Lou,
Apologies for the late reply.
> Mike, thanks for the detailed explanation. Is there a reason for doing the
> channel selection via FFT/IFFT method as opposed to a frequency xlating
> filter? Maybe it's more efficient since you needed to do the FFT anyway for
> the spectrum sense.
I utilised the FFT/IFFT method throughout because the maximum signal was being identified and passed on as a single FFT bin number which is then set as the central bin number for the 'Keep M in N' block. In this 'Keep M in N' block, the FFT bins to the right and left are used to make up the correct channel bandwidth so that the wanted sampling rate is obtained when the IFFT takes place to convert the FFT frequency domain vector stream to the time domain.
> I see the frequency selection with the mouse is a feature of the WX FFT sink
> (Freq Set Varname). I wonder if the same functionality is in the QT sink.
It looks as though in GNU Radio 3.7.11.1 the feature to set a GRC variable with the 'click to tune' function is broken in both the 'WX FFT GUI sink' and the 'QT FFT GUI sink' blocks. Fortunately, the 'QT FFT GUI sink' block has a message output which sends a message when a new frequency is clicked on. To utilise this message, I've created a 'message_to_variable' block which receives a message from the 'QT FFT GUI sink' block containing the new frequency information and then the center_freq GRC variable is updated by an associated 'function probe' block.
> Anyway, yours is a good example to wrap my head around how the Probe Signals
> work. Looks like it may be a better method than the message queues I was
> experimenting with. I was using a queue to grab the FFT data and compute
> the max value.
I hope you have been successful with your message queue experiments. A Scanoo QT version is in progress which will utilise message sending functionality. One essential on the list is to add a message output to the QT push button so that when clicked, I can use the message to trigger something.
Regards,