These pulses trigger IC2, another 555 timer, this time wired as a variable monostable. IC2's pin 3 output is normally low which means that its internal discharge transistor is on and the 1nF capacitor on pins 6 & 7 is discharged. However, when the monostable is triggered (by IC1), its output goes high, the internal discharge transistor turns off and the 1nF capacitor charges via VR1 & VR2 until it reaches 2/3Vcc.
At this point, the output at pin 3 switches low again. Each time pin 3 of IC2 goes high, it turns on power Mosfet transistor Q1 which in turn switches on the lamp. Potentiometer VR2 is used to control the time it takes the 1nF capacitor to charge to the threshold voltage and thus sets the width of the output pulses. At maximum resistance, the pulse width is 55ms. This is longer that the 45ms period of oscillator IC1, and so IC2's pin 3 output is high for 100% of the time and the lamp operates with maximum brightness. Now consider what happens if the monostable's period is shorter than the astable's.
In this case, each time IC1's pin 3 output goes low, pin 7 of IC1 also
goes low and discharges IC2's 1nF timing capacitor via D3. This
retriggers the monostable. As a result, IC2 is triggered at a 22kHz rate
and produces variable width pulses depending on the setting of VR2.
It's output in turn pulses Q1 to control the lamp brightness. D2
isolates IC1's timing circuitry from IC2's. VR1 is used to set the
minimum lamp brightness when VR2 is at minimum resistance. If this
control is not required, VR1 can be replaced with a 1.8kO resistor. The
220µF capacitor on pin 5 of IC2 provides a soft-start facility to
prolong lamp life.
Initially, when power is first applied, the 220µF capacitor is discharged and this lowers the threshold voltage (which is normally 2/3Vcc). That in turn results in shorter pulses at the output. As the 220µF capacitor charges, the threshold voltage gradually increases until the circuit operates "normally". For the prototype, Q1 was a BUK553-60A, rated at 60V, 20A & 75W. Q1's maximum on-state resistance is 0.1O, so switching a 4A lamp load results in a maximum power dissipation of 1.6W. The bridge rectifier comes in at around 5W and so both should be mounted on suitable heatsinks.
The power dissipation in the bridge rectifier can be reduced by using power Schottky diodes rated at 5A or more. The output of 555 timer IC2 is capable of directly driving several Mosfets (up to four in tests). Note, that if the Mosfet is going to be some distance from the 555, it will be necessary to buffer it. Power for the control circuitry is derived from 3-terminal regulator REG1 which produces an 8V rail. This in turn is fed from the output of the bridge rectifier via diode D1
Initially, when power is first applied, the 220µF capacitor is discharged and this lowers the threshold voltage (which is normally 2/3Vcc). That in turn results in shorter pulses at the output. As the 220µF capacitor charges, the threshold voltage gradually increases until the circuit operates "normally". For the prototype, Q1 was a BUK553-60A, rated at 60V, 20A & 75W. Q1's maximum on-state resistance is 0.1O, so switching a 4A lamp load results in a maximum power dissipation of 1.6W. The bridge rectifier comes in at around 5W and so both should be mounted on suitable heatsinks.
The power dissipation in the bridge rectifier can be reduced by using power Schottky diodes rated at 5A or more. The output of 555 timer IC2 is capable of directly driving several Mosfets (up to four in tests). Note, that if the Mosfet is going to be some distance from the 555, it will be necessary to buffer it. Power for the control circuitry is derived from 3-terminal regulator REG1 which produces an 8V rail. This in turn is fed from the output of the bridge rectifier via diode D1
Tidak ada komentar:
Posting Komentar