2 Inductor Acoustic Noise and Designs

Identifying acoustic noise sources is essential. In LED drivers, the two main types are inductor-generated and capacitor-generated noise. To determine which component generates the acoustic noise, a simple method can be used: use a long solid object (such as a pencil), place one end against the inductor, and hold the other end against your ear. If noise is detected through the pencil, the inductor is the noise source.

Once the noise source is confirmed as the inductor, measure the SW node waveform and the inductor current waveform. Typically, frequency components within the 20Hz to 20kHz acoustic range can be observed in these waveforms.

Several factors can contribute to inductor acoustic noise, including a discrepancy between the inductance value and design specifications, insufficient inductor saturation current, excessive input-output voltage differential, and loop instability. Upon identification of the inductor exhibiting acoustic noise, the current waveform may be measured utilizing a current probe.

Inappropriate inductor selection can contribute to acoustic noise. As shown in Figure 2-1 , the detection of periodic spikes in the inductor current waveform is indicative of inductor saturation-induced anomalies, which subsequently result in acoustic noise generation. To mitigate this phenomenon, replacement with an inductor exhibiting an elevated saturation current is recommended, which prevents complications associated with saturation effects.

The observation of periodic dropout in the inductor current waveform can be attributed to output overvoltage or an excessive input-output voltage differential, leading to the device's inability to sustain output regulation under fixed switching frequency conditions. To address this phenomenon, it is recommended to increase the input voltage or reduce the switching frequency, thereby verifying stable output voltage regulation.

The observation of periodic dropout in the inductor current waveform may be attributed to output overvoltage or an excessive input-output voltage differential, leading to the device's inability to sustain output regulation under fixed switching frequency conditions. To address this phenomenon, TI recommends increasing the input voltage or reduce the switching frequency, thereby verifying stable output voltage regulation.

As depicted in , the inductor current waveform exhibits unstable characteristics, with low-frequency ripple components coupled into the switching frequency, resulting in acoustic noise emission from the inductor. At this juncture, the schematic design should be systematically examined with particular attention to loop compensation parameters. All TI LED drivers are typically accompanied by proprietary calculation files, which enable preliminary verification of loop response characteristics, as demonstrated in Figure 2-3.

In BOOST switching converter configurations, the phase margin should generally exceed 45 degrees while maintaining a gain margin above 10 dB. Additionally, the crossover frequency is required to surpass 1 kHz as a minimum specification. Loop response conducted on the LED driver circuit shown in Figure 2-4 revealed a phase margin of less than 45 degrees, which fails to satisfy the specified design requirements. Following iterative optimization of the loop compensation network, the recalculated loop response parameters are shown in Figure 2-5.

Furthermore, inappropriate component selection constitutes an additional potential noise source. Particular attention must be paid to thermal characteristics of inductive components, as certain ferrite core materials exhibit significant parameter drift across operational temperature ranges, which can precipitate instability under varying thermal conditions.

For comprehensive validation, full-load testing across specified temperature profiles must be conducted, with particular emphasis on low-temperature characterization. This procedure verifies that loop stability margins remain within acceptable bounds throughout the entire operational temperature range. Deviations observed during thermal cycling must be systematically addressed through iterative compensation network adjustments.