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

[linrad] 1: Questions and assumptions [by OH1ZAA - Chapter_1]

1: Questions and assumptions [by OH1ZAA - Chapter_1]
(thread started June 14, 2004 with Chapter 'zero')

'Linrad' can be classified as a unique branch in the present
developments of Software Defined Radio (SDR). It has several
features that you will not easily find in any other SDR-branch.

In my view the most important characteristic of 'Linrad' is its
ability eliminate many forms of intermittent or persistent man-
made noise completely, especially the sort of noise coming from
all kinds of spiky sources like brush engines, power-line leaks,
welding arch's, faulty ignition and the like. This means that
for 95% of potential weak signal operators the suffering has
come to an end, and the impossible situation can be turned to
a fruitful working environment that is virtually noiseless.

Since the spiky sort of industrial or man-made noise does not
produce much of a power spectrum at microwave frequencies, my
guess is, that the biggest benefit of 'Linrad'-techniques can
be gained in the 25-300 MHz range when considering trade-offs
with noise cancellation and dynamic range. However, due to all
the other features, the technology is beneficial for general
use at any given part of the spectrum. At the present stage
of development everything seems to be optimized for weak CW
reception, but much more is coming our way.

Given an operational mid-range PC with the appropriate SVGA-
and audio cards, a top quality receiver can be obtained with
less than 50 dollars/euros of home brew additions. For noise
suppression that little setup will beat commercially available
receivers in over 95% of circumstances, beating even the most
expensive equipment costing 10k...50k units or more... 
After the tweaking of the noise suppression properties, dynamic
range can be improved with increasing cost per additional decibel
of improvement. However, if one is willing to operate the gain knob
of the receiver just a bit more often, probably 99% of signals can
be heard instantly with the 50 dollar circuits, compared with a 5000
dollar realization of 'Linrad'. The cost for extra dynamic range is
quickly mounting to over 500 dollars/euros per dB, and the need
for that is only in extreme situations with permanent changes
for strong nearby signals. It does not mean that the more
expensive setup would be any more sensitive, i.e. during
absence of extreme signals elsewhere in the bandwidth, the
whole spectrum would be equally observable, with a given
equal sampling rate of the sound card in both situations.

So far I have installed 'Linrad' software under SuSE 8.0 Linux
with kernel 2.4.18, Knoppix 3.4 (kernel 2.4.26/2.6.5) and Gentoo
Linux 2.4.25-r2 with OSS sound drivers in all cases, and svgalib
1.4.3 (1.9.17 under Gentoo). My biggest battles have been with
the 1.9.19 svgalib_helper.o module, like reported by many others,
and I have never succeeded to get it going. These days I am able
to install and configure the full set of linrad 01-20 and svgalib
1.4.3 software in an hour, but I spent 15 months to understand
that in SuSE 8.0 the mouse has to be set right for both the SuSE
main package and the libvga.config file. With almost 20 years of
DOS/Windows experience a lot of energy has to be spent on learning
the tricks and philosophy of Linux (surely a worthwhile activity).

Old SVGA cards like the i740-based Gigabyte (AGP), S3-based (PCI), 
ATI Radeon-7000 and Powercolor M64 cards work with svgalib 1.4.3,
but Matrox G400MAX does not work. Gentoo is very coherent, and its
"native" 1.9.17-r2 svgalib took the unknown SVGA-card automatically.
Many modern SVGA_3D cards (like a Ti4200) operate flawlessy (VESA).
Gentoo is probably the best Linux-version for people that want to
know all the nuts and bolts and backgrounds in the long run, and it
can be updated (synchronized) on a daily basis (by Internet sync's). 
Sound cards are under study, and there is a lot of confusion to sort
out, just to understand that e.g an AC'97 mixer indication does not
mean that it has anything to do with a VIA AC97 OSS driver. For the
newcomer it is initially a dark jungle out there, and it will take
a lot of time just to find out how to disconnect the internal path
of the analog input and output of a sound card with the mixer(s)!!

So far I have only once been able to feed actual audio signals through
a functional 'Linrad' setup, and now I have realized that I should have
tried that much earlier. In Chapter_2 I will try to indicate some advice
for the newcomer, in order to achieve faster results with this system.

I must say that the first contact with the 'Linrad'-processed signals
was quite a shock in the positive sense (a long awaited experience).    
A simple pulse generator will be soon available in order to be able
to calibrate a simple and inexpensive front-end for the sound card.

Keep us informed of YOUR experiments!
                                          73, "Zaba"  OH1ZAA

Initial Chapter_0 below .....
(future chapters will be stored in the archive)


Questions and assumptions [by OH1ZAA - Chapter_0]

'Linrad' is an extremely interesting development, and it is
great to see many of the former VHF/UHF/SHF pioneers working
with a new type of challenge/tool for weak signal reception.

There is much more behind the visible detail, and for the
newcomer it is very hard to know where to begin. My guess
is that Leif, SM5BSZ, is investing at least 14 hours a day
into this development, and he has been doing that for many
years now, possibly adding up to 30,000 hours. Also he is
admirably commenting on the majority of the email passing
through the 'Linrad'-list.

This is my first time to post through the list. However, I
have been communicating with Leif directly over the last two
years, and I have done a couple of attempts to get 'Linrad' in
action on several PC's. Initially it has mainly been a lesson
in Linux, and between failures there have been long breaks...

During last weekend's RATS VHF/UHF/SHF-meeting in the OHo_Aland
Islands it was good to see Leif's solid 'Linrad'-setup in full
operation, and we did some thorough measurements on a FT-857D.
So besides being the ultimate receiver for weak signals in very
difficult environments, it is also an instrument to characterize
the performance of commercial or homemade communications gear.

When reading through all the 'Linrad' documentation and related
QEX-publications, one gets easily overwhelmed by the huge amount
of information. However, I have noticed that I get to know more
and more, and maybe soon I will know "too much" to be able to ask
the numerous questions that I have had before. Therefore I thought
that it would be good to make maybe a dozen short email chapters
with issues related to the interpretations of 'Linrad' features,
and maybe a guided approach for the newcomer, who will be asking
possibly the same sort of questions. This is chapter 'zero'; more
of these will follow at random, but in correct numbering sequence.

In the meantime I wish you all good luck with Linrad experiments.

                                       73, "Zaba" OH1ZAA/2