Standard TPU PWM functions use only one channel to produce a PWM output, but the minimum and maximum pulse widths (other than 0% or 100%) that can be obtained are limited by the latencies of TPU functions running on other channels. To produce a very short pulse, the same channel must be serviced quickly twice in succession, and this may not be possible when there is activity on other channels. Although this restriction does not present a problem in many applications, there are others, including motor control, where high resolution control of the high time is required over the full 0 to 100% duty cycle range. In motor control applications, it is also desirable to have defined timing relationships between multiple PWM signals in order to reduce ripple currents and to prevent shoot-through currents on the phase drivers.