Nice solution, Andrew!
Gerry
On 2/17/2014 11:24 PM, Andrew Martens wrote:
Hi Rolando
Congratulations
How I can confirm that the correlator function properly?
A good way to test a correlator is to do the following;
1. Get a wideband noise source and add a low pass filter so that the
signal will fit into the bandwidth of your system.
2. Split this signal and feed it into two inputs of your correlator.
3. Capture the output of the correlator and plot the phase
(arctan(real/imag)) of the cross product of the two input signals.
4. The phase of the cross product in the bandpass of the filter should
be quite clean and close to 0. Outside of the bandpass of the filter the
phase should be noisy.
5. Now lengthen one of the two signal inputs into your system by making
the cable longer.
6. Capture the cross product of the two again. The phase in the filter's
bandpass should still be clean but should have a slope in it.
7. Making the cable difference larger should make the slope steeper.
Removing the cable difference should make the phase slope return to 0.
Cheers
Andrew
--
----------------------
Gerald R. Harp, Ph.D.
Director, Center for SETI Research
SETI Institute
