[Date Prev][Date Next][Thread Prev][Thread Next] - [Date Index][Thread Index][Author Index]

RE: Computer controlled radio seen at AMSAT Annual Meeting 03


Thanks for your reply.  Your guess about the 703/706/897 is not correct.
The sound card is the actual last IF of the SDR radio.  The last IF of the
SDR-1000 (and again, this is software configurable and has been chosen
for the current code) is at 11025 Hz.  The sound card is the primary
determinant in the system performance given the radio.  The 703/706/897
receivers do not compare favorably to receivers such as

Yaesu FT-1000D, Yaesu FT-1000MP, Elecraft K2, Ten Tec Orion and now the
SDR-1000 WITH A GOOD SOUND CARD.  You are right on the money about too
many cables, too many sources of errors but this reminds almost exactly
of patching together of pieces of things in homebrewing.  The SDR revolution
has arrived on amateur shores.  It is hear to stay and will make great
changes in our hobby in terms of our ability to experiment, and make our
radios unique to our wants and needs.  The industry standard digital
voice is a pretty easy add on to the software defined radio and I agree
that digital speech compression is here to stay and hope we get involved
in it.

In the next few weeks (shortly after the New Year), you will be able to
hook Digipan, DRM, etc. to the SDR-1000 and this will usher in its first
step into the more traditional digital modes.

Sorry for rambling on but I have not been this into anything amateur radio
for a decade!


-----Original Message-----
From: Emily Clarke [mailto:w0eec@amsat.org]
Sent: Monday, December 15, 2003 5:54 PM
To: Robert McGwier; Amsat Bulletin Board
Subject: RE: [amsat-bb] Computer controlled radio seen at AMSAT Annual
Meeting 03


Although I think you also make a good point, I wouldn't expect the
703/706/897 by themselves to produce the performance you are measuring in
your lab, and I hope I didn't imply that.  But you are measuring results
using features afforded by the sound card.  I would say if you add the same
sound card, computer and the right software to the 703/706/897 I believe
they will as well.  I've been experimenting with my Icom 910 and PC based
software including a software DSP and the results are good!  Maybe not as
good as your lab tests, but they are nonetheless fun for me.

I also believe that 703/706/897 styles of radios (and price points) will
evolve over time and embrace new features and advanced modulation
schemes.   I don't think the big three (I/K/Y) will sit idly by if there is
a market.  I believe the next to be incorporated into radios will be an
industry standard digital voice protocol (TNCs and SSTV are already there),
and after that will be things like spread spectrum and high error rate
corrected digital modes.  I don't mean this as a criticism of I/K/Y but
they are commercial companies that commit funds to put technology into
production after there is a market demand, generally not before.  It's not
there for SDRs at this time however I believe all manufacturers are looking
very carefully this way and will make decisions on near-term designs based
on the commercial success or failure of existing products (like the
TS-B2000 and FT-1000MP) that have similar or lead-in features.

I am happy to hear that Gerald is improving designs and that the sound card
will be part of the design.  I'm very leery of products making statements
like "good receiver performance [...] is directly related to the quality of
the sound card." when one can control neither the sound card nor the
software.  One weakness of the current design is that built-in PC sound
cards are the norm, not the exception, and if you buy an advanced sound
card system today what is the likelihood that software will support it in
the future?  So an integrated sound subsystem will at least provide a
measurable baseline, configuration management and survivability.  And
that's a good thing.

I'm pleased to hear people are writing software in other platforms and
languages, but I wonder.  Will any embrace ISO-9001 quality standards and
produce very well tested and documented high reliability software and
firmware?  Maybe, but at what additional cost?

73 and Season's Greetings,


At 03:07 PM 12/15/2003 +0000, Robert McGwier wrote:
>Some of your critique is right on the mark and Gerald is likely
>to tell you that the people he considers his friends and advisors
>are telling him exact what you have been saying.  I can tell you
>however, that not one of the receivers in the list you have mentioned
>can come close to the performance of the SDR-1000.  In my labs
>at work, and in the lab of W3IP (who is VP of SAIC) we have measured
>blocking dynamic ranges and spur free dynamic ranges that EQUAL to
>the FT-1000D.  You are very right that we had to use top of
>the line audio cards to get these numbers.  It might surprise you
>however to note that top of the line is under $100 for the purposes
>you mention.  (M-Audio Revolution 7.1 is $70-ish at most stores).
>If you want an extra 3 dB in most of the linearity numbers and
>spur free numbers,  you need to spend 10 dB more money for a Lynx
>L22 but I do not recommend that for a number of reasons including
>that I believe it will be of no use in a year when a new version
>of this kind of radio comes out (at least I hope it does!).
>So as it now stands, I can get just about the best receiver money
>can buy for $600.  It does have drawbacks.  These have been realized
>by Gerald and are being addressed in a clever way.  The SDR-1000
>will be a 50-100w transceiver with the receive problems addressed
>by a simple add on board.  Gerald did the design for all of this
>"in his basement" and he made a boo-boo or two.  I might also
>say that Gerald has expressed some amazement at the performance
>levels being measured now.  The primary uh-oh is he put the
>noise and receiver merit determining amplifier BEHIND the mixer
>and of course this amplifies the DDS spurs, mixer noise, phase
>noise from the oscillator/limiter/mixers as well as the signal
>of interest.  This is being addressed in the add on board which
>will move large pieces of that gain in front of the quadrature
>sampling detector.  Most of the problems with this receiver's
>potential disappear once this has been done.  You still have to
>contend with the sound card and the interfacing.
>Gerald and his friends and owners have all decided that the "next
>radio" whenever that comes out, can not depend on the sound card.
>As such, it is to have a high speed digital interface such as USB
>2.0 for both audio and control.  I am sure Gerald would tell you
>that he is not interested in having an opinion of the radio given
>by someone who owns an original SB-16 card!
>AB2KT and I are testing the Linux version of the radio software
>now.  It will be available on my home website soon
>( http://rwmcgwier.homelinux.net )
>under an open source license.  There will be a php based bulletin
>board to discuss this work and other work on that web site. The
>site is very rudimentary now as I have just gotten the pieces
>assembled and have been concentrating on the linux code with Frank.
>Two people are writing Visual.NET versions of the code.  Both have
>pieces in Visual C++ and C#.  The primary reasons for going to
>.NET development is that it is supported under XP and it has a
>real threading model that allows you to write real code (even if
>you do have to hide it behind the MicroSH%^ .NET stuff).
>I almost puked when I got the code with the radio and saw that it was
>VB 6.0.  Gerald can tell you that I was less than gracious about it.
>On the other hand, I am one of the better contributors to its
>upgrades(in terms of volume and added features, not quality of code).
>Now we have folks from everywhere adding features.  VB 6.0 was,
>in my opinion, having shut my mouth and thought about it, likely
>the very best vehicle for Gerald to have introduced this SDR
>to the world.  Many have contributed and VB 6 has made this easy
>to do.  Gerald very best business decision was to decide to
>do open source for the code and to make it doable in an easily
>accessible language with decent signal processing library support
>from Intel.
>It has absolutely NO PLACE in the future.  VB 6 will be replaced,
>and soon, on multiple platforms.  I am building a Qt interface
>that can be used on Windoze and Linux with code works for
>both platforms (with enough ifdef's to allow you to compile on
>Stay tuned and do not despair.  SDR is a revolution and it has
>just begun.

W0EEC - CM87tm
AMSAT Area Coordinator - San Francisco Bay Area
http://www.projectoscar.net    http://www.emilyshouse.com/W0EEC

Help Launch Echo - http://www.amsat.org/amsat/sats/echo/index.html

Sent via amsat-bb@amsat.org. Opinions expressed are those of the author.
Not an AMSAT member? Join now to support the amateur satellite program!
To unsubscribe, send "unsubscribe amsat-bb" to Majordomo@amsat.org