Phase Spectrum from Amplitude Spectrum

[ivotebexreni]

Hello,

Can somebody tell me if is there a way to get the phase spectrum of a signal when I just know the amplitude spectrum of the signal?

 

[MacGyverS2000]

How can you get phase from amplitude and no other info? That's like trying to get the sign of a number from absolute value data.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[ivotebexreni]

Thanks for your reply.
I'm not sure if is possible, but is not so absurd how its sound if we talk about a entire signal. For example, its possible to get imaginary part of a frequency spectrum from the real part using Hilbert transform.

 

[MacGyverS2000]

I don't think you understand what a Hilbert transformation is, then. To get one, you are simply taking a real signal and shifting it by 90 degrees... you haven't added/removed any phase info from the signal, you've simply changed your viewpoint, creating a quadrature phase. Forward/reverse Fourier transforms are applied (with negative frequencies removed), but that's beside the main point.

In your OP, you only have the amplitude... the phase information has been completely lost already.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[VEBill]

I suppose if one has 'amplitude' samples of the signal in both I and Q, then you might be able to span this chasm.

 

[dicer]

Use the proper test instrument, either frequency domain or time domain. Find the proper phase in time domain?