Transition time analysis of sine wave while changing LO frequency from f0 to f1

[iselha]

Hi guys,

I have a sine wave and by using LO freq I have changed to new LO freq.. And of course we have a time that transition between f0 to f1.
I need to learn which methods can be used to determine this time by using attached wave data


.

Note: Please just consider one plot (green or yellow).

Thanks.

 

[SherpaDoug]

Consider the time it takes one signal to die out and the other signal to come up. These happen simultaneously so the transition time is whichever is longer.

 

[iselha]

To detect how many microsecond need to settling down, I started to analyze with spectrogram. For the each time step, I observe ratio of expected frequency component power to total power at all frequency range. I tried to decide at which time (...10^-6 seconds) expected frequency component has stable power value.
However, I thought maybe this is not reliable way.

 

[albbg]

I presume the two plots refer to I and Q signals. You could demodulate the signal using a FM demodulator and extract the instantaneous frequency. You should see stable frequency fo, "strange variation of frequency, stable frequency f1. From this shouldn't be difficult to calculate the transition time. This method is independent from the amplitude of the signal.

 

[FvM]

In the present example (single clean sinewave without noise), the frequency can be easily determined by measuring the period of zero crossings. The method is simple and fast.

- - - Updated - - -

A slightly more complex method would calculate the instantaneous signal phase from I and Q voltages and calculate frequency as dphi/dt. In the present case, the direction of rotation changes during frequency transition, in other words the frequency sign is changed.