Specifications

PWM Waveform Generation With Compare Units and PWM Circuits
6-55 Event Manager (EV)
6.6 PWM Waveform Generation With Compare Units and PWM Circuits
A pulse width modulated (PWM) signal is a sequence of pulses with changing
pulse widths. The pulses are spread over a number of fixed-length periods so
that there is one pulse in each period. The fixed period is called the PWM (carri-
er) period and its inverse is called the PWM (carrier) frequency. The widths of
the PWM pulses are determined, or modulated, from pulse to pulse according
to another sequence of desired values, the modulating signal.
In a motor control system, PWM signals are used to control the on and off time
of switching power devices that deliver the desired current and energy to the
motor windings (see Figure 6–24 on page 6-60). The shape and frequency of
the phase currents and the amount of energy delivered to the motor windings
control the required speed and torque of the motor. In this case, the command
voltage or current to be applied to the motor is the modulating signal. The fre-
quency of the modulating signal is typically much lower than the PWM carrier
frequency.
PWM Signal Generation
To generate a PWM signal, an appropriate timer is needed to repeat a counting
period that is the same as the PWM period. A compare register is used to hold
the modulating values. The value of the compare register is constantly
compared with the value of the timer counter. When the values match, a transi-
tion (from low to high, or high to low) happens on the associated output. When
a second match is made between the values, or when the end of a timer period
is reached, another transition (from high to low, or low to high) happens on the
associated output. In this way, an output pulse is generated whose on (or off)
duration is proportional to the value in the compare register. This process is
repeated for each timer period with different (modulating) values in the
compare register. As a result, a PWM signal is generated at the associated out-
put.
Dead Band
In many motion/motor and power electronics applications, two power devices,
an upper and a lower, are placed in series on one power converter leg. The
turn-on periods of the two devices must not overlap with each other in order
to avoid a shoot-through fault. Thus, a pair of non-overlapping PWM outputs
is often required to properly turn on and off the two devices. A dead time (dead-
band) is often inserted between the turning-off of one transistor and the turn-
ing-on of the other transistor. This delay allows complete turning-off of one
transistor before the turning-on of the other transistor. The required time delay
is specified by the turning-on and turning-off characteristics of the power tran-
sistors and the load characteristics in a specific application.