10

Why do many SDR control apps have two frequency settings, even when the app is designed or set up only for listening to or decoding one signal? Should the two SDR frequency controls be set differently or the same? If different, how much for which modulation schemes (CW, USB, LSB, AM, etc.), in which direction, and why?

Kevin Reid AG6YO
  • 24,506
  • 7
  • 51
  • 103
hotpaw2
  • 13,316
  • 6
  • 45
  • 76
  • 1
    What apps have 2 frequency settings? Screenshot? – Phil Frost - W8II Feb 09 '14 at 21:55
  • @Hotpaw2 Could you edit the question to make it more clear what you are asking? As it is, it's too broad for us to attempt a canonical answer. Thanks. – WPrecht Feb 09 '14 at 23:58
  • 1
    I think there's a little difference of background assumptions involved here: as far as I know, all of the apps that RTL-SDR users might use, i.e. host-computer-using software-defined radios, have two frequency controls, and what this question is referring to is obvious from that perspective. I suppose that (probably dedicated hardware) SDRs without panadapters might not have such controls since they have less use for them, but I'm not as familiar with that type. – Kevin Reid AG6YO Feb 10 '14 at 01:52
  • @KevinReid You can guess, if you want. The point of putting questions on hold is to discourage that until the question is edited to make it clear. – Phil Frost - W8II Feb 10 '14 at 13:02
  • The question will purposely be left unclear because I have seen this type of SDR control interface on more apps than I can name (ones that don't run on my Mac for instance). – hotpaw2 Feb 10 '14 at 16:18
  • 1
    @hotpaw2 I think the objections would be reduced, and the question would be improved, by including a screenshot, say, of any one of those apps. You can say this is just one example. – Kevin Reid AG6YO Feb 10 '14 at 16:30
  • @hotpaw2 I'm sure you can find screenshots that other people have made, if you can't figure out how to make your own. Even if you just tell us the name of the software and the name of the two frequency settings, the specificity of your question will be greatly improved. Someone might even be nice and edit a screenshot into the question for you. – Phil Frost - W8II Feb 10 '14 at 19:31
  • 1
    I made up a cropped screenshot of Gqrx I think would illustrate the question well: http://i.stack.imgur.com/FYsbR.png I'll leave it to you whether/how to edit it into your question. – Kevin Reid AG6YO Feb 11 '14 at 00:55

1 Answers1

12

One controls the hardware, and the other controls the software.

  1. The hardware selects some section of the entire RF spectrum (by a local oscillator and mixer), and down-converts it into a frequency an analog-to-digital converter can handle, filters it (to discard out-of-band signals), samples it, and delivers that data to the computer.

    This data determines what you can see in the so-called “waterfall” or “panadapter” displays. It also sets the limit for the widest-bandwidth signal you can possibly demodulate.

  2. Then the software does a similar process in order to select a single signal to demodulate; it shifts it to baseband (a “0 Hz” signal which only varies according to the modulation), applies a low-pass filter, and demodulates appropriately.

Both stages have their own controllable local oscillator, and those two controls are what you are seeing.

Here are some reasons for there to be these two separate stages:

  • Analog hardware is imperfect; there are various sorts of garbage you can get in the digital signal (DC offset, IQ imbalance, insufficiently filtered out-of-band signals). You can change the hardware center frequency to shift the garbage away from the signal of interest.

    For example, if there is a DC offset (RTL-SDRs with E4000 tuners do) then you have garbage at the center frequency, so you would set it to be slightly different from the signal of interest so that the following digital filter removes that part. This is one of the main reasons to specifically set the two frequencies differently, if your receiver has this problem. It doesn't matter which direction you offset, as long as the offset is enough that the bandwidth of the desired signal doesn't overlap the unwanted signal.

    On the other hand, filtering is imperfect and the way this shows up in the signal from the hardware is that signals which are out of the tuned band, but strong, will be seen to “wrap around” and appear at an in-band frequency modulo the hardware bandwidth (sample rate). This is a reason to receive close to the center; the hardware filtering is at its best at that point.

    I highly recommend playing with changing the (hardware) center frequency slowly and watching and listening to how the displayed signals change or don't. You will learn what to do.

  • If you are interested in monitoring a whole band rather than a single station (e.g. a single amateur HF band) then keeping the hardware settings fixed leaves your waterfall display undisturbed to watch activity over the whole band while you are free to select which stations you are currently demodulating (listening to).

  • Digital filters can be optimized for the particular mode in use, and, if you don't care about power consumption, be extremely sharp (good at selecting exactly what is wanted) compared to analog filters. They can also be adjusted in bandwidth or shape to trade off filtering out nearby unwanted signals vs. better intelligibility in the absence of nearby signals. (Note that this is a reason for the two-stage architecture, but it isn't directly a reason to have independent tuning controls as you've asked.)

Kevin Reid AG6YO
  • 24,506
  • 7
  • 51
  • 103