Unable to set filter width larger than 80kHz

[Colonel Hacker]

I have a hackrf one.  In gqrx, I cannot set the filter width any larger than 80 kHz (40k low cut and 40k highcut) even when I have the device sample rate set to 20 Msps.

If I use SDRSharp, I can set the filter width all the way up to 20 MHz if I set the device bandwidth/sample rate  to 20 Msps.

What am I doing wrong?  Is this a limitation of GQRX?  Do I need to make a custom flow chart in gnu radio companion to achieve a higher tuning filter/bandwidth?

Please forgive me if my terminology is off, some SDR softwares refer to the high and low cut filter as tuning bandwidth.  just to be sure I am conveying the right idea: if I am tuned to 20 MHz as the center freq. the left side would go all the way down to 10 MHz while the right side would go up to 30 MHz.  

 

[Robin Gape]

Colonel, or should I call you Hack,

the filter width set in GQRX depends on the modulation that GQRX is set to receive. For CW, the default filter widths are ~200Hz, ~500Hz, and ~2.0kHz. For WFM, the default filter widths are ~120kHz, ~160kHz, and 200kHz, extendable to 240kHz. These filter widths are generally set to useful values for the modulation being received. (Arguably, for WFM, the filter width should be capable of a wider setting for lowest distortion.) For narrow FM, the maximum filter width on the user position is 80kHz, to allow reception of FM modulated weather satellites.

Why are the filter widths limited? "The wider you open the window, the more dirt blows in!"

If you want to look at interesting digital signals, then writing your own GRC (GNU Radio Companion) flowgraph is a possible good way to do it, though best results come from clear thinking and a sound grasp of theory—as with many other enterprises. Other bits of toolkit that may come in handy include:

* inspectrum (https://github.com/miek/inspectrum)

* Universal Radio Hacker (https://github.com/jopohl/urh)

HTH, 73,

Robin, G8DQX

--

[Colonel Hacker]

Hi Robin, thank you for the reply.  I understand why the filter widths are limited by the type of demodulation, I am trying to experiment with digital signals that utilize higher bandwidth like you mentioned.  Michael Osman has a nice class detailing how to use GNU Radio companion, and I guess that is route I will need to go, or keep running SDRSharp in a virtual machine.