*Hello!*
**
*Comments and suggestions for further experiments, please? John is looking
for some other*
*stations to help with the experiements. Post your comments here and / or to
John KD6OZH directly.*
**
*Thanks,*
*John K8OCL*



---------- Forwarded message ----------
From: John B. Stephensen <kd6ozh@...>
Date: Thu, Jun 25, 2009 at 7:32 AM
Subject: OFDM Activity
To: k8ocl@...


Hi John,

I haven't done anything on 6 meters since the STA expired. I've been making
a less expensive version of the DCP-1 hardware that is more flexible. The
external microcontroller has been replaced by a soft CPU inside the FPGA.
The winter TAPR newsletter has an article on this CPU and the source code is
on the TAPR web site. I've ported all of the Verilog code developed during
HSMM working group era to the new FPGA and a later version of the Xilinx
develoment tools. The modules that were tied together in hardware are now
peripherals for the soft CPU and several more have been added to support
traditional analog modes. I've attached a block diagram and some pictures of
the new PCBs. I plan to pulish more on the PCB and the rest of the Verilog
and assembly language code as testing progresses.

This new board can generate and process 1-30 MHz RF directly. I'm now
testing soft CPU code for the various modes and have SSB and ISB
transmission and reception working plus FM and OFDM transmission. I should
have FM and OFDM reception working soon. The soft CPU scheme allows the
DCP-3 hardware to be easily switched between analog and digital modes so it
can be used to build a general-purpose radio or fed into a transverter. The
final configuration is intended to be a device that will switch between
analog voice and TCP/IP data via the ethernet port.

I plan to do more on-the-air testing this summer. The new Verilog tuner and
modem modules can handle signal bandwidths between 1.6 MHz and 500 Hz, so
I'm no longer restricted to VHF and UHF. I may do initial testing on 80
meters with 6 kHz wide OFDM that could support 9.6 kbps under good
conditions. Anything that will work on HF will work even better on VHF and
UHF frequencies. Switching to 750 kHz wide 1.2 Mbps signals is now a matter
of loading new filter coefficients from flash memory.

After testing on 80 m and 70 cm, I could apply for renewal of the STA for
wideband 6-meter tests. However, it's probably more useful to find few other
stations that could test on 80 m or 70 cm first. I had to use up 120 square
inches of 4-layer PCB to get a prototype, so I made four DCP-3 PCBs and two
other variations. One simple route to more widespread testing may be to
produce a few adapters for transverters with 28 MHz IFs. I'll see what
happens in the fall after I've done more on-air testing myself.

Since the DCP-3 is easily programmable there is another alternative for the
6-meter bands that is currenlty legal and could be developed. OFDM is
limited to AM voice bandwidth on the VHF bands. However, 8PSK could fit in
the same 20 kHz bandwidth as 9600 bps FSK and could operate at 3 times the
data rate. I did put AX.25 accelerator logic in the FPGA, but it will be a
while before I have time to test it on anything but the logic simulator.
It's intended be used to implement 9.6-76.8 kbps FSK modems for amateur
satellites which is what originally got me interested in doing a DSP
project.

73,

John
KD6OZH



--
Nisciune nasce imparate

"When you can measure what you are speaking about, and express it in
numbers, you know something about it." [Lord Kelvin 1824-1907]


[Non-text portions of this message have been removed]