Hi,
one thing came to my mind: Don't average the phase, average the complex
value instead. If you have a lot of noise, the phase will jump around
and will also wrap around 2*pi, which will "confuse" the averaging.
After averaging the complex value, convert it to the phase and you are done.
Fabian
Am 12.10.21 um 15:53 schrieb mehtap özkan:
Dear All,
I try to measure the phase difference between a delayed signal and a
reference one. I use the following blocks:
Multiply conjugate block (Inputs are the delayed and reference signals)
Complex to Arg
Moving Average
The flowgraph works good when there is no noise. However when I add
noise the variations are pretty wild.
These blocks constitute the basic structure of a PLL's phase comparator
which should work under noisy Input conditions.
What am I doing wrong?
