Programming help with Charleston SDR project
tfox at knology.net
Mon Jul 20 17:37:00 CDT 2009
Some of you may be aware of the Charleston SDR board, developed by Dr. John
Schwacke. It uses the Burr-Brown/TI AFEDRI8201, a 12-bit A/D chip, designed
as an IF sampler for HD radios and other applications. John has designed a
small board with the 8201, a programmable gain amplifier, and an oscillator,
that will plug into a Digilent Nexys2 FPGA development board. The Nexys2 is
used primarily as an interface between the 8201 chip and the host computer
See John's website: http://sdrtrack.drupalcafe.com
I am also putting information and files on the AMRAD web site
(www.amrad.org), under SDR & DSP, then click on the Charleston SDR link.
I've placed a parts list for the Charleston SDR board there, along with the
bit files for the Nexys2 board. I can upload more files as necessary.
Some folks in AMRAD are interested in using this (relatively) inexpensive
design to learn more about direct-sampling Software Defined Radios (SDRs).
The total of the two boards is around $210. It should be noted that this is
NOT a high-end device such as the HPSDR, QS-1R, usrp, or Perseus. With only
a 12-bit A/D, and about 80MHz sampling clock, this SDR is not going to match
the performance of those otehr devices. But, the cost is significantly
lower, and it provides a very flexible SDR training tool. Frank is making a
prototype run of the SDR board, which should make them more readily
I have used one of these boards to listen to HF, plus FM broadcast and 2M FM
repeaters, these last two using subsampling. The major limitation right now
is software. With Windows, there is only a spectrum display program, no
demodulation. Under Linux, there is a nexsdr "source" available, so the GNU
Radio receivers can be used.
I think that this project could use some help from one or two software
gurus. I have three specific pieces in mind.
The first is for Windows, we could use someone to create an interface that
changes the rf samples (16-bit I & Q) from the Digilent DLL into a
"PortAudio" format, that could then be passed to PowerSDR or other similar
The second is also for Windows. This would be a more daunting project.
This would take the same samples, but magically convert them into something
that looks like a sound card. This would be used with SDR programs like
Rocky, that interface with sound cards, possibly through WDM, ASIO4ALL, or
some other bit of obtuse Gatesonian magic.
The third, probably the easiest, is for Linux. We already have software
that creates a GNU Radio "source", so using GNU Radio is already being done.
It would be great to convert the I/Q samples to something that alsa, oss,
jack, etc can use. This way, Linux programs such as sdr-shell could use the
I don't think the first or third project would be very hard at all, probably
just researching the various formats in question, then writing a "wrapper"
for the samples. The Windows sound card seems like it would be more
If there is anybody here on the tacos reflector that can help in any way,
please let me know. Mel, Frank, and I are already either using this SDR, or
will be in the very near future. I think this project is a stepping stone
to the future of SDRs, and ham rigs in general.
More information about the Tacos