How Can You Measure Phase Angle with Distorted Current Without Filters?

[hisham.i]

Am trying to measure the phase angle between voltage and current, so i thought about zero crossing detection method, but the problem is that the current i am sensing is highly distorted, so i am trying to find a solution without using filters.

After searching i found a phase meter on the internet that works using multiple zero crossing detection but i didn't understand how this method works.

If anyone has idea of how this algorithm works please share your ideas.

This is the link of the phase meter that i found
http://www.powertekuk.com/phasemeter.htm

Thanks

 

[Mbert]

The angle you are looking for is the angle between the fundamental components of voltage and current. You need to look at Fourier transform theory. This is the method used by the phase meter you are referring to (Discrete Fourier Transform).

M.

 

[hisham.i]

But what can i understand from multiple zero crossing detection?, how does this algorithm works?

This word is mentioned in the picture to the right.

 

[uart]

The usual methods are either,

1. BPF and zero crossing detect.

Or

2. DSP and FFT (as Mbert suggested).

The DSP method is more reliable as analog BPF's with high Q can have a significant phase error if you're off the center frequency by even a small amount. So it's kind of a no win situation if you've got a lot of distortion or noise. You either use a low Q filter with small phase error but suffer from the noise and distortion not fully removed, or you use a high Q filter which adequately removes the noise and distortion but introduces it's own errors if there is even a slight center frequency error/drift.

One way I have seen the above (non DSP) problem reasonable well handled however was with the use of a switched capacitor filter that had it's clock input frequency locked to that of the voltage signal, so that the center frequency always tracked the signal frequency. (that one was designed and built by me btw. )

 

[Mbert]

There is also a more recent approach that uses DWT (discrete wavelet tranform), but you're probably better off with just a Discrete Fourier Transform. Forget about the multiple zero crossing detection.

M.