[hpsdr] HPSDR future (dreams, ideas)

[Ben Witvliet]

I myself am not yet so far on the digital radio path that I can do more than
dream and visualize. I will already be happy when I have my complete HPSDR
transceiver so that my trustworthy FT990 will become decoration gathering
dust. Give me a year or two (four.. six?). 

 

Yet I'd would encourage work in the HPSDR group on:

 

1)      Two (or more)  hardware receivers on a single HPSDR

I would like to install  a second and/or third Mercury receiver on the same
Atlas. Don't know if there is any software supporting such an HPSDR
configuration already??  So really two simultaneous antenna inputs on the
HPSDR, so that we can do spatial filtering. 

PAoSIM has already built a very effective application for two spatial
filtering, much better than the thing FlexRadio demonstrates. He first
splits up a 3 kHz passband in small FFT (IQ) bins, than PER BIN compares
phase between both antennas and shows the results statistically. He than
introduces a filter window on the phase difference. His filter than puts all
the FFT samples that don't pass the selection to zero, and reconstructs the
3 kHz channel from the remaining IQ bins. PAoSIM has demonstrated to PA3DES
and me that in that way he can very effectively reconstruct weak signals
from one particular direction from a load of strong interfering signals from
several other directions. Not just some gain, really deleting all the
interference!

PAoSIM used two identical loop antennas, through two synchronized
FT1000MP's, to high-end audio cards. But you can imagine you can use this
system for brick wall elevation angle selectivity on a stack of two 14 MHz
Yagi's pointing at Europe, effectively killing all east Coast interference
and only hearing clean DX signals. Any takers?

 

2)      Very high dynamic range receiver

Although our 16-bit A/D is probably the best on the market, and the Mercury
is holding up fine, I would still like to have some 10 or 20 dB more dynamic
range (IFDR). There are no RF high-realtime-bandwidth A/D-convertors better
than the one we use yet. But would it be possible to use two in parallel,
one with the preamplifier and one with preamp off (or attenuation on), and
to process the simultaneous output of the two to get one high dynamic range
A/D? 

That way the highest bit values would come from the A/D without preamp and
have no distortion from the high level of the combined shortwave, while at
the same time the lowest bit values would come from the other A/D with
preamp, giving us the fine granularity needed for detecting signals deep in
the noise.

 

Kindest regards,

 

Ben, PA5BW 

pa5bw at xs4all.nl

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/attachments/20100723/dd62395c/attachment-0003.htm>