I'm on a team of college students that is doing a senior design project involving the SDR 1000. We are trying to create a portable SDR with similar hardware and software to the SDR 1000. We have a couple of questions about exactly how the SDR 1000 works that we couldn't answer from the QEX papers. If anyone has any thoughts on any of our questions, please let us know.

1.   When we are receiving an FM signal in FM mode, we can physically remove the I and Q inputs that run from the SDR 1000 to our soundcard (a Delta M-Audio 44). How is it possible for the PowerSDR software to demodulate an FM signal if either I or Q is all zeros?

2.   How does the SDR 1000 do phase matching? We were thinking that the local oscillator inside the SDR 1000 would need to be phase matched to the received signal, but we can't find any hardware that performs that operation.

3.   How does the input RF signal relate mathematically to the I and Q signals? We have found that the definition of "in-phase" and "quadrature" used in our textbooks isn't the same as what is used in this system, and we're having trouble with the mathematic analysis on account of that.

4.   How does the Tayloe detector convert to an IF frequency? When we looked at the patent for the device, it only said that it converted to baseband, not to an intermediate frequency (namely 11.025 KHz).

Thanks for your help!

Leah --KB3NMO

1. The FM demodulator is passive in the sense that it doesn't care whether it actually has a signal or not.  It will continue to attempt to demodulate whatever you send to it.  I suspect the PLL will thrash around with no signal.  Note that even when there is no signal on the cables, it is unlikely that the samples are 0s due to the inherent noise in the soundcard.

2. In software.  There is a PLL in both the FM demodulator and the SAM (synchronous AM detector).  I am not an expert on how this software PLL works, but this is how we "match the phase"

3. I'm afraid I'm not much help here.  This would be a better question for someone on the hardware side (N4HY or K5SDR). 

4. The detector does indeed mix to baseband.  However, in our software, we have arbitrarily chosen to tune the hardware DDS 11kHz low and then use a software oscillator to mix back down 11kHz.  This allows us to avoid some of the nasties around DC that would otherwise get stuck right in the passband if we went straight to baseband.

Best of luck on your project and please let me know if I can answer any further questions on the software side.

Eric Wachsmann

FlexRadio Systems