There are 4 messages in this thread.
You are currently looking at messages 0 to 4.
Hi All, What is the utility of "demodulated noise"? Why do DEMON processing? I am working on acoustic signal processing applications involving rotating machinery. Once the carrier (generated by a rotating shaft) has been demodulated what is significant about the remaining signal? Might it contain information about the source (i.e., the machinery) such as a cavitation signature for the case of propellers on a vessels or perhaps some other telltale trait? If so, would these signals qualify as noise??? How would a time vs frequency (waterfall) plot of the FFT'd data look different from a plot for which noise demodulation was not performed? Thanks, Mikial______________________________
On Mar 11, 11:59 pm, "Mikial Chubarov" <mikialchuba...@ieee.org> wrote: > Hi All, > > What is the utility of "demodulated noise"? Why do DEMON processing? > > I am working on acoustic signal processing applications involving rotating > machinery. Once the carrier (generated by a rotating shaft) has been > demodulated what is significant about the remaining signal? Might it > contain information about the source (i.e., the machinery) such as a > cavitation signature for the case of propellers on a vessels or perhaps > some other telltale trait? If so, would these signals qualify as > noise??? > > How would a time vs frequency (waterfall) plot of the FFT'd data look > different from a plot for which noise demodulation was not performed? > > Thanks, > > Mikial The original DEMON process was intended to detect a broadband signal (cavitation impulses) generated by a cyclical process (blade passage near the keel). The signal processor performed narrowband analysis and showed no indication of the impulses. So, a detection process was performed (with a simple very low frequency DC removal) to produce the envelope that was a narrowband process suitable for the processor. The impulses were 'noise' with respect to a narrowband processor. Machinery might produce similar signals from the passage of a broken tooth or other mechanisms. The DEMON process also works for conventional sidebands. In the SONAR case, a band from 1 to 4 kHZ might be run through a detector, DC removed, low pass filtered and desampled for analysis in a band from a fraction to several Hertz. A waterfall of the 1-4 kHz DEMON input band might show no signal related to the cavitation source. The waterfall of the several Hertz wide DEMON output band might show a strong tone at the blade passage rate. This process would also detect conventional side bands within several Hertz of any carrier in the 1-4 kHz band. Dale B. Dalrymple http://dbdimages.com______________________________
Thank you Dale. Mikial______________________________
Dale wrote: ... > Dale B. Dalrymple > http://dbdimages.com Hello Dale, nice photos! Here in Switzerland, snowboarding season is drawing close, and I haven't yet managed to get my board out of the cellar. Your photos have motivated me to take one of the next days off from work and enjoy the beautiful weather in the mountains. Thanks :-)! Regards, Andor______________________________