Strategies for Mitigating Noise in Audio Devices

In audio technology, impeccable sound quality is a fundamental objective. Nevertheless, undesired auditory disturbances, such as hissing, buzzing, or interference, can considerably impair the overall sound quality. These disturbances hold special significance in the context of headphones and microphones since users seek accurate and unaltered sound replication.

This article examines different approaches to reducing unwanted noise in audio devices such as headphones and microphones. The TDK Audio Sample Kit is referenced as an example of a solution that provides all the components required for noise suppression and ESD countermeasures for microphone lines without degrading sound quality.

The rise of Bluetooth and TWS

Bluetooth technology was originally intended for hands-free communication. That said, Bluetooth applications rapidly grew to include a variety of devices such as headsets, speakers, car systems, and more. The technology's low energy consumption and universal compatibility have made it an indispensable component of the constantly expanding ecosystem of connected devices.

True Wireless Stereo (TWS) emerged after Bluetooth became the de facto standard for wireless audio transmission. Taking the idea of wireless audio one step further, TWS earphones untethered each earpiece. This was the start of a new age in portable music. The tiny, wire-free earphones represented a trend toward simpler, more portable music equipment. TWS technology liberated consumers, allowing them greater mobility and convenience.

Many of the latest trends in music and audio consumption depend on smartphone services such as wireless content streaming to Bluetooth speakers and earbuds. Although speakers and earbuds have become the standard for audio output, obtaining flawless sound quality in audio devices like Bluetooth earbuds, speakers, and voice assistant microphones presents a few obstacles.

Issues affecting wireless audio devices

Audio equipment free of wired connections is convenient in many ways. However, because these devices rely on a wireless signal, they are more likely to experience issues than wired headphones, microphones, or speakers.

In wireless devices, transmission, reception, device performance, and battery life are all affected by the quality of the RF link. Whenever RF capability is integrated into small wireless devices, the PCB traces and the wiring interconnections for each audio input and output are typically placed close to the antenna. Due to this proximity, when audio is sent to the microphone or speaker, the RF signals emitted by the antenna can create EMI noise and decrease the audio quality. This issue, commonly known as crosstalk, affects the signal integrity.

Similarly, the switching that occurs in digital amplifiers used in battery-powered portable music equipment can emit noise, creating multiple harmonics. These harmonics pose a threat to the antenna's output and input RF signals. Since the antenna and the wire are so close together, coupling occurs, resulting in decreased receiving sensitivity. All these possible EMI noise sources are shown in Figure 1.

Figure 1: A typical wireless audio configuration with potential noise sources. (Source: TDK)

Mitigating RF noise in speaker lines

When using Bluetooth Classic Audio, as opposed to BLE audio, devices exchange data at regular intervals. When an RF signal is fed into an audio amplifier, an envelope waveform is produced due to non-linear effects. This envelope waveform is detectable as background noise when it is transmitted to the speakers along with the intended signal. This type of noise is commonly referred to as Time Division Duplexing (TDD) noise, Time Division Multiple Access (TDMA) noise, or simply “buzz” noise.

This difficulty with the RF radio envelope waveform manifests not only in Bluetooth applications, but also in cellular networks and Wi-Fi. During a phone call, GSM modules generate an RF burst transmission every 4.615 ms. When radiated to an acoustic circuit, the envelope waveform of the RF burst can produce audible TDMA noise at a frequency of 217 Hz along with related harmonics (Figure 2).

Figure 2: How TDMA noise is generated in GSM communication. (Source: TDK)

A standard wired connection between a speaker and a Bluetooth SoC is shown in Figure 3. Here, the wired connections pick up the RF signal and propagate it to the SoC.

Figure 3: An RF signal affecting the audio on wired speaker lines. (Source: TDK)

Therefore, it is necessary to filter out the audible noise produced by the RF envelope waveform and any RF signals picked up by the antenna circuit before they are fed into the speaker. Reducing the strength of the Bluetooth RF signal (2.4 GHz band) that generates the envelope waveform is the key mitigation strategy. Mitigation is achievable through a thorough understanding of small passive filters and careful study. Noise can be reduced by filters like those found in TDK's MAF series.

Chip beads are usually used to reduce the background noise in audio cables. They are made of a coil laminated onto the inside of a ferrite core. The impedance of a chip bead is defined in terms of the reactance and AC resistance of the coil. The reactance component is mostly responsible for noise reflection in the low-frequency range, whereas the AC resistance component is primarily responsible for noise absorption and heat generation in the high-frequency range.

TDK has created a novel ferrite material that is both low-distortion and effective in canceling out noise. The MAF series of multilayer chip components was developed in response to the emerging market for noise cancellation in the audio lines of portable electronic devices like smartphones. The letters M, A, and F in MAF respectively stand for Multilayer, High-Fi Audio, and Noise Suppression Filter.

Electrostatic discharge (ESD) protection for the wiring that connects the microphone and the speaker is also required, since TWS earphones come into physical contact with the user's hands when being used. TDK has designed a notch filter (AVRF series) to mitigate this potential issue by shielding audio signal lines from electromagnetic interference (EMI) and static discharge (ESD). The insertion loss versus frequency performance of several AVRF notch filters is shown in Figure 4.

Figure 4: Insertion loss vs. frequency for different TDK AVRF notch filters. (Source: TDK)

Combining an MAF series noise suppression filter (with its series inductor) and an AVRF series notch filter (with its series capacitor) brings about the low-pass output filter shown in Figure 5. This setup produces high attenuation characteristics in the 2.4 GHz band and prevents relevant noise from accessing the audio amplifier. As a result, the envelope waveform does not generate any unwanted noise.

Figure 5: (a) Configuration with MAF and AVRF filters, (b) FFT of the corresponding filtered signal, (c) High attenuation centered around the 2.4 GHz band. (Source: TDK)

Mitigating RF noise in microphone lines

In the same way it does with speaker lines, transposing a Bluetooth RF signal onto microphone lines results in an envelope waveform that is sent to the audio processor's input. The audio processor will then send the unwanted audible noise to the speakers. Figure 6 shows one possible route for the wireless Bluetooth signal to be converted into a wired connection within the microphone's circuit. The noise is coupled to the original audio signal after processing.

Figure 6: An RF signal affecting the audio on wired microphone connections. (Source: TDK)

To effectively minimize noise, MAF filters are a better choice than regular chip beads due to their higher impedance and lower noise attenuation in the 2.4 GHz frequency. An MAF filter can reduce audible output noise to undetectable levels by increasing attenuation at lower frequencies.

The MAF + AVRF solution prevents an increase in THD+N, in contrast to the use of ordinary ferrite chip beads and multilayer ceramic capacitors (MLCCs). There is no harmonic distortion since neither the MAF nor the AVRF components create non-linear variations in voltage or current within their respective operational ranges. When it comes to signal distortion, the MAF + AVRF solution is virtually indistinguishable from using no filter at all.

The outcome of the TWS earbuds' reception sensitivity with and without mitigation is shown in Figure 7. Approximately 6 dB of receiving sensitivity enhancement was seen after the introduction of MAF, AVRF, and MAF + AVRF countermeasures, all of which have noise reduction effects in the Bluetooth 2.4 GHz band.

Figure 7: Receiving sensitivity in TWS earbuds with and without filters. (Source: TDK)

TDK’s Audio Sample Kit

Smart appliances and consumer electronics such as smart speakers are on the rise as society moves toward the Internet of Things (IoT) and connected products. The fundamental components of smart speakers are microphones, which also operate as sound sensors, making a person's speech an interface to connect them with the device. TDK's semiconductor micro-fabrication technology was used to build a wide range of MEMS microphones for use in such contexts.

To address the need to suppress RF and ESD noise in MEMS microphones, TDK provides the Audio Sample Kit (Figure 8). This product combines TDK InvenSense MEMS microphones with MAF noise suppression filters and AVRF ESD notch filters. These filters are designed to specifically combat typical issues in audio lines, while offering additional benefits such as improving reception sensitivity in wireless or cellular communication.

Figure 8: TDK’s Audio Sample Kit. (Source: TDK)

Providing noise suppression and ESD countermeasures for speaker and microphone lines, the Audio Sample Kit includes the following components:

• 20 MEMS microphones
• 80 MAF series noise suppression filters
• 120 AVRF series ESD notch filters

The main features of the audio solution sample kit are:

• Improvement in reception sensitivity of cellular and Wi-Fi communication
• High sound quality because of low distortion due to low THD+N characteristics
• Suppression of TDMA noise
• Small signal degradation due to low resistance
• Achievement of both ESD and noise countermeasures

Conclusion

The combined use of noise suppression filters and ESD notch filters provides an effective countermeasure against the noise affecting wireless headsets and microphones. TDK’s Audio Sample Kit is a ready-to-use solution including all the components engineers can use to mitigate RF noise in their wireless audio designs without compromising sound quality.