[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xylo-SDR] Firewire, DSP and FPGA on one module



The main thrust right now not necessarily in priority order is to;

1. Eliminate the sound card, with one of superior quality, with elimination of ground loop noise. 2. Correction of the 200MHz and A/D clocks clock to super accuracy and stability. 3. Replacing of the parallel port interface in the SDR-1000 which controls band switching, Autotuner control.
4. Control and audio for improving CW operations.

The ultimate goal being of having one USB cable between the PC and the SDR 1000 radio, all data and control being done through USB.

Once that is accomplished some in the group are interested in going further, to at least reduce the load on the CPU by doing some of the PC's task in the FPGA so slower laptops and PC's will function, if not outright replace the PC.

Outright replacement of the PC is a very tall task, you will be replacing all the current software with hardware and with our budgets all we can afford are limited free tools. I would most likely involve adding a floating point DSP chip to the mix.

One of the reasons I have pushed for usage of a Cyclone II chip is the added hardware multipliers that can be used in making a data crunching unit, most of the FFT and filter code consist of additions, subtractions, shifting and multiplication.

All these task will be modules of software in the FPGA and can be added or removed at will for use in other projects.

At 10:20 AM 12/18/2005, you wrote:
Robert,

I may have read a lot into the project that I wanted to see but I
thought one of the reasons for the efforts were to replace the desktop /
laptop. This thought may apply to the disscussion of the USB 1 or 2.0,
it is a link to the desktop / laptop and we know that can be done, but
in order to have a SDR free from the big computer we need hardware that
is fieldable and relatively low power.

Cash

-----Original Message-----
From: xylo-sdr-bounces@lists.ae5k.us
[mailto:xylo-sdr-bounces@lists.ae5k.us] On Behalf Of Robert McGwier
Sent: Sunday, December 18, 2005 4:31 AM
To: Xylo-SDR Discussion
Subject: Re: [Xylo-SDR] Firewire, DSP and FPGA on one module


Hello Leon!

The 16 bit fixed point DSP chip could be an issue which is dependent on
how it is used.  Dealing with fixed point is an art few can master.  I
did it with the AEA DSP boxes 15 years ago because there was no other
choice and I am doing it now in FPGA and some smaller spacecraft
embedded microsystems.  I am uncertain about its suitability for general

experimentation with the goal of a high performance SDR.  If it were a
32 bit fixed point system,  we would be in business.  The FPGA seems
nice enough.  The USRP uses Cyclone with 20,000-ish logical elements.
This has 50K.  So that is a win plus it has the 18x18 multipliers.  All
positives.  It is my personal opinion that we are better served at this
time by mostly avoiding the use of DSP chips, especially fixed point
ones and to rely on our ability to speedily process data on our modern
desktops.  The SDR-1000 DSP core,  written by AB2KT and I,  recently was

hung on the end of an Agilent multi receiver system.  It successfully
prosecuted 92 separate receivers, each sending I/Q at 8000 samples per
second,  simultaneously on a 2.4 GHz Xeon based Dell Server.  The
processor was not sweating at all.   It will easily do this up to 48000
samples per second since we were not consuming 1/6 of the CPU resources.

I went to Reno to work with the Gnu Radio folks for a week.  Using the
GnuRadio software elements, plus some blocks I helped write,  we
implemented a 1 Mbps GMSK packet SYSTEM  that transfers data over USB
2.0 (the preferred medium unless you can demonstrate WHY you must have
firewire) and we did it from scratch in 3 days .  You may download that
software now from GnuRadio.   This ran successfully on the N2MJI (USRP
hardware designer) abacus,  a 1.8 GHz celeron laptop.  We do need to
stop underestimating these desktops and to learn how to optimize for
them because they are so much easier to deal with than DSP chips.  DSP
chips are your last resort and intended primarily for embedded systems
that must have them.  I would limit there use to those places where I
need a completely standalone device and could afford to hang an external

system on the SDR.

The nice thing about using the FX2 and USB 2.0, is that we can and
should adapt our toy to use the HUGE body of software that GnuRadio has
produced in addition to those things we can and should add.

73's
Bob
N4HY


Leon Heller wrote:

>Here is an interesting product with a TI DSP, Spartan 3 FPGA and
>Firewire:
>
>http://www.traquair.com/products/ultracompact/uc1394a-3.html
>
>Just add an ADC for a high-performance SDR!
>
>Leon
>--
>Leon Heller, G1HSM
>leon.heller@bulldoghome.com http://www.geocities.com/leon_heller
>---
>[This E-mail has been scanned for viruses but it is your responsibility

>to maintain up to date anti virus software on the device that you are
>currently using to read this email. ]
>
>_______________________________________________
>Xylo-SDR mailing list
>To post msg: Xylo-SDR@ae5k.us
>Subscription help: http://lists.ae5k.us/listinfo.cgi/xylo-sdr-ae5k.us
>Xylo-SDR web page: http://xylo-sdr.ae5k.us
>Forum pages: http://www.hamsdr.com/hamsdrforum/
>
>
>


--
Laziness is the number one inspiration for ingenuity.  Guilty as
charged!

_______________________________________________
Xylo-SDR mailing list
To post msg: Xylo-SDR@ae5k.us
Subscription help: http://lists.ae5k.us/listinfo.cgi/xylo-sdr-ae5k.us
Xylo-SDR web page: http://xylo-sdr.ae5k.us
Forum pages: http://www.hamsdr.com/hamsdrforum/


_______________________________________________
Xylo-SDR mailing list
To post msg: Xylo-SDR@ae5k.us
Subscription help: http://lists.ae5k.us/listinfo.cgi/xylo-sdr-ae5k.us
Xylo-SDR web page: http://xylo-sdr.ae5k.us
Forum pages: http://www.hamsdr.com/hamsdrforum/


Cecil Bayona
KD5NWA
www.qrpradio.com

"I fail to see why doing the same thing over and over and getting the same results every time is insanity: I've almost proved it isn't; only a few more tests now and I'm sure results will differ this time ... "