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

[linrad] RE: EL2RL an UNKN422 record

Hi Matthieu and all,

> I have downloaded the EL2RL record file as processed by LINRAD.
It is not Linrad, it is one of it's predecessors, but the result
should not differ much.

> The EME operators around me find it impossible to copy any full 
> call. 
Yes, this is just one single minute from a long qso. I do
not remember the details any more but I think I needed about
10 one minute periods to convince myself that EL2RL had my
call correctly in this random qso. It was impossible to
really copy a full call sign at any time , but the summed
information of many fragments, the last of which you have
listened to, made me sure that EL2RL had written SM5BSZ
into the log-book.  

> With the UNKN42 file I can copy "9y" , the end of af9y call.
> Someone could tell me how many extra dB would be necessary to 
> copy exactely the weak station, maybe 1dB? 
Depends on who you are;)
Mike, AF9Y himself claims to have copied the call by ear just by
listening to the file through an ordinary (narrow) filter.
The challenge has been around for many years now and there is one
person who has solved the problem and received USD 100 from Mike.

So the answer to your question is: AF9Y and Gary Huntress (the 
first winner and this far the only one) do not need any more
signal to copy the weak signal.

The average EME operator would probably need 3dB but I am sure 
a single dB would have given too many winners for the challenge
to still be around here: http://af9y.com/radio30.htm
Personally I think I could copy the unknown call with about 1.5
dB more signal. Had the stability been better so coherent processing
had been possible I am pretty sure I had solved it already.
Had the keying been done with a machine so averaging would have
worked, incoherent averaging would have solved the problem easily.

Since my ears/brain is inadequate for the problem I am working
on making the computer (Linrad) clever enough to give the assistance 
I need. Maybe, one day...........

> How strong is the libration in these two records? I would like to 
> know how to train my ears to be able to recognize the various 
> effects of the EME path, and I think these two records are good examples.
> Thanks for trained eme operators to give me there feelings about 
> the level of difficulties of these to records.

These recordings are at the very limit. The UNKN422 is pretty strong
but the QSB happens to make AF9Y decodeable at three different places
but the unknown signal is in the qsb minima in between. The frequency
drift is fairly rapid and I have personally not been able to track
the phase of the carrier through the weaker sections of the recording.

I would suggest you train your ears on "normal" EME signals.
There are plenty of them on the Internet, some at the EF9Y site.

Libration fading is always the same in strength. The difference
is the rate that may change from time to time. These recordings
are typical for 144 MHz EME. Much slower rate is very unusual, but
it happens. The qsb rate may go from a peak duration of typically
two seconds to a peak duration of 20 seconds. I heard that only
one single time in my life in a qso with VE3ASO. The quiet periods 
lasted in the order of a minute. It was really like short distance
tropo conditions. Do not expect the qsb rate to differ from what
you hear in "typical recordings" it is really rare.

Note: If you do not have X-pol, you may have ionospheric qsb caused
by random variations in Faraday rotation when your antenna is close 
to 90 degrees away from the correct polarisation. When the signal
is far below it's optimum, the level depends critically
on the polarisation angle.


Leif / SM5BSZ