Saturday, February 16, 2008

SINPFEMO A better signal tech analysis by using F/TDM

SINPFEMO
-------------
A better signal tech analysis by using F/TDM analysis drawings in specialized reports
(version blogger )
Considering that SINPFEMO does not 100 % cover the truth from listeners point of view in regards to signal readability QRM and fading I have used for more than 4 years derived from more specialized monitoring equipment such as Hewlett Packard , Rohde & Shwartz etc
I think however that both terms shown below are very easy to understand however there is still a little analysis behind them on how i used them so far.

#1. Frequency domain analysis
...................#1b. About computer DSP analysis on FDM
#2. Time domain analysis ( fading analysis)
...................#2a.Cases of 'Short' time fading found on SW:
...................#2b.Cases of long term fadings
#3. More advanced way for domain analysis
#4 Using R75 and specialized programs New.gif (908 bytes)
....................#4a Smart Icom 32 and R75
....................#4b:Use of DSP and R75




1. Frequency domain analysis
I used two types of analysis -with the help of a drawing. The first way is with the use of bars and the second with the use of bell shaped curves that still can show on how a signal could be QRMed from a nearby station . As that times my older receiver HF150 or previously SONY IDCF7600 D did not include S meters I was unfortunate to deliver drawings with accurate drawings, Instead of it my drawings were made with the totally subjective audibility standards as used by the SINP(FEM)O standards
Nowadays the usage of ICOM R75 with its AFC derived S meter fixed a bit of this problem. There are still not 100% accurate readings as (modulated) signals exceeding S9 levels are still not correctly read !For a level with the R75 you can still do a little more by using ATT (-25 db ) and then preamp ( in any scale from 10 or 16 dB ) in order to record the fading levels
This case be done more easily with more specialized interfaces such as SDU from Grove ( direct to IF stage if receiver permits ) or by using laboratory surveillance equipment ( the o.m. models of HP , R&S etc)
However there is still a question on which is the best way to measure the signals for receivers as R75 :
1.The one is reading on the S indicator during AM / AM Narrow
2. The second is checking the signal with CW analysis and using as narrow as possible bandwidth in order to narrow down the width of carrier (??) and clear out from the modulated signal.
Below is the idea with pictures. Please save this file and these pictures as under this analysis they may not be clear
Please open picture in a separate web pagePicture with bars:
The above picture is from very old logs for Vo Malaysia on 15295 during the times 93-94 with VoLiberty shown at left part and RFI ( France ) at right. Of course it can not show any QRM problems encountered
Dotted lines show the depth of fading.
This is my main way for signal reporting
Picture with bell curves
This can show clearly the QRM problem from both stations though still cannot deliver any problems encountered by modulation though this curve can oversimplify the term modulation when used on that manner (I found many signals being quite strong but low modulation)
Picture with modulation analysis
This is possibly the most accurate way to describe the characteristics of a signal that can also show the modulation index as well as tha audio quality. But how easy id for anyone to do it? I think that ONLY computer made techniques can do it or real time analysis by the specialized equipment can do it (?)
+With the spectrum analyzers you can too see any carriers and other signals that can lie inside the passband++
Of curse still co channel QRM cannot be seen this way except the aural way ....or with the use of LPDA antennas
+++added 5-5++
1b. About computer DSP analysis on FDM
Fortunately with the use of modern computer analysis with soundcards and the shareware audio programs such as Cooledit part of FDM can be analyzed relatively easily. For example Cooledit has a spectrum analyzer. This way audio snippets from either side of signal ( LSB and USB separately ) can show not only the modulation levels (and calculating the index - providing that signal is off tuned ) but also show any other carriers that lie inside the (pass)band with greater accuracy a/o strength that any HF receiver can show

2. Time domain analysis ( fading analysis)
This is the second part of the most significant information required for a station (I think ). From many measurements I did so far over the years I found that there are two types (at least) of fading , the one is a 'temporary' fading that can occur in a less than 2-3 secs and another fading that can occur in times of minutes
2a.Cases of 'Short' time fading found on SW:
Short time fading #1
This is the main technique to show in most cases. Though a little arbitrary drawing this is one of the most common cases found on HF. As anyone may understand from this drawing this is a sinusoidal curve ' and refers to only 'clear' signals ( free from QRM)
Example #2
Another clear signal found esp. on 5 MHz band Notice on how it was the lower part with a small fade up
Example #3 ( Angola 4950 )
This was another example with signal a little steady and sudden fade outs
Example 4 Co channel QRM
Unfortunately this drawing could not be easy to made by computer this is the reason i remade it by hand.

You see here how the signal fluctuates. It comes mainly when another signal is on (accurately saying: nearly ) co channel QRM with 'shortest fading levels- depths ' coming from the signal difference between QRMer and QRMed signals while the frequency comes from the freq deference between the two carriers

This case can happen also to clear signals with much higher fading levels ( depths) sometimes as in 18-4 -2000 when most of the signals on tropical bands had a fluttery of more than 4-6 S units with a >3 Hz rate probably due to ionospheric reasons
These are 4 examples found on the SW band but there are too many others I found so far
2b.Cases of long term fadings
From my observations there are not so many long term fadings. One can be a steady signal over the time and another with very deep fades occurring every say 5 mins ( this can happen mainly to stations less than 2500 km far from the RX system ) but here also are the fadings caused from fade ins , fade outs and ionospheric conditions as will be shown below:
Long term #1
Here is an example - case of R Tirana some years ago and nearly the same for R Netherlands IND service
The signal was steady for abt 6 mins then jumped down to S3 for less than 1 min
Here the 'short term' fading levels are not shown
Long term #2 fade in curves
Here is a reference to SBC Warna 7235 ( Signapore) two months ago. The signal on abt 1400 was too low of the levels S5 but gradually enhancing to stand up to S 20 from at least 1515
Nearly the same can happen with fade our curve

Case #3 : ionospheric problems
Here is a case found on 18-4 -2000 when listening to Namibia on 3290 and 3270 . The case here is for 3270 with a signal steady until 1810 at S9+10 db but afterwards gradually lowering toS7 and standing there
3. More advanced way for domain analysis
There is still another way for making a exact F/TDM analysis that can be completely made by computer. The method is relatively simple by showing a predetermined spectrum of say + /- 15 kHz - or a wide as cases require to show the spectrum in a given time say 1 hr and on a 3D analysis.
Though not 100 % sure for this it was a special model tracer from HP that with the combination for spectrum analyzer has given the results in a thermal paper

UPDATE COM ON BOYS!
4 Using R75 and specialized programs
4a Smart Icom 32 and R75
ICOM offered a control software CS-R75 or RS-R75 as optional . This year Aug 2001 i asked th e importer to order this control software at a price of USD 42.5 as they previouslsy told me.
However after two months of waiting, they finally advised me that this control software has been given up production. Fortunately from the ICOM R75 mailing list have found the software of Mark Fine Smart Icom 32 which operates directly with the receiver with the use of a serial cable 9 pin RS232.
Using this software for spectrum analysis was a but tricky at first but then it was very easy. Just you have to determine the l two VFOs first the A as starting point and B the ending point. Sweep radio was however very low and I had to adjust the line pacing ( in the preferences) as low as 15-18 in order for the software to control as fast as possible ( In the preferences dialog - communications tab) with speed as high as 19200 And in the scanning parameters with delay as low as 10 ms. And still the speed was very low at a rate of 5-6 steps per second!
With as many sweeps as possible you can have a real conditions result that can show minima and maxima for each stations you determine.. Stopping the sweep , for two sweeps you can have the average level , but for more sweeps i m not usre if thsi can show the latest sweep or the average level ( RMS)
The result of sweeps can be saved for later reference.
Notice that still under this way you can still see only the contiguous channels ,but not what s inside the channel.
And here is an example:
Here is ana example for spectrum analysis. This is for a report for V of Nigeria on 0909 on 11-11
4b : Use of DSP and R75
The idea has been shown above on the 1b and is hown here with more detail .The basic infrormation however is that you have to use SSB with offtuining( IF shift) so that carrier can be shown. as reference for all mesurements. All is then rereed with pictures
Part 1 : Spectrum analyzer. It has been refered posibly cprectly on the 1b, and can show in parts the side band separately for each band selection .
Here is a capture of a signal in USB . carrier is shown at abt 240 Hz . YOU see a low level carrier at 500 Hz or 250 HZ above carrier
RMS levels of 20 sec of signal on USB for 11785 and shift of 500 and other carrier presumably same level .
Part 2 Fading analyzer : In contrast to above . i think that the correct for identifying the fading is with the AGC off . Tis is because the AGC can alter the S meter indications. Also in other cases I filter out the modulation parameters so that carreri to be pure and not interfere with the resulted carrier information
Notice that i am not 100 % sure if this way can really show the fading. level as it is audio based.
Case 1 : Fading due to co channel interference. Abt 4.5 HZz of fading? Trace for abt 10 secs Normalized , full wave
A relatively stable signal ! ( BBC)
15060 CBS Taiwan at 1200 for10 secs short term fading This picture is a close up of the picture shown as below.
15060 CBS Taiwan at 1200 for 2 mins -middle term fading . Includes also the above picture You can see the fading is somehow periodic.

Please tell me your opinions by return email to me at

5 comments:

  1. Καταπληκτική δουλειά που μπορεί να φανεί πολύ χρήσιμη στην θεωρητική ανάλυση της ασύρματης μετάδοσης σημάτων σε κανάλια πολλαπλών διαδρομών με διαλείψεις.

    ReplyDelete
  2. Poli endiaferontes oi paratiriseis sou. To fading apo ti dimiourgeitai stis efarmoges autes?

    ReplyDelete
  3. Dear friend Ζαχαριας Λιαγκας

    I very much in the way you do your analysis and I think this material would be very useful to be released by the newsletter ATIVIDADE DX of Dx club of Brazil.
    So ask your permission to do the translation into Portuguese and published in ATIVIDADE DX.
    Iform the source that mention giving them the credit for the text.

    A hug,

    Adalberto Marques de Azevedo
    Adviser to the DX Club of Brazil
    Adalberto.azevedo@gmail.com

    ReplyDelete
  4. Adalberto Azevedo....
    YOU surely can link and translate this text to your page
    sorry that i did not seen your message but i use too many accounts in many webpages

    ReplyDelete
  5. sakis and vasilis :
    This study here is referred to the shortwave fading . Signals are passed and reflected to the ionosphere

    ReplyDelete