06-04-2025 06:50 AM
If you want to restore the phase of a IQ signal, the common solution in interferrometrie is the Heydemann correction.
one paper:
https://www.scilit.com/publications/ea2b1a23eba36d2ec2bde571bf1b926d
https://iopscience.iop.org/article/10.1088/0957-0233/26/8/084004
06-04-2025 09:18 AM - edited 06-04-2025 09:19 AM
@leesh63 wrote:Due to internal security issues, I cannot attach the file. I measured two sine function form DAQ, ans if I correspond one to X-axis and the other to Y- axis, I can obtain an elliptical XY graph. From this graph, I have to calculate the major axis, the monor axis, and rotation angle ( radian or degree). I attach bmp file.
Measuring the phase difference of two sinusoidal signals does not require fitting an ellipse and you can can calculate all interesting parameters from the phase and relative amplitude of the two signals directly. What else do you know about the signals beforehand? For example do you know the exact frequency already? Are you trying to simulate a digital lock-in amplifier?? Characterize the frequency response of a device? Something else?
I think this discussion would have developed very differently if you had explained the primary problem you are trying to solve, and not something tangential, operating on a transform of the primary data. Still, fitting arbitrary XY data to an ellipse is an interesting geometric problem by itself and I don't mind writing any example that involves nonlinear fitting, so thanks for the distraction. 😄