Filter selectivity shape factor of the SRM SDR radio receiver

Follow Quadrus SDR

SDR radio selectivity

Band filtering is responsible for selecting the signal of interest from the incoming radio spectrum in a radio receiver. SDR radio receivers perform this step by employing digital signal processing as opposed to traditional crystal (ladder) filters. This make it possible to select the necessary channel bandwidth more freely, which is the main feature of the software-defined technology. In order to change the receiver, simply a new set of algorithm parameters have to be chosen.

Recently, I’ve looked at the IF selectivity of the SRM SDR radio receiver. The industry standard way to characterize selectivity is known as the shape factor. It is the ratio of the -6 dB and -60 dB bandwidth of the filter; a single number as a figure of merit for selectivity.

Performance of traditional and SDR radio receivers

A good list of traditional crystal filter performances can be found here:

A great discussion dealing with the shape factor of digital filters in professional SDR radios can be found here:[settings][filter_by]=all&topic-reply-list[settings][reply_id]=13829055#reply_13829055

So, let us now see how the SRM is performing, but first a couple words on the testing methodology. Traditionally, filter response measurements require a sweep signal and a tracking analyzer to measure the amplitude level at various frequencies. In case we have no sync capability, we may use the maxhold function of our spectrum analyzer to keep readings for different inputs on screen. After a while this will plot the transfer characteristic on the display. As we have no maxhold on the SRM display (I’ll request it…), I had to find an other way.

I ended up using the built-in internal generator of the SRM receiver, more specifically, its noise generation capability. I switched off all other signals, and I’ve increased the level of noise to -45 dBm/Hz level. This is a wide band signal, so the filter output display showed the filter response itself. With averaging turned on (yes, this traditional spectrum analyzer feature we do have in the SRM) we can evaluate the filter’s shape factor.

The shape factor results for 4, 2, and 0.1 kHz are respectively 1.1, 1.2, and 2.

2k 4k k1


Share Quadrus SDR

2 thoughts on “Filter selectivity shape factor of the SRM SDR radio receiver”

  1. It is puzzling that people are not concerning themselves with the real performance concern about selectivity.

    Selectivity achieved before the A to D converter is very much more effective in dealing well with strong nearby interference, which is the only real purpose or value in selectivity. Specsmanship games can be played with digitally achieved selectivity, but to no real advantage. If a large signal cannot be filtered out before the A to D, then the only way to not clip the signal is to give the lions share of the bits of resolution to the strong interference, leaving very few bits of resolution left over for the smaller desired signal, possibly none if the signal is small enough, relative to the interference and the resolution of the A to D. No amount of steep filtering after the A to D, can restore resolution to the desired signal. So why should anyone care about digitally generated selectivity? It accomplishes little to nothing of any real value in improving the S/N ratio of the desired signal reception, in the presence of a strong interference.

Leave a Reply

Your email address will not be published. Required fields are marked *