This FAQ is an attempt to answer some of the more common technical questions about HSCW - questions not covered in the general FAQ. However, this will not go deeply into any one method of generating HSCW, such as MS_DSP or CoolEdit, as they have or soon will have their own sections, either on this or another Web site.
WHY CAN YOU NOT USE THE TRANSMITTER'S REGULAR KEY JACK FOR HSCW?
The shaping circuits prevent most modern transmitters from being keyed this
way at speeds above 80-100 wpm (400-500 lpm) without the keying becoming very
bad and even completely unreadable. A few may go higher, some won't go that
high. While it is possible to modify many rigs so that they will transmit
at higher speeds, the keying is then not usable for normal speed CW. This
is why nearly everyone, both in North America and in Europe, uses an injected
audio tone.
Usually. However, some modern rigs have a "data in" or "AFSK in" jack which may be
more suitable. Check your rig's manual or try out both, if available.
This varies with what program or device is being used. More details can be found in other
sections covering the specific programs or devices. But here is a summary.
MSDSP uses the Sound Blaster audio card to generate a tone. For the transmit-capable
versions, a two-line stereo cable is plugged into the SB board's Line Out. This is because
the left channel is used for the transmit tone, while the right channel is used to
simultaneously carry the playback audio. From the line going to the left channel,
a shielded cable is run to the mike (or AFSK In) jack. But there should be a small
capacitor in series with this line at the mike jack. This is to block any DC voltage
that may be present the at the mike jack, to drop the level slightly, and to pass only
the higher audio tones. The exact value will probably be small, several hundred pF,
up to .01 uF. Use the smallest that gives approximately the same amount of audio to the
mike jack as you get from your mike. This, like many other things, will have to be
determined empirically ("cut and try"). (See the section on MSDSP when it is available
for more information on MSDSP's transmit audio. An R-C shaping circuit should soon be on:
Several programs have a keyed output on the RS-232 port. These require a simple
transistor or opto-isolator as an interface between the computer and a separate audio
oscillator. The output of the audio oscillator is then fed into the transmitter's mike
input, again using a shielded cable and a series capacitor. Diagrams for these can be
found on the N1BUG Web pages. Also see several
of the following questions.
K7CACW uses the computer's speaker output.
A programmable keyer, such as the CMOS Super Keyer 3, simply keys a separate audio oscillator, which then feeds the mike jack. Note, however, that with at least the standard XR-2206 circuit, there MUST NOT be ANY capacitor across the output of the keyer or the input of the XR-2206 circuit. Anything more than a few pF will cause all of the elements to completely run together.
BUT WHEN USING A PROGRAMMABLE KEYER, DON'T YOU NEED SOME RFI BYPASSING OR SOMETHING ON THAT
OUTPUT LEAD?
Probably. But most HSCW operators who use this keyer put the audio oscillator into the same box. The capacitor across the output of the keyer (C1, a 0.01 uF, across the C & E of Q2) is removed. Since it makes good sense to also be able to use this very versatile keyer for regular CW, they put a switch on the front panel of the keyer to switch the output of the keyer between the audio oscillator and a jack on the back panel. You can then put a capacitor across this jack, since it is switched out of the circuit when the audio oscillator is being used. This has worked fine for use with 1.5 Kw finals on 144 MHz.
When you slow MSDSP's transmissions down to 20 wpm and play them through a speaker, they are
filled with a lot of pops. This is because of the way the tone is being generated and the
timing involved. Using the small cap in series with the audio line will remove most of the
low-frequency component (and the 60 Hz hum that is often present). The pops generated by the
computer are primarily high-frequency, and are filtered out by the circuits in the rig itself.
(In actual use, it is not run at 20 wpm, but rather at a much higher speed. An injection tone
of 2000 Hz also helps keep the unwanted components in the range where they will most easily be
filtered out). This was initially a major concern to those using MSDSP; but the author of the
program assured everyone that it would work fine. So while it may sound bad to you, WAV files
of even very strong bursts have produced no evidence at all of these pops when used on the air.
Also, be sure that the audio compressor is turned OFF to cut down any white noise from the audio
board or any other extraneous signal on the audio line. For more, see the separate technical
notes on MSDSP (which are still being prepared, 12/97).
Hook a shielded line from an audio output of the receiver and feed it to the input of the
Sound Blaster board. You can use the speaker or headphones output, or a separate audio
output, depending upon your rig and your preferences. Since you will be working primarily
with tones in the 1500-2500 Hz range, you will probably want a small series capacitor in this
line, also. You may need some bypassing, ferrite beads, etc., for keeping your signal out of
the audio board. Also, the power supplies of many high-power amplifiers have been known to
put a considerable 60-Hz hum on this line. The series cap will help with this; but also
consider where your cables run and where your amp's power supply is located, if you're using a
tube-type amplifier.
Most computers have an external audio amplifier and speaker set-up. Most operators put a
stereo jack in the amplifier/speaker enclosure to disable the speakers and drive a pair of
headphones. (K6STI says that you need a 100-Ohm resistor in the lead to each of the pair of
earphones). The $3.00 lightweight phones sold in any drugstore for use with a
portable "Walkman"-type unit have been found to be good for HSCW. They do need to have a good
frequency response, as you're dealing with both low and high pitches. And don't forget that
the headphone cable can be a major source of RF pickup, which will tend to mess up your audio
amplifier if you're running high power and are using MSDSP for transmitting and playback at
the same time.
HSCW is definitely a weak-signal mode. But the bandwidth is somewhat wider than the optimum
for EME work. At this time, the use of an audio filter is controversial, as not enough tests
have been run to produce any agreement on this question. You will have to experiment with this
yourself.
You need to receive in the USB position of the receiver, using a 2.4 kHz or so filter.
Several feel that putting a DSP audio filter, in bandpass position, between the receiver and
the computer does not help and probably makes copy more difficult; others disagree. When the
signal has been slowed down and heterodyned to a more comfortable level, it appears that this
is where the audio filter would really help. Some have found that it does improve the S/N
ratio, others have not seen much difference. All agree, however, that it is tricky to get the
filter's frequency and width set just right. At this time (01/98), this is a matter that is
still being tested. But since everybody's "between-the-ears filter" is somewhat different,
your results may not be the same as others. Try it, and please let us know your results.
For a start - put it in USB, AGC to Fast or OFF, Squelch open, Audio Gain about normal, USB
filter, RIT to zero (or maybe +0.5 kHz), Pre-amp on, IF Shift rotated to near the maximum for
giving a high pitched no-signal white noise from the speaker, Main Tuning 2 kHz below the
ZeroBeat frequency of the schedule (more on this under the "Procedures" for HSCW paper).
Some rigs handle HSCW fine with the Noise Blanker on, some cannot receive HSCW at all unless
it is turned off. These are general settings. You will have to experiment to see what is
best for your equipment, especially the NB and AGC.
You need for the received pitch to be as high as you and your equipment can handle. The amount
you slow down the keying is also the amount that you will reduce the pitch of the tone. It
becomes quite difficult to copy a tone below 200 Hz or so. MSDSP allows you to heterodyne it
higher in pitch, but with a loss of intelligibility. SBMS does not have the heterodyne
function. Either way, if you are running at low or medium speed HSCW and can receive at a
high pitch, then you should still be able to copy the slowed-down low pitch tone. A received
tone between 1500-2000 Hz seems optimum for most rigs and operators. Again, you will have to
find out what best suits your rig and hearing.
This has to do with the difference in transmit speeds, and (probably) the fact that a modern
rig operated in CW mode offsets the signal by about 800 Hz.
If your rig is capable of working SSB or CW MS, aurora, or tropo, it is probably capable of
HSCW MS. So even a simple SSB rig should be usable. There are several additional "bells and
whistles" that are especially useful for HSCW. The most important would be an IF Shift Control,
a RIT with a range of +/- 5 kHz or more, and for calling CQ a 2nd VFO or 2nd receiver can be a
big help.
For the transmitter, the more power the better. Unlike other modes, doubling the power on
HSCW seems to give a 1.4-times increase in the number of underdense pings (while the number of
overdense bursts stays relatively constant). This is another reason why smaller 144 MHz SSB
stations have tended to operate primarily during the peaks of showers, trying to catch the
overdense bursts, where they are on a more equal footing with the big guns. An amplifier in
the 150-watt range is adequate for much HSCW operation, tho more power, of course, is better.
But if you don't have an amp, try it anyway. HSCW contacts have been made with as little as 5
watts!
On VHF, the antenna system is usually the most important piece of equipment. It does not
have to be large - in fact, an array large enough for EME operation is really too sharp for the
best MS work. You need gain, but not at the expense of a very sharp front lobe. The old-timers
swore by the 16- or 32-element collinears or a quad array of 5- to 10-element Yagis. These are
still hard to beat for MS! A single medium-size Yagi is a good "modern"
choice.
There are three notes that must be made here, however. The first is to use an efficient
Yagi design. Do NOT be misled by "gain" figures as published in some catalogs. Many of the
antenna companies, including even some of the large ones, play the "Gainsmanship Game", as it
has been called, trying to lure the less-knowledgeable operator to buy their products. So as
not to recommend specific brands, if you are buying an antenna (for 2 meters) larger than 10
elements, be sure to get one that is advertised for the LOW end of the band, rather than for
FM work. The gains on these tend to be more honest, for the operators who usually purchase
these (especially the large ones) already know their capabilities and will shun those with
inflated gain figures. Second, get it just as high as possible. Then put it 10 feet (3 M)
higher! A high 7-element Yagi will outperform a low huge Yagi array on MS! And you need this
height for everything except EME. So, before you plan a new!
installation, consider making it yet a little higher! And third, use good feedline. Height
means distance from the rig. For a low Yagi, Belden 8214 coax is OK. But this really is a
weak-signal mode. Consider one of the even better types of coax for longer runs. (Cable TV
coax is often available, in throw-away tails. You make a 50-to-75-Ohm transformer, and these
can also be purchased ready-made. Or you can make the run a multiple of a half-wavelength.
See the N1BUG Web site for more). And if you have a really long run, consider a mast-mounted
preamp. However, do NOT install this until you know what you're doing, or you will fry the
preamp. This requires an extra run of coax and a sequencer, which we can't go into here.
(All EME operators use this).
And one more note. While you can't operate HSCW without a method of generating it and
slowing it down, don't forget the one absolutely vital piece - the OPERATOR. A good operator,
who has had the patience to teach himself what to do and who has practiced it, can work stations
that others can't even hear! Ultimately, the "nut behind the wheel" is the final bit in the
equation of how well a rig will work!
At this time, all of them require a '386 or better (the faster the better, of course, but the
'386 seems to be adequate for all but CoolEdit which really needs a Pentimum for on-air use),
VGA monitor, and the usual expected items. Most important for HSCW, and for all the other
programs that are of the "DSP" variety, is a true Creative Labs Sound Blaster audio card. An
SB clone will usually *NOT* work. This is because all of these HSCW and the other DSP programs
(FFTDSP, DSP Blaster, etc.) use the SB audio board for much of the actual processing. These
are DOS programs; and while some will run in a DOS window under Windows 95, several (including
MSDSP) will NOT run at all under Windows 3.1x. (SBMS runs under DOS or W3.1, not W95, and has
been known to work under at least one SoundBlaster clone).
A special note for those considering MSDSP - The current version (0.51) is still a Beta
version, and has several bugs that have not yet been corrected. One of these bugs prevents it
from running on some seemingly compatible computers. The reason for this is not known. (If the
latest version will not run and you have a true SB audio card, try V. 0.34). For more on
MSDSP, see the MSDSP section on this Web site, which hopefully will be available by the end of
January 1998. Or contact one of those now using MSDSP).
Answer: Any and all that will run on your computer! First, check the various Web sites. By
the time you read this, it may be greatly out of date and several other versions or even whole
programs may be available. So check the links to the various Web sites to see what is new.
But here is a summary as of 01/98.
SBMS, available from:
The higher the speed, the higher the tone needs to be. An even higher tone, such as
2500 or 2700 Hz, might be better (so long as it passes easily through the rig's filters).
However, 2000 Hz has worked well at the highest speeds currently possible (about 8500 lpm).
At the lower tones and higher speeds, it is possible that a dot will be shorter than one cycle
at the tone frequency! There is nothing "magic" or even necessary about 2000 Hz. Having it an
even 2 kHz simply makes it easier to figure the dial reading offset. 1000 Hz has proved to be
OK up to at least 4000 lpm. Nearly all North American HSCW operators use tone injection and
run schedules at about 4000 lpm, so an injection tone of 2000 Hz has simply become the standard.
MS_DSP (or MSDSP), available from:
Cool Edit, a shareware WAV file editing program, from Syntrillium Software, was never designed for this purpose. It is a powerful audio program. But, with practice, it can be used for both transmitting and receiving HSCW. In fact, it was discovered by two European operators who could not work CW but wanted to operate HSCW MS! They used this program to read it off the screen. For a full tutorial of this program, see:
K7CACW. This is a relatively new transmit-only program. Originally designed for EME and similar work, modified for HSCW. In its current version, V. 2.2, it has a maximum speed of 2000 lpm. It sends its audio to the PC speaker, which is then tapped off for the transmitter. It is freeware, available at:
KB0VUK is attempting to translate the data on a French program. See:
for further details.
There probably are more MS and/or transmit/receive programs available. If not, no doubt there soon will be. Most of the above, except OH5IY's multi-feature program, are small and don't take up a lot of hard drive space. (Unless you use MSDSP or CoolEdit and save a lot of WAV files, that is).
I DOWNLOADED SBMS BUT IT WON'T RUN. WHAT'S WRONG.
Two possibilities.
The first is that you do not have a Sound Blaster audio board.
A second problem that many have run into is this program's need of a virtual
ramdisk.
This is not really a problem, but it is not adequately explained in the program's instructions. So you have two choices. If you already have a ramdisk set up, use it. Edit the second line of the SBMS.INI file to point to it. (For more on ramdisks and the commands necessary to create one, see your DOS manual).
Many have found that there is a better method. You simply set up a
directory named RMDSK, and the program will use this in place of a ramdisk.
(It takes up almost no hard drive space,
for anything in it is erased when the program is terminated). To do this, first change to the
root directory of your hard drive (probably C:\). Then, using the MKDIR command, make a new
directory there named RMDSK. This directory will function as a "simulated" ramdisk. (On slower
computers it may cause some hesitations in playback, but has worked satisfactorily for most).
Then edit SBMS.INI as follows: On line 1, put the correct path to the program. Line two should
read C:\RMDSK\. If it does not, change it as necessary (or change it to reflect where your
"simulated" ramdisk directory is located). SBMS should now run. You will also need to edit
the line to give the proper delay for your reaction time, and the line for slowing down the
received material to the correct speed. But these can be changed later.
No! Not if you are willing to spend a little time in the workshop. HSMS was popular in Europe
for many years before computers were in any ham shack. The use of high speed keyers like the
CMOS Super Keyer III is discussed below. For receiving, an ordinary (and inexpensive) cassette
tape recorder can be modified to record the CW at high speed and play it back at a slower speed. The exact modifications and operating procedures vary from unit to unit, and it is not possible to give exact details here. See N1BUG's Web site for further information. The basic technique is to add a motor speed control, which can be as simple as half a dozen components. The motor is then run as fast as possible while recording, and slowed down to play back the HSCW at slow speed. The speed reduction possible varies from unit to unit and depends also on the motor speed control methods employed. However, you should be able to handle 1200 LPM easily, and some machines will handle 2000 LPM and beyond. Ev!
en 1200 LPM is much faster than any SSB, and will allow you to make a lot of otherwise
impossible contacts. Open-reel tape decks have also been used, and if you are lucky enough to
find a 4 speed unit you should be able to work speeds of 1000 to 1500 LPM without making any
modifications at all. These speeds may be somewhat slower than typical for North American
skeds. Don't worry. Other operators will run with you at these speeds. Not having a computer
is no reason to give up the idea of working HSMS!
It is believed that several can be modified to run at lower HSCW speeds, but no information is
currently available on any of then, nor any easy way to calibrate the speed control.
The one keyer known to work well is the CMOS Super Keyer 3, described in QST for August
1995. (Kits are still available from Idiom Press, Box 1025, Geyserville, CA 95441 for $55 + $3
S&H US [$5 S&H surface foreign]). It is capable of 4950 lpm and must drive a separate audio
oscillator for HSCW (see previous questions on that). It is not as convenient to program in
real time as MSDSP, but is much easier than CoolEdit. Also, it is an excellent general-purpose
programmable keyer with up to 18 memories available in 3 banks. See the QST article for more.
Because of its cost and the fact that it was designed for European style HSCW, none are in use
over here yet, although several are considering a purchase. The Europeans say that it is
great. However, its speeds, tones, and filters are designed for European HSCW, in the
1000-3000 lpm range. Since most of the experienced North American operators are now running
many skeds at 4000-5000 lpm, it may require a firmware upgrade before it is suitable for use
over here at the higher speeds. But until one or more have actually been tried, nothing more
can be stated at this time. Obviously, everyone is anxiously awaiting an actual hands-on test
report when used under North American procedures! Follow the hyperlinks on this board for
more information and technical data on the unit.
Sure. After the usual playing with the functions and some practice, do this for easier HSCW
operation. Make a little card showing the /U codes for the various HSCW speeds (1000 lpm - /U20,
4000 lpm - /U80, etc.). Then put a piece of masking tape or a slip of paper in front of the 6
memory buttons for labels. (For HSCW, here's mine - #1, calls. #2, calls, report. #3, R,
report. #4, RRR. #5, 73. #6, open). Also, on your sked sheet, put the Bank (and, if
different, the memories) to use for each station. Since you can't tell for sure what you are
sending, you must use these memory aids for this or any similar memory keyer. Even so, you
will occasionally find yourself sending the wrong data, for we ALL have done this. Use all 3
banks, one bank for each sked. Or, if you are running several skeds at the same speed, memories
#4 & #5 would be the same, so you could have even more skeds set up ahead of time. #3, the
report, can probably be set up ahead of time. Unless it is near the peak of a shower, the
chances are that your report will be "26". You can pre-set this, changing it if necessary.
Since the keyer is programmed via the paddles, and since you are going to have a function
code to put at the beginning and end of each message, there are two possible ways to set up the
first two memory positions:
Method 1
- In memory #1, put /Uxx (speed), his call, your call, /1 (to repeat memory #1).
In memory #2 put /Uxx, his call, your call, report, /2. This is the simplest. But then you
may have to reprogram it if you get a major burst very early in the sked and the report already
loaded is not correct. (Depending upon what he is sending, it equally well may be memory #3 that has to be reloaded).
Method 2
- In memory #1, put /Uxx (speed), his call, your call, /2 (to go to memory #2). In
memory #2 put ONLY /1 (to go back immediately to memory #1). If you need to send calls and
report, reprogram memory #2 with /Uxx, report, report, /1 (to go back to memory #1).
Any of these two (actually, 3) methods works equally well, both take a number of seconds to
program. It's a matter of personal preference, the number of skeds and how close they are
together, whether you can make a good guess of the report you will send to him, etc.
A stability of perhaps +/- 100 Hz is more than adequate for HSCW. More important is whether or
not it generates good sine waves, and how well it will key.
It sure does for me! I have a strip of paper reminding me what the first 6 number keys transmit. That way I don't have to look at the screen to remind me. HSCW can move fast if you get many pings!
This can be difficult. But it is necessary, for several stations have found that their keying
was completely unreadable when they first tired to transmit.
If you have a separate receiver, separate computer, and oscilloscope, it should be rather
easy.
Lacking all that, but having a separate receiver, there are several tests you can make.
Transmit (into a dummy load) and record a snatch of your signal onto audio tape. A simple
hand-held audio cassette recorder is adequate. Re-cable and play the output of the recorder
into the computer using one of the HSCW receiving programs. Slow it down, play it back, and listen carefully. But it has been found that the ear can copy code which really is not keying well. So record a snatch of this playback and look at using CoolEdit, MS Wave Studio, or other audio program. Select several characters and zoom in. Zoom in and out, looking at the beginning and ending of each dot and dash, and, if using an outboard audio oscillator, see if the space between elements is very close to being the same as the length of a dot. Do this with several different speeds, especially the higher ones.
If you do not have a separate receiver of any type, about all that you can check is the
keying from your keyer or computer. Again (if using a computer), use a tape recorder to record
the tones directly from the speaker.
If there is another Amateur not too far away with an all-mode rig, ask him to record a
little bit and play it back, perhaps on FM, over the telephone, or send you a short wave file
over the Internet. Then "receive" this and study it.
Once you have the obvious problems solved, try to run an "easy" sked with another HSCW
operator. Request him to send you a wave file of at least one strong ping on you, then load
this into your two programs and study them.
Occasionally check the quality of your signals! You may have developed excessive hum or
other problems as cables have been moved, etc.
If you hear a problem on another HSCW signal, try to save a wave file of it and send it to
him, along with a description of a problem. Remember, they have no method of knowing that
problems may have developed! (Note in the "Procedures" that the "U" signal means "Ur keying
is defective". Once you have seen the difficulty of properly checking your own signals, you
will appreciate having this signal in your arsenal of HSCW techniques).
You probably do. Record a snatch of a high-speed ping or your own high-speed HSCW signal into
a playback buffer, then play it back at all possible speeds. (Practice pings and pings of
various speeds are also available as wave files on several of the Web sites). You may find
that there are certain speed ranges where the slowed-down speed seems to be very much different
than what it should be, and the keying is gibberish. But when you get the speed into the
correct range, it suddenly becomes readable. You must slow it down the proper amount -
neither too fast nor too slow.
In general, the direct path is used. There are, however, some special circumstances which
would make it desirable to offset the antenna from the direct path. This may be because of
the position of the meteor shower's radiant (and thus, the angle of the arriving meteors) at
the time of the sked. Or, because the other station is less than about 500 miles from you, or
there is a major obstruction in the way of the direct path. Yet another reason is when a
really big antenna array (EME size) is used, which would have a very narrow front lobe.
(In theory, there is a "dead" spot of a few degrees wide in the exact direction of the other
station at the optimum sked time). Since the ionization of the typical trail is about the
height of the E layer, it is difficult to work stations closer than 500-600 miles without
using side scatter (that is, each station aiming their antennas at a common area of space
about 400-500 miles distance from each).
This subject is too large to cover further in this FAQ. This may be covered further if
we ever get a major work on MS itself ready for the Web pages (which is doubtful, since a number
of very good articles have already been printed). The most important article is the second one
by W4LTU, published in QST for May 1974, and reprinted in "Beyond Line of Sight", available
from the ARRL. It is strongly recommended that this article, the one by W9IP, and the one by
W1JR be studied.
Probably, if you can do so. The exact amount depends upon the distance of the other station and
the amount of offset in azimuth you are using. Only the first can be covered here.
For a direct heading, the approximate elevations and distances are:
400 miles, 15 degrees elevation. 600 miles, 10 degrees elevation. 800 miles, 5 degrees
elevation. 1000 miles, 0 degrees elevation.
When running many daily skeds from central KY to central FL with large antennas on each
end, a very definite peak was found at about 7 degrees elevation. Above 10 degrees or below
5 degrees, the number of pings dropped off considerably. (Offsets in azimuth were less
definite). On daily skeds from central KY to FN42, there seems to be a peak at 5 degrees; but
it is less definite than the peak on the FL skeds, possibly because of the difference in local
terrain in the two different directions.
For more information on the elevation for various offsets in azimuth, see the OH5IY
MSSOFT program and the W4LTU article.
The exact answer depends upon what you mean by your question. An overdense burn requires a
larger meteor, produces a much more densely ionized trail, and produces a much stronger,
specular reflection than an underdense burn.
More specifically, "Underdense trails are those wherein the electron density is low enough so
that the incident wave passes through the trail and the trail can be considered as an array of
independent scatterers. Overdense trails are those wherein the electron density is high enough
to prevent complete penetration of the incident wave and to cause reflection of waves in the
same sense that the ordinary ionospheric reflections occur. A rough sorting of trails into
these two categories can be done on the basis of trail lifetime or duration."
(Ionospheric Radio Propagation, U. S. Department of Commerce Monograph 80, 1965, p. 356). Cf.
The W4LTU article, above.
Speed, speed, speed. For a small fraction of a second, an underdense "ping" may produce a
signal which has been reflected from the very head of the trail (where the ionization is most
dense) that is quite strong. These are many, many, times more frequent than
overdense bursts,
except near the peak of a shower. Since SSB requires a ping about 1 second long to propagate
even the smallest amount of useful data but HSCW can move a large amount of data in only a
fraction of a second, and since there are usually hundreds (thousands?) of fractional-second
underdense pings for every overdense burst, HSCW is set up to use these very short pings instead
of waiting and hoping to catch that once-per-day "blue whizzer".
Nope. Like everything else, there is a trade-off. HSCW is far better than SSB most of the
time; that is, during non-shower periods. But near the peak of a major shower, SSB may be more
effective. This is because during a shower peak, there usually are more overdense pings of 5
seconds or more, allowing SSB to sometimes complete an entire QSO exchange during one overdense
burst. HSCW is not a "break-in" mode, and, in fact, HSCW operators find long overdense bursts
to be exciting but unnecessary and even a nuisance!
Here are some references on meteor scatter, as contributed by a number of people. They are by
NO means complete:
Many issues of the old VHF Horizons magazines.
A number of the VHF columns in QST and CQ magazines.
J. Reisert, W1JR has an article on MS with a good reference list in Ham Radio, June 1984.
Handbook of Visual Meteor Observations, the International Meteor Organization.
Ionospheric Radio Propagation, published by the National Bureau of Standards,1965.
Diurnal variations in forward-scattered meteor signals, C.O.Hines, Journ. of Atmospheric
and Terrestrial Physics, 1956, vol. 9, pp.229-232
Proceedings of the IRE, December 1957, p. 1642-1743. This is 100 pages on MS! The old
bible on MS and highly recommended for the MS operator, if you can locate it.
Radio Propagation by Reflection from Meteor Trails, George R. Sugar, Proceedings of the
IEEE, February 1961, p. 116-136.
AGARD Conference Proceedings No. 382, NATO 1985-86: Propagation Effects on Military
Systems in the High Latitude Region: Paper 9.1 and 9.2 on Meteor Burst Communications.
AGARD Conference Proceedings No. 419, NATO 1988: Scattering and Propagation in Random
Media: Paper 44.1
AGARD Conference Proceedings No. 486, ISBN 92-835-0601-4
AGARD Conference Proceedings No. 244, 1978, ISBN 92-835-0219-1
(Hughes Aircraft Co. did research meteorscatter around 1957-61. They have built a 2x3
stack of 8 elm yagis for 50 MHz, 48 elements! Quite impressive picture of this monster
in above references. Guess some hams can tell their experience 40 years ago).