JAJSC54E May 2006 – January 2024 TPS28225
PRODUCTION DATA
The dead-time control circuit is critical for highest efficiency and no shoot through current operation throughout the whole duty cycle range with the different power MOSFETs. By sensing the output of driver going low, this circuit does not allow the gate drive output of another driver to go high until the first driver output falls below the specified threshold. This approach to control the dead time is called adaptive. The overall dead time also includes the fixed portion to ensure that overlapping never exists. The typical dead time is around 14 ns, although it varies over the driver internal tolerances, layout and external MOSFET parasitic inductances. The proper dead time is maintained whenever the current through the output inductor of the power stage flows in the forward or reverse direction. Reverse current could happen in a buck configuration during the transients or while dynamically changing the output voltage on the fly, as some microprocessors require. Because the dead time does not depend on inductor current direction, this driver can be used both in buck and boost regulators or in any bridge configuration where the power MOSFETs are switching in a complementary manner. Keeping the dead time at short optimal level boosts efficiency by 1% to 2% depending on the switching frequency. Measured switching waveforms in one of the practical designs show 10-ns dead time for the rising edge of PHASE node and 22 ns for the falling edge (Figure 9-1 and Figure 9-2 in the Application Section of the data sheet).
Large non-optimal dead time can cause duty cycle modulation of the DC-to-DC converter during the operation point where the output inductor current changes its direction right before the turn ON of the high-side MOSFET. This modulation can interfere with the controller operation and it impacts the power stage frequency response transfer function. As the result, some output ripple increase can be observed. The TPS28225 driver is designed with the short adaptive dead time having fixed delay portion that eliminates risk of the effective duty cycle modulation at the described boundary condition.