Welcome Customer !

Membership

Zhejiang Saikong Electrical Technology Co., Ltd.
Free Member ·1YRS
,
Other

Main Products: 变频器

Xwboo>Technical Article>Do you know how to deal with the harmonic problem of the inverter?

Zhejiang Saikong Electrical Technology Co., Ltd.Free·1YRS

  • E-mail

    amy@safeinvert.com

  • Phone

  • Address

    #22 Liujiang Avenue, Liushi Town, Yueqing City, Zhejiang Province, China

Contact Now
Do you know how to deal with the harmonic problem of the inverter?
Date: 2022/11/3Read: 35

Do you know how to deal with the harmonic problem of the inverter?

Mar 17, 2022

Do you know how to deal with the harmonic problem of the inverter?

The input and output of the frequency converter (hereinafter referred to as -VFD) are not standard sine waves, especially the output PWM wave contains a large number of high-order harmonics, and there is a large error when using a general power analyzer and harmonic analysis method.



1. Insufficient sampling frequency leads to VFD harmonic analysis errors


When analyzing the VFD output waveform, the signal should be sampled first. According to the sampling theorem, the sampling frequency fs must be greater than twice the highest frequency fm of the signal (fs≥2fm) to obtain the full spectrum of the signal. when fs


Second, the actual bandwidth of the measuring instrument is not enough to cause the VFD harmonic analysis error.


In order to avoid aliasing, the common practice of general power analyzers is to add an anti-aliasing filter, limit the signal bandwidth, and filter out the signal higher than fs/2 in the VFD output PWM wave. The bandwidth of the antialiasing filter is the actual bandwidth of the analyzer. This method of filtering out the harmonics first and then analyzing the harmonics has a very large error, and it can even be said to be wrong.


3. Analysis algorithm VFD harmonic analysis error reasons


At present, most analyzers use the FFT algorithm for harmonic analysis. With FFT harmonic analysis, the signal is first intercepted with a data sequence, then the signal is sampled, and finally the signal is discretized in the frequency domain by FFT transform. Due to the characteristics of FFT, the data sequence required to be intercepted must contain an integer number of signal periods, and the number of sampling points in the data sequence must be 2 to the nth power. In fact, the signal period used for the FFT transform is the last detected signal period. Because the signal period of the VFD output is not completely stable, errors are bound to occur. As can be seen from the requirements of the FFT, the limitations of the FFT algorithm are obvious: if the data sequence intercepted from the signal is not an integer number of signal periods, it will cause spectral leakage and severe errors.


When using FFT harmonic analysis, spectral leakage affects the measurement accuracy of harmonic analysis, and the most fundamental cause of spectral leakage is the error of period synchronization in sampling. In order to deal with the phenomenon of spectral leakage, the general analyzer adopts a window function (such as Hanning window, Blackman window, etc.), which only suppresses the spectral leakage to a certain extent, which can be said to bring new errors.


4. Harmonic analysis error caused by VFD measurement accuracy


The VFD harmonic analysis is based on the measurement of the electrical signal, and the error of the power analyzer's measurement of the electrical signal will inevitably lead to the error of the analysis result. To reduce the impact of measurement errors on harmonic analysis, choosing a truly high-accuracy power analyzer is key.


5. Calculation error


After the electrical signal is sampled into a digital signal that can be processed by the power analyzer, it is necessary to calculate the RMS, average, power, total harmonic distortion, harmonic factor and other related parameters according to the formula, which will lead to calculation errors.


Sixth, choose the right power analyzer.


Depending on the source of error in VFD harmonic analysis, the following requirements can be placed on the power analyzer:


The sampling frequency should be high enough to meet the requirements of VFD harmonic analysis, not less than 200kHz.


The bandwidth should be 6 times higher than the switching frequency of the VFD, and the bandwidth should not be lower than 100kHz.


The harmonic analysis algorithm uses the DFT algorithm and does not require the entire period and number of points of the data sequence.


With powerful hardware support, it can meet the requirements of DFT algorithm for calculation speed and amount of calculation.


Use a high-precision power analyzer with well-defined error reference conditions, including actual usage conditions.