To All from PA0SE
In an e-mail to Bob, ZL2CA. discussing the PA0LQ audio filters, I mentioned
a thesis
on the reading of morse code signals in noise.
As a result Bob asked:
Do you know a way of accessing the university and thesis details? >
I gave the following reply which I also sent to the reflector because others
may be interested in it.
Bob,
The thesis has the title "Signal detection in noise, with special reference
to telegraphy".
The author is Peter Montnémery, SM7CMY, who is employed at the Department
of Otorhinolaryngology [say that fast ten times :-)], Head and Neck
Surgery, University Hospital, Lund, Sweden. Code number is
ISRN LUMEDW/MERL -- 1038--SE.
Some of the subjects treated are:
1. Recognition of telegraphy disturbed by noise at different S/N-ratios and
different telegraphy speeds.
2. Recognition of telegraphy signs at different listening levels and
frequencies.
3. Effect of dichotic presentation on the recognition of telegraphy signals.
4. Recognition of telegraphy in hearing-impaired telegraph operators.
5. Performance of electronic morse decoders in decoding telegraphy masked by
noise.
6. Detection of trains of tone pulses masked by noise.
I learned about the thesis by a message from Peter Schnoor, DF2FL, via the
LF-Group reflector, considerable time ago.
Via a friend at the Leyden University I managed to obtain a copy which took
many months to arrive.
After writing an article about it for our VERON magazine "Electron" (still
waiting to appear) I have passed it on to my good friend PA0CX/F2ZI who will
or already has sent it to Pat Hawker, G3VA.
As a further aside, do you know why Harry picked 1000 Hz for his filter?
I don't know.
For my filter I had to use 1000Hz because my German LF-receiver of the
sixties has a fixed BFO that produces an audio note of 1000Hz when the
signal is centred in the IF filters.
But when I am busy in my shack doing other jobs I often leave the receiver
on with a wide IF-filter selected, no audiofilter and such high volume
that the headphones function as speaker.
I have noted several times that I could detect a weak signal this way.
After putting on the headphones and switching on the 30Hz audio filter it
was often very difficult to read the signal! It again goes to show that
the ear-brain combination acts as tracking bandpass filter that should not
be "assisted" by too much pre-filtering.
Your "Morse Man", writing the column on CW in "Break-In", may be interested
in the subject as well but I don't have his e-mail address. So please be so
kind to pass this message on to him. Tnx!
73, Dick, PA0SE
|
