Gaurang:
I have implemented two different systems for controlling the effective power delivered to a load from an AC power source.

In the first application I equipped the triac with zero crossing turn-on time control and then I had a PWM control that was set to run at a frequency of line frequency divided by 16. In my application I synchronized the PWM period start time to the output of the zero crossing detector. As the PWM control signal duty cycle varied from near zero toward one hundred percent I was able to obtain 32 distinct steps in the amount of power delivered to the load. These steps follow a linear relationship. Keep in mind that this procedure may not be suitable for all types of loads because the load may not integrate the bursts of delivered power in an effective manner. After all, with a 60 Hz AC frequency the PWM frequency is only 3.75 Hz. I was driving a heating coil that boiled liquid nitrogen in a closed containter system and the PWM control was used as part of the control system to regulate the pressure in the container. The picture below shows how this type of concept works.



The second scheme I have used for AC power control is to once again synchronize the PWM waveform with the zero crossings of the AC waveform. This time the PWM period is set to match the period of one half cycle of the AC waveform. Then the PWM duty cycle is varied to delay the time during the half cycle that the triac is allowed to turn on. In this scheme the power delivered to the load is not a linear function of the PWM signal duty cycle. It is however possible to create a translate table that converts a linear input step sequence to the appropriate PWM duty cycle in order to achieve the effect of a linear control of power delivered to the load. I have used this scheme to control the brightness of lamps and speed of small AD/DC type appliance motors. The picture below shows how this concept operates.



Good Luck
Michael Karas