Sample Rate and audio underflow ....

[Rob Roschewsk]

Hi all!

I'm working on a "Hello, World!" flow in GRC .... just a simple NBFM receiver tuned to my local NOAA station.

Using an RTLSDR and the osmocom block ... into a low-pass that decimates by 5 into the NBFM block and then an audio sink. It doesn't get any more simple. Here is a image of the flow http://pabut.org/wiki/images/a/aa/My-fmrcv.grc.png

So with the sampling rate set to 2M .... works fine ..... if I REDUCE the sample rate, say 240k, I start to get audio sink underruns "aUaUaUaUaUaUaUaUaU" and I only get small blurbs of audio.

I'm sure to set the sample rate in the downstream blocks appropriately.

240k samples/sec should be plenty for a 5khz deviation FM signal ... right??

This is completely counterintuitive to what I was expecting.  I thought by reducing the sample rate I would lower CPU consumption ... and mitigate the exact problem I'm experiencing at the lower rate.

What am I not grocking here??

Thanks,

--> Rob