Test Case Study
- Updated2022-11-09
- 5 minute(s) read
Test Case Study
Refer to the following test case study to learn how to find root causes and solutions for abnormal test results.
High THD
Analyze the root cause of a high THD and find solutions to avoid the high THD.
The following figure shows the THD of a speaker measured with a reference microphone at 94 dB SPL and 100 dB SPL. The maximum THD under 94 dB SPL is about 2%, which is higher than the speaker specification provided by the speaker vendor.
The speaker THD is vulnerable to standing waves and noise. The following table analyzes the possible causes of the high THD focusing on standing waves and noise, provides verification methods, and describes expected results.
| Possible Cause | Verification Method | Expected Result |
|---|---|---|
| Standing waves heavily attenuate the fundamental frequency. Meanwhile, the harmonics do not suffer attenuation or even get a boost. | Check the speaker frequency response at the frequency where the speaker THD peaks. | The speaker frequency response is low at the frequency where the speaker THD peaks. |
| The noise weakens the fundamental frequency, which increases the speaker THD. | Prolong the sweeping duration of stimulus signals. | The speaker THD becomes lower. |
To determine which possible cause is true, measure the speaker frequency response at 94 dB SPL and 100 dB SPL. The following figure shows the measurement results.
The speaker frequency response is low at the frequency where the speaker THD peaks.
Prolong the sweeping duration of stimulus signals from 2.72s to 16.6s and then measure the speaker THD at 94 dB SPL. The following figure compares the measurement results.
The speaker THD changes slightly after the sweeping duration of stimulus signals is prolonged.
The previous measurement results indicate that standing waves cause the high THD. A fixture constitutes a surface parallel to the generation surface in the chamber, which results in reflections and generates standing waves.
To further verify that standing waves is the root cause of the high THD, remove the parallel surface, cover all the flat hard surfaces in the chamber with acoustic absorbent materials, and then measure the speaker THD. The following figure compares the speaker THDs before and after changes.
The speaker THD greatly decreases across 1 kHz to 2 kHz after changes.
Therefore, the root cause of the high THD is standing waves. To avoid a high THD, you need to minimize reflections by removing hard surfaces, especially parallel surfaces, in the chamber.
Large Frequency Response Difference
Analyze the root cause of a large frequency response difference between reference microphones and find solutions to avoid the large frequency response difference.
In a microphone test, two reference microphones are placed side by side. The reference microphones receive the same signals. The following figure shows the frequency responses of the reference microphones.
The following figure shows the difference between the frequency responses of the reference microphones.
The frequency response difference is large, around 1.6 kHz and between 4 kHz and 5 kHz. The following table analyzes the possible causes of the large frequency response difference and provides verification methods.
| Issue | Possible Cause | Verification Method |
|---|---|---|
| Large frequency response difference, around 1.6 kHz |
Standing waves may decrease the frequency responses of the reference microphones around 1.6 kHz to the minimum value, around −42 dB. Even a slight difference in positioning significantly affects the frequency responses of the reference microphones. |
Avoid standing waves. |
| Large frequency response difference, between 4 kHz and 5 kHz |
The distance between the reference microphones has a bigger impact on the frequency responses at higher frequencies. For example, the wavelength at 3.4 kHz is 100 mm due to the speed of sound 340 m/s. Even a distance of 20 mm between the reference microphones leads to a big difference between signals received by the reference microphones at 3.4 kHz. |
Reduce the distance between the reference microphones. An ideal distance is shorter than or equal to 5 mm. |
To verify the possible causes, cover the flat hard surface above the reference microphones with soft acoustic absorbent materials, place the reference microphones face to face, and then measure the frequency responses of the reference microphones. The following figure shows the measurement results.
The following figure shows the difference between the frequency responses of the reference microphones after changes.
After the previous changes, the frequency response difference is consistent with the microphone specifications provided by the microphone vendor.
Therefore, the possible causes are true. To avoid a large frequency response difference, you need to reduce resonances or standing waves and place the reference microphones as close to each other as possible.