FFT Analyzer Basic FAQ
Question number FAQ-0769
What are PSD and ESD, and how are they different from power spectrum?
PSD (P ower S pectral D The density function (power spectral density function) is the frequency resolution Δ of the FFT calculation. f A spectral function that expresses power values per unit frequency width (1 Hz width) in a way that is independent of the signal, is often used to evaluate non-periodic signals that have a continuous spectrum, i.e., irregular signals (random signals). Its unit is the signal x (t If the physical unit of ) is EU, then EU 2 It will be /Hz. Specifically, if the vertical axis is voltage (V), then V 2 /Hz or its square root V/ √Hz Therefore, it is often used as an evaluation value for noise in amplifiers and other devices. Also, in the field of random vibration testing, the vertical axis is vibration acceleration (its unit is m/s²). 2 Therefore, ASD (A Acceleration S pectral D Also known as ensity (acceleration spectral density), its unit is (m/s²). 2) 2 /Hz, or m 2 /s 3 This is the result. Please refer to the related items below for the calculation method.
ESD (E nergy S pectral D The energy spectral density function (Essence Function) is a function of time signal x (t This is a function that represents the frequency distribution of the energy of a transient signal, and is used for spectral evaluation of transient signals such as impulse signals. The method for obtaining it is spectral-related. FAQ: "What is signal power?" In this section, the averaging operation is omitted, so the PSD value is equal to the FFT time window length T It can be calculated by multiplying by [a certain factor].
Table 1 Comparison of three types of spectra
| Types of spectra | Physical meaning | Target signal |
|---|---|---|
| Power Spectrum | Power distribution by frequency band | periodic signal |
| PSD | Power distribution per unit frequency | Continuous random signal |
| ESD | Energy distribution per unit frequency | transient signal |
(Note) PSD and ESD are mainly used for vibration analysis and are not commonly used for noise analysis.
Last updated: 2009-11-16