Greetings all:
Following thread re: tone enhancement et al and knew that I had some
references buried deeply within the catacombs of my basement. The
following is an article that I sent to Arthur Owen G2PUD several years
ago, when we were chasing NDBs and transatlantic openings over several
years. We chatted on a semi-weekly basis and this subject came up one
day, and I typed in the article and forwarded.... so please excuse any typing errors.
Attached is a GIF showing the two figures referred to in the article.
THE ENCLOSED ARTICLE is from ham radio magazine - October, 1974, pp 53-55.
THE SECOND ARTICLE ( which I have not included, but do have the original if
anyone is interested ) is from QST, April 1983, pp.32-34. Titled
"A Dichotic Detector for CW", by Douglas Kohl, W0HTM.
A technique is described based on the ability of the human ear to
distinguish between very close audio frequencies -
to quote " the dichotic detector takes advantage of the brain's higher
level processing of the auditory neural impulses - physco-acoustics!
The result is the same as narrowing the receiver's bandwidth to just
a few hertz, but without reducing the speed of response, as with a
high-Q filter." Circuitry is included......
This seems to be a legitimate article - but I have seen some very
interesting concepts appear in other -APRIL- issues, and some of them even work !!
In the same issue, pp.35-36 is an article by J.B.Heaton, G8JFY and
R.V.Heaton, G3JIS, entitled " An Electro-Acoustic CW Filter". Though
not using a "crystal" resonator, it does make use of a glass tumbler ! ! ! !
therefore saving significantly on costs !!!!! :-)
73 to all
Mitch VE3OT
ARTICLE ONE FOLLOWS
Enhancing CW reception through a simulated-stereo technique
Ham Radio - October 1974, pp. 53-55
A sharp audio filter is a great help in copying CW signals through noise
and interference. A filter passband of 100HZ will accomodate the keying
bandwidth at 20 wpm and tolerate some short-term drift of transmitter
and receiver. However, such a narrow filter makes scanning of the band
slow and difficult, and with high skirt selectivity the character of the received
signal and its attack and decay may become distorted from ringing.
Threshold Gating - Higer noise reduction than possible from a tuned filter
can be accomplished through threshold-gating, a technique in which the
received and filtered CW signal is used to key a relay or an electronic switch.
In turn, the relay or switch feeds the received and filtered signal or a sidetone
oscillator to the headphones or speaker. Threshold gating achieves is slectivity
through the switching process. No signal is heard off-frequency or between the
dots and dashes. However, the original signal is highly distorted, especially in
the attack and decay, or is completely replaced by the sidetone. As a result,
feel for band conditions and for the quality and signature of the signal are lost,
and any relpy slightly off the filter frequency will not be heard at all.
This discussion suggest that you cannot, at the same time, use high filter
selectivity and retain an essentially unaltered CW signal and full feel for the
signal and the band. Therefore, sharp filters seem to have little vale in contest
and net operations, where you want to respond rapidly, often to signals which
appear to one side of your receiver center frequency.
Another approach - This is certainly true as long as you listen to the filtered
signal only. However, if you were to listen simultaneously to one speaker fed
with processed signal and another speaker reproducing the raw CW signal, you
could retain a feel for the band and would even hear chirp and clicks that extend
beyond the filter passband. Also, if the received signal were drifting you would
not lose it, and you could hear replies at some distance from the center frequency
of your filter. Obviously, you would lose some of the advantage of the filter and
QRM and QRN would have returned to some degree.
To this end, you can be helped by the ability of the brain to discriminate between
signals which appear at both ears or only at one. To do this, the incoming CW
signal is divided into two channels as shown in Fig. 1. Half of the signal goes through
a sharp filter to one ear, and the other half is passed unfiltered to the other ear. The
level or balance control in the unfiltered side compensates for any attenuation of the
filter. It is set so that a tone centered in the filter passband is heard by both ears at
the same loudness level. This tone is then perceived stereo-phonically and appears
centerd within your head. (Because this effect may be enhanced or diminished by the
phase relationship of the audio in the two earpieces, it would be desirable to try
transposing the leads to one of the earpieces to see what would happen - editor )
Any other tone is attenuated by the filter and will be heard predominately from the
unfiltered channel. Consequently, all signals that are not passed by the filter appear
to come from that side which is receiving the unfiltered channel.
Results - The effect of this simulated stereo reception of CW signals is dramatic.
Interfering signals and broadband noise appear to come from a point off to one side
of the head. The desired signal, centered in the filter passband, appears within or just
in front of the head and assumes a transparent clarity that is hard to describe. The
signal-to-noise ratio is much improved. The character of the signal is preserved and
ringing is wither absent or less apparent than in monaural reception with the same filter.
Chirps and clicks are readily noticed, and even drifting signals or DX signals with
multipath distortion are readily copied.
The mind seems to concentrate automatically of the desired signal and to be relatively
unaware of and undisturbed by the signals outside the filter passband. Yet, that
information is present and an off-frequency reply is heard just as well as if
no filter were in use.
Practical Considerations - I have used this approach with a single passive toroid filter
and with a more complex filter and threshold-gate combination. Both are effective.
The latter has an advantage on some occasions, but for its simplicity and ease of
adjustment the toroid filter is superior. The high impedance audio signal from my
HW101 is matched to the low impedance of the series-tuned toroid filter by the
transformer, as shown in Fig. 2. The centre-tapped secondary provides two equal
signals, one attenuated by the balance control and then fed to the right channel of
a low impedance stereo headset.
The toroid filter in the other half of the secondary resonates at 790 Hz and has a
3-dB bandwidth of 60 Hz; this frequency was chosen because the apparent stereo
separation increases at low frequencies. The output of the filter is fed directly to
the left channel of the headset. The balance control should be adjusted on a
moderately weak CW signal; a strong, steady tone, such as from a calibrator, gives
a slightly different balance.
First, peak the signal in the filter by listening only to the left earphone, then put
on both phones and adjust the balance control until the signal appears centered.
The range of the balance control is sufficient to move the signal from the far right
across the center to the left. Little further adjustment is required under differing
band conditions.
The principle can be applied in various ways. Other input and output impedances
can be accomodated with different transformers, or a parallel-tuned toroid filter
( approximately 500 ohms) could be used. Other filters could be substituted,
and instead of earphones a stereo amplifier and speaker combination used to
give a good demonstration of simulated-stereo CW reception to interested
listeners at a club meeting or hamfest.
References: Joh J. Duda, W2ELV, ³Noise reduction for CW reception,
² ham radio, September, 1973 (threshold-gating/ limiting and energy integration
to reduce noise interference on CW )
FIG 1 Block diagram ( attached)
FIG 2 Simulated-stereo setup uncluding simple series tuned toroidal filter.
(attached) showing a small centre-tapped xfmr (the splitter), in this case
500 ohm primary with 8 ohm centre-tapped secondary.
|
|
