JAJSRZ4 January 2024 TPSM843620E
PRODUCTION DATA
The first step is to decide on a switching frequency. The TPSM843620E can operate at five different frequencies from 500kHz to 2.2MHz. The fSW is set by the resistor value from the FSEL pin to ground. Typically, the highest switching frequency possible is desired because this switching frequency produces the smallest design size. A high switching frequency allows for smaller inductors and output capacitors compared to a power supply that switches at a lower frequency. The main tradeoff made with selecting a higher switching frequency is extra switching power loss, which hurts the efficiency of the regulator.
The maximum switching frequency for a given application can be limited by the minimum on-time of the regulator. The maximum fSW can be estimated with Equation 4. Using the minimum on-time of 40 ns and 18V maximum input voltage for this application, the maximum switching frequency is 1390kHz. The selected switching frequency must also consider the tolerance of the switching frequency. A switching frequency of 1000kHz was selected for a good balance of design size and efficiency. To set the frequency to 1000kHz, the selected FSEL resistor is 11.8kΩ per Table 6-1.
Figure 7-2 shows the maximum recommended input voltage versus output voltage for each FSEL frequency. This graph uses a minimum on-time of 40ns and includes the 10% tolerance of the switching frequency. A minimum on-time of 40ns is used in this graph to provide margin to the minimum controllable on-time to make sure pulses are not skipped at no load. At light loads, the dead-time between the low-side MOSFET turning off and high-side MOSFET turning on contributes to the minimum SW node pulse-width.
In high output voltage applications, the minimum off-time must also be considered when selecting the switching frequency. When hitting the minimum off-time limits, the operating duty cycle maxes out and the output voltage begins to drop with the input voltage. Equation 5 calculates the maximum switching frequency to avoid this limit. This equation requires the DC resistance of the inductor, RDCR, selected in the following step. A preliminary estimate of 10mΩ can be used but this must be recalculated based on the specifications of the inductor selected. If operating near the maximum fSW limited by the minimum off-time, the increase in resistance at higher temperature must be considered.