[hpsdr] Strange question about Penelope

[Tom Clark, K3IO]

Chris Stratton wrote:
> In a transponder application, the necessity of the
> data going through the computer (and thus the USB)
> should probably be re-evaluated.   
>   
The more sophisticated SDX implementations look at the power spectrum of 
the signal, applying adaptive AGC (actually variable notch filters) to 
better equalize the power split among the different users.

Less sophisticated SDXs (like the Suitsat-2/Odyssey implementation) 
don't have enough CPU cycles for the FFT implementation; instead they 
implement multi-tap FIR or IIR filters to define the passband 
definition. In this case, a full-band AGC can be implemented by changing 
the multiplier coefficients in the FIRs.
> What is gained from that round trip vs. keeping it
> entirely in the FPGA's? (with USB maybe reserved for
> monitoring)
>   
For spacecraft implementations, the added time delay for signal 
processing is comparable to the round-trip path delay to the satellite, 
so we don't worry about the time delays.
> I know this group has done a lot with the PC-based-DSP
> paradigm, but it's not the only way of doing this. 
> Some combination of either a programmable FIR filter
> in the FPGA and varying the decimation ratios and
> tuned frequencies should be able to accomplish any
> desired contiguous block translation.
>
>   
Yep

73, Tom


 1211301964.0