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

[linrad] UKSMG (50 MHz) awareness for noise-canceling; Re: [linrad] Re: Winrad V0.90


Hello [linrad]-group!

Thanks to Leif and the alert users of Linrad the technology
is leaping forward. I am sure that there are still many not
so obvious improvements waiting to be discovered. I have put
some comments on the UKSMG-pages as 50 MHz is certainly one
of the bands to profit from this. Personally I have had much
trouble from outboard motors, pleasure yachts that will buzz
in my 50/144 main lobes for miles and miles during summertime
for hours on end... I noticed I have repeated myself in those
messages, but maybe it is good to stress the need for input
bandwidth regarding efficient operation of the main feature.

                                73, "Zaba" OH1ZAA/2


P.S. Richard/G8JVM - I located the Spanish BT878A link:

http://www.domenech.org/bt878a-adc/index-decimator-e.htm

It goes up to 896 kilosamples/s (effectively 10-bits or more)



(recent) UKSMG announcements by OH1ZAA:
Linrad/Winrad: SM5BSZ/Leif has now markedly improved the Linrad Noise Blanker for the additional refined removal of weak man-made interference with the release of Linrad 2.02 (March 1st) [http://www.sm5bsz.com/linuxdsp/linroot.htm >> wlr2-02.zip] --- This version works flawlessly on Linux systems, but exhibits still occasional (rare) memory leakage problems in Windows. I noticed also incidental program termination by overrule of other Win-audio applications ... My general impression is that Linrad noise-blanking is globally still the best available --- I2PHD/Alberto's Winrad page (see 6 messages lower down) has counted over 1200 unique visits... You can run Linrad and Winrad subsequently on the same PC for a comparison ... input volume settings are a bit different usually --- A common misunderstanding with these noise-canceling systems is, that one could achieve top performance by heavy pre-filtering using narrow crystal filters ... NO NO NO ... it is exactly the opposite: in order to achieve COMPLETE elimination of man-made noise, one should provide MAXIMUM audio bandwidth in direct relation to the sampling frequency; i.e. with 48 kHz sampling about 22 kHz real audio output (or -22/+22 kHz I/Q-audio), and double values for 96 kHz sampling --- Do not expect miracles by filtering a 4 kHz wide spectrum, though a marked difference is noticeable even there --- These new technologies will change 50 MHz operating thoroughly for most of us, and it keeps operator involvement very vivid, as signals are still interpreted by ear (but without unnecessary suffering due to excessive noise) ---- Good luck (you'll be up and running in minutes, after a quick Linrad/Winrad install)! --- 73, "Zaba" OH1ZAA/OHoMZA --- Friday, March 03 2006 at 00:10 (GMT)
The WINRAD page [http://www.weaksignals.com/] has seen already more than one thousand visits --- According to I2PHD/Alberto there will be a build 21 soon in order to correct a few small issues. Please note the two small demo MP3-files of the same transmission at the bottom of the Winrad-page. You will be shocked to see the difference! It is a pre-taste of what you will experience in a man-made noise environment.... Please note that in contrast to traditional receivers - where selectivity is achieved with ever narrower crystal filters - in Winrad/Linrad the noise reduction is the better the more audio bandwidth is presented to the software. While 4 kHz wide audio can do something (bypassing a 2.4 kHz SSB filter) it is preferable to feed e.g. 20 kHz wide audio with 48 kHz sampling to either channel, or more than 40 kHz (+/-20) in case of a I/Q-type RF-frontend in order to get superior noise-blanking... For ultimate performance the number 1 option is still SM5BSZ/Leif's Linrad (now also with an early multi-threaded Windows-version: http://www.sm5bsz.com/linuxdsp/linroot.htm >> wlr2-01.zip) with two parallel channels... Winrad (installed in less than a minute) will of course evolve in the same direction with future added features... Make your next Es-season "different" with either of these programs; a separate RX-frontend can be constructed for I/Q-output with a couple of fixed crystal oscillators to cover the 50.000 - 50.250 band segment. With a 96 kHz audio card three oscillators will suffice for a start. It makes a nice beacon band monitor as well, with a possibility for a time-rollback to study the exact beginning of the opening (recording feature of the full monitored bandwidth in Linrad)... or you can go back for the last hour to see what you have missed within the DX-segment (live through it again by tuning around in the past) --- Another reception method is to mix down IF output from an existing transceiver for similar audio bandwidth --- 73, "Zaba" OH1ZAA/OHoMZA --- Wednesday, March 01 2006 at 03:07 (GMT)
WINRAD: As a follow-up to Linrad (below) we just got the announcement by I2PHD/Alberto of a new Windows development in digital radio. This is pretty amazing stuff! Installed in 20 seconds, and immediate operation after selecting the sound card and pushing the START-button. I could even switch from 22k to 48k samples in-flight. The graphics is great. It turned me speechless for the moment.... I guess that the internals are differing from SM5BSZ/Leif's code, but I am sure that we will hear much more about this. It is always good to have various systems side-by-side for evaluation! I guess that we are most hungry for all noise-cancelling properties of SDR-type systems, and the related bandwidth. Also the blocking dynamic range of modern 24-bit sound cards (given a high-quality RF-frontend) can't be beat even by expensive professional receivers. WINRAD V0.90 can be downloaded at http://www.weaksignals.com/ --- 73, "Zaba" OH1ZAA/OHoMZA --- Saturday, February 25 2006 at 22:00 (GMT) 


At 08:07 2.3.2006, G8JVM wrote:


Giancarlo Moda wrote:



Hi,
FST switches have quite an interesting bandwidth but
switching does not permit to work at 144 or even worst
at 432MHz.
FST3253 maybe OK up to 30 MHz then they may have
internal unbalanced switching. You should see the
conversion loss will increase and also the noise.
FST3125 (3.3V) have been tested OK up to 50 MHz in the
I7SWX 2T H-Mode Mixer with up and down conversion.
73
Gian
I7SWX





Morning all,



Looks like I've managed to kill the pcm side on two sound cards using winrad.
Still gotta find out how.
Now I notice one of the clever Italians has managed to use the video input of an BT878 TV card for sampling, I did find his web site although the mods to the card were obvious , but alas not listed. 
Anyone know more about this ?
I agree with you comments above Gian, you can I&Q demod easily up to 2.4 GHz , but the penalty is noise, the higher you go in frequency the wider the bandwidth used, WBCDMA is running at 4.7 Mchips/s. Way beyond bouncing a signal off the moon , I would n't even take guess at path loss.
The way I'm going is to use the finally IF for all VHF and microwave apps, this way I can gain on demod and keep the brick wall filtering and strong signal handling at VHF/UHF, That rig was built for contest work 25 years ago when there was activity in a contest. 
My thoughts


Richard G8JVM



#############################################################
This message is sent to you because you are subscribed to
 the mailing list <linrad@xxxxxxxxxxxxxxxxxxxxx>.
To unsubscribe, E-mail to: <linrad-off@xxxxxxxxxxxxxxxxxxxxx>
To switch to the DIGEST mode, E-mail to <linrad-digest@xxxxxxxxxxxxxxxxxxxxx>
To switch to the INDEX mode, E-mail to <linrad-index@xxxxxxxxxxxxxxxxxxxxx>
Send administrative queries to  <linrad-request@xxxxxxxxxxxxxxxxxxxxx>
LINRADDARNIL