SDR for Spectroscope

Question

I'm looking into getting a cheap SDR (perhaps the RTL-SDR RTL2832U) to use as a spectroscope, especially to test the frequency output of Baofeng radios. I currently have reason to believe that one of my radios, on high power in 2m, transmits most of its power below 140 MHz but I'd like to characterize where.

I'm fine writing software/scripts on Linux to do this, and I've found that there's apparently an rtlsdr toolkit for this. However, I can't find whether or not it can give me an AM-demodulated signal (i.e. I want samples of data within, say, 1 MHz of the specified frequency).

Does anybody have experience (or software!) doing this? Or have an idea of which devices have which capabilities?

Answer 1

You could just use an RTL-SDR and any SDR software (Gqrx, SDR#, etc) that supports it, and manually scroll around 140 Mhz until you see the signal (the RTL-SDR allows you to view 3.2Mhz max at one time). To not pick up other people's transmissions/other noise, use a few inches of 22ga solid wire stuck in the center of the SMA port of the RTL-SDR for a really poor antenna. And most SDR software allows you to demodulate AM or FM or SSB or CW or... :)

I have used this method to see how clean the oscillator in a microcontroller is, with great success.

Answer 2

If you are concerned about checking for a wide variety of potential spurious emissions, you will want to look into the many software packages for using the RTL-SDR as a spectrum analyzer. The RTL-SDR.com blog has a number of articles on this, e.g. http://www.rtl-sdr.com/rtl-sdr-spectrum-analyzer/.

What these do is programmatically tune across a whole wide range of frequencies and sample the spectrum at each "stop" along the way. So instead of seeing only distortion products in the ~2MHz spectrum around one particular center frequency, you can get a single composite view from the whole tunable range with one of the "spectrum analyzer" helpers.

Of course, you can manually accomplish something similar simply by tuning to various windows where you might expect to see spurious signal, e.g. 2x the transmitter frequency and further harmonics, or at various combinations of signal/IF/LO interactions. But scanning through an entire wider range while continuously transmitting might help you catch something you wouldn't have thought to check.

Note also that the signal on your handheld transmitter's antenna port will be waaaay to strong to inject directly into pretty much any receiver, including your SDR. So if you want a direct cable connection, make sure to use an appropriate amount of attenuation. Even picking up the signal via antenna at the same workbench could easily overload the SDR's receiver circuitry and cause a variety of additional signal distortions to show up, which aren't actually emitted by the transmitter itself. Make sure your receiver is not overloaded by your (or any other nearby transmitter!) before considering its output valid.

Answer 3

In addition to easy-to-use software packages suggested, I would also suggest that you consider gnu radio. Gnu radio is a drag and drop sdr framework. It does also allow programming in python and c++ with hundreds of signal processing libraries needed for your application.