Juan, > ubject: Re: What is the best way in a SDR? > > Well, guys, there are a lot of ideas, hi! > > There are projects that They make the QSD process by DSP, like the > DSP-10. > But in this case you make some operations like FIR filter, Hilbert > Filter, > Sine/Cosine Wavelet,... Maybe it's not a problem because there are DPS > too > fast, 40 or 60 MIPS. > > Also there are DSC, cheap and easy to work with them (Motorola's > 56F8165, > Microchip's dsPIC30/33F). And they have a fast ADC's (Microchip's > dsPIC30F2010 500KSPS). In ANALOG DEVICES there have some ADC faster > than > it, until 250MSPS. But I think it's easier to make a Tayloe Mixer and > after > input a ADC. > > So, Which bandwith do we have to use? I can to input 455KHz from a ADC > because Nyquist's rule is about the maximum bandwith that I can get > from a > signal. I can put the 455KHz to a ADC about 100KSPS but I get only > 50KHz > bandwith from it.
No comprendo when you say "I can to input". Nyquist (Shannon) talks about the portion of the spectrum range that is non-zero in power. The 455kHz filter output has a very low bandwidth. Maybe it's just the language barrier here, if not do a web search on "undersampling" and check out where the spectrum ends up aliased when you sample lower than the IF. The choice of ADC depends on the technology too. Are there sufficient bits? Is it an audio codec with sharp filtering at the input? Is it a sample and hold? Do any clock signals give high order harmonics that you pick up in the RF input? You also mentioned about an ADC, giving only half the bandwidth of the sample rate. You can use two with the ADC clocks 90 degrees apart and get a bandwidth equal to the sample rate. One thing you will notice is that if you sample at a rate higher than your audio signal bandwidth, you will waste a lot of CPU time. I had a project in a place where I worked at in 2001 we used a dedicated chip (the AD6620) to take an IF and bring it down to zero IF. If you want to stay cheap, not spend development time and keep plenty of CPU cycles, you can use two DSPs, with one dedicated to the downconversion to zero IF. If the ADC doesn't have enough bits you can model the quantization noise as white noise (it's a simplification) and just sample at a higher rate and filter the output digitally. > > If I get 30KHz, it's the 40M's CW part! (almost). If I use a DSC with a > Audio codec, I can make a QSD transceiver. > > But, Where do I put the QSD process, in a 455, 50 or 30KHz IF? IF to a > AD & > DSC or to a Tayloe mixer, AD & DSC? > > Well, guys, it's sooooooooo interesting, isn't it? > > Juan Machuca EA7FNJ > [EMAIL PROTECTED] - [EMAIL PROTECTED] > http://www.qsl.net/ea7fnj > _______________ Frank ------------------------ Yahoo! Groups Sponsor --------------------~--> Get fast access to your favorite Yahoo! Groups. Make Yahoo! your home page http://us.click.yahoo.com/dpRU5A/wUILAA/yQLSAA/ELTolB/TM --------------------------------------------------------------------~-> Yahoo! Groups Links <*> To visit your group on the web, go to: http://groups.yahoo.com/group/soft_radio/ <*> To unsubscribe from this group, send an email to: [EMAIL PROTECTED] <*> Your use of Yahoo! Groups is subject to: http://docs.yahoo.com/info/terms/
