Sound quality metrics

Description

Manatee e-NVH software can compute the Sound Pressure Level and synthesize the sound of the electrical system, combining magnetic noise produced by the electrical machine as well as imported non-magnetic noise. Total sound signature can then be analyzed using Sound Quality metrics to accurately quantify the perceived sound attributes, beyond the simple Sound Power Level value. 

Besides A-weighting, the following psychoacoustic metrics are available at the end of every simulation:

  • Loudness and specific loudness (Zwicker method for stationary signals)
  • Sharpness (DIN 45692)
  • Roughness and specific roughness (Daniel & Weber method)

 Sound quality metrics are calculated using open-source library Mosqito. If some sound quality metrics are missing for your specific application, you can contribute to the open-source project.

Application

The physical Sound Power level is sometimes insufficient to quantify the acoustic annoyance of an electric drive. In some cases, e-motor prototypes are rejected by a customer just because it sounds more unpleasant than the competitor prototype. This can be due to tonalities or modulation effects introduced by eccentricities or Pulse Width Modulation. Electric drives are particularly prone to tonalities as magnetic forces occur at some particular frequencies

Some standards therefore use psychoacoustic indices to quantify more objectively the subjective perception of sound. As an example, tonalities are penalized in wind turbine applications and railway applications. Sound quality metrics enable acoustic engineers or NVH engineers to quantify and analyze system-level acoustic annoyances.

Demonstration

Gallery