Spread spectrum is a factory option. To find which devices have spread spectrum enabled, see Section 5. The purpose of spread spectrum is to eliminate peak emissions at specific frequencies by spreading these emissions across a wider range of frequencies rather than a part with fixed frequency operation. In most systems containing the chip, low frequency-conducted emissions from the first few harmonics of the switching frequency can be easily filtered. A more difficult design criterion is reduction of emissions at higher harmonics that fall in the FM band. These harmonics often couple to the environment through electric fields around the switch node and inductor. The device uses a ±2% spread of frequencies that can spread energy smoothly across the FM and TV bands, but is small enough to limit subharmonic emissions below the device switching frequency. Peak emissions at the switching frequency of the part are only reduced slightly, by less than 1 dB, while peaks in the FM band are typically reduced by more than 6 dB.

The device uses a cycle-to-cycle frequency hopping method based on a linear feedback shift register (LFSR). This intelligent pseudo-random generator limits cycle-to-cycle frequency changes to limit output ripple. The pseudo-random pattern repeats at less than 1.5 Hz, which is below the audio band.

The spread spectrum is only available while the clock of the device is free running at their natural frequency. Any of the following conditions overrides spread spectrum, turning it off:

• The clock is slowed during dropout.
• The clock is slowed at light load in auto mode. In FPWM mode, spread spectrum is active even if there is no load.
• At a high input voltage-to-low output voltage ratio when the device operates at minimum on time, the internal clock is slowed disabling spread spectrum. See Section 7.6.
• The clock is synchronized with an external clock.