phase sensitive detection

[gary36]

I have two signals A and B. Both are 50 Hz signals. A is of the form V1sinwt and B is V2sine(wt+phi). Now I have used multiplier and the output is DC=(V1*V2/2)*cos(phi). I am facing issue that if V1 changes then the DC output also changes giving in correct phase readings. How can I make the DC output insensitive to the magnitude of reference A?

 

[danadakk]

Handbook on multipliers and their circuits :

https://www.analog.com/media/en/training-seminars/design-handbooks/ADI_Multiplier_Applications_Guide.pdf?doc=AD633.pdf

Not sure where offset comes from....

Regards, Dana.

 

[gary36]

In page no: 10 (pdf page no.15), the average output from phase detection is proportional to V.Vr cos(phi), which means that slight magnitude change in either reference or incoming signal will result in incorrect phase readings.

 

[BradtheRad]

Sense moments of peak amplitude. That way you ignore DC component as well as signal intensities.

 

[gary36]

Sense moments of peak amplitude

Could you please elaborate on this. Do you mean peak detection?

 

[crutschow]

One option is to use an AGC circuit to keep the A signal amplitude constant (which can be done with another multiplier).
Or use another multiplier to adjust the DC output gain (signal level) inversely proportional to the A signal amplitude.

--- Updated Mar 27, 2023 ---

Could you please elaborate on this. Do you mean peak detection?

A possible disadvantage of peak detection (or zero-crossing detection) for phase-shift measurement, is that they are rather sensitive to any signal noise or distortion, whereas the multiplying technique is generally less so.

 

[BradtheRad]

Could you please elaborate on this. Do you mean peak detection?

Exactly. This simple method works for any amplitude or DC component.
It indicates bursts of current at each moment incoming voltage exceeds charge on the capacitor.