Recent visit, SDR work, etc

[Terry Fox]

Hey guys, it was great to see those of you that made it to tacos while I was
there.

After tacos, I tried a simple demo of the Brainerd DDS/QSD board for some of
you.  It turned out that there was no RF leaking into the ARRL meeting room,
at least none at 40 meters, that I could find.  I did see some artifacts
from an AM or FM broadcast station.  I tried several QSD-style receivers
(Brainerd board, SoftRocks, etc).

That got me thinking that for demos, I need to include at least an active
antenna.  I'm going to resurrect the AMRAD active antenna that Frank did a
few years back, and put it in my bag of tricks.

The other way to do a demo would be with another, "standard" rig to bounce
signals with.  I had my FT817 in the trunk, and if we had more time, I could
have set it up as well.  This would be HIGHLY recommended for future demos.

At Winterfest, maybe a demo of a SoftRock RxTx running Rocky, and decoding
PSK-31, which Rocky now does natively.  Also, maybe a demo of a modified
RxTx doing the N2PK network analyzer.  You could use a regular rig as the
other end of the demo if there aren't many real signals on the bands.

What we need (AMRAD, and the general amateur community) to further our SDR
work is a couple more hams that have an in-depth knowledge of programming in
Linux.  I'm not talking about installing, configuring, or managing Linux
(although that would help as well).  I'm talking about hardware drivers, C
programming X user itnerfaces, DSP interfacing, and other down-and-dirty
programming.

A recent example of this is trying to configure Linux to use two sound cards
to be used for SDR work.  One (a better unit) would be used for the Rx IQ
signals in and transmit IQ signals out.  The other (lesser quality) would be
for demod audio out to speakers, and a mic input for transmit.  As it turns
out, this is NOT POSSIBLE right now under Linux, thanks to some long-running
brain damage in jack.  It is easy to do under Windows.

Another example is improving the GUI for Linux SDR programs.  The (arguably)
best SDR program witht eh (arguably) best UI right now is Edson's sdr-shell.
It has source code available.  But, the spectrum display is limited by the
width of the display, not the number of samples.  Edson does not do any
averaging or other method to combine "bins" to allow a wider bandwidth
spectrum display - unlike Rocky.  Rocky has (arguably) the best and easiest
user interface for simple SDR work.  The window can be resized, and the
display can show as much of a band as the sound card can support.  Modifying
sdr-shell to work like Rocky is another area of improvement that someone
could help with.

Mel (W4MEL) here in Charleston has modified Edson's sdr-shell to use the
newer version of the QT toolkit.  He also has added basic DDS support, so
the Brainerd board can be used for receive, along with any other DDS driven
by my PIC code.  Mel is also adding transmit capability to sdr-shell.  He
has it generating the necessary audio IF signals already, except for CW.
Edson is also back playing with software.  But, others could help out a lot
in this area as well.  The DTTSP core is not easy to understand at first,
but it can be figured out.

Windows is great, but the handwriting is on the wall.  Vista, and future
releases of Windows, will make it increasing difficult to interface to
hardware, and do significant IO handling and processing.  The big media
companies have won out, protecting content is more important than
functionality.  Therefore, a move to Linux is in the cards for us
experimenters.

So, this is a shout-out to other AMRAD amateurs to help with SDR, especially
in the Linux world.  How about it????

Terry
WB4JFI