Hi all,


I've been trying to learn DSP on my own, and I could use some pointer on the subject from DSP experts on this.


Attached link is a PDF document on a Software Defined Radio that I bought and have been tinkering with.


http://www.simplecircuits.com/SimpleSDR.html


On page 6 of the manual is the block diagram for this radio.


What I am having a hard time understand, is how the Hilbert Transformer, together with the adder and the 64 tap FIR low pass filter, (second and third row of the block diagram), somehow deciphers the AM data before feeding into the DAC.

This is what I know so far from doing my own homework. Looks like the frequency in question, between 3.5~18 MHz range, is amplitude modulated as DSB, which is a popular technique for ham radio. From what I understand, the Hilbert Transform shift the signal's phase by 90 degrees, which, by virtues of complex signal analysis, turn the DSB signal into SSB by either adding or subtracting the phased shifted signal. It will either add the LSB to the USB, or vice versa, and the resulting single side band will contain the total signal power of both LSB and USB. The low pass FIR filters out the frequency components above 2 KHz, and pass through the ‘voice data’ contained below the 2 kHz.

Still I don’t quite see how the amplitude modulated signal is picked up beyond that. Looking at the schematic at the end of that PDF file I don't see any circuits that stood out.


This SDR aside, in general, how does DSP combine/separate the data from the carrier frequency? I have this concept that an ideal SDR involves an antenna to ADC and the rest is all DSP, which means however it's modulated it's somehow demodulated using DSP.



So can someone explain me what is going on here? Or if it's too much to explain on this, would anyone happen to have some technical documentation on the subject that I could study up on? I have a few books on DSP but none of which really talks about how it deciphers the information in a signal, whether it's amplitude or frequency modulated signal.

Thanks.

 

Only a few hams use AM signals anymore. If you look at current practice you find USB on 20M and down, while LSB is used on 40, 80, and 160M. In analog signal processing when you combine two signals with what is essentially a multiplier you get four things out. If we call the original frequencies f1 and f2 then what comes out is those original frequencies and two new signals which are the sum (f1+f2) and the difference (f1-f2). If you choose the values of f1 and f2 then a signal in one range of frequencies can be moved to another range of frequencies. The same effect can be achieved in the digital domain.

To show this multiply two cosine functions together of frequencies f1 and f2 and take a look at the result.

Now for the digital domain you have to get very familiar and very comfortable with convolution. It is the key to understanding what is going on. If you haven't found it yet Steve Smith's excellent free e-book on DSP just might be the ticket for you.

http://www.dspguide.com/