On Oct 31, 2009, at 6:29 PM, Bill, W6WRT wrote:
>> Like it or not, RTTY (170 Hz shift 45.45 baud FSK) is not
>> suited to the MF range - even in times of low sunspot numbers
>> which should represent the most benign conditions. In normal
>> conditions, FSK RTTY and other modes that are highly sensitive
>> to multi-path should not even be permitted at MF just like
>> wideband FM is not permitted below 29 MHz as it is simply
>> wasteful of the limited bandwidth.
>
> REPLY:
>
> Absolute nonsense. Anyone who has spent much time actually operating
> 160 meter
> RTTY knows it works just fine. This is one canard that should die a
> quick and
> merciful death.
The point is not that you cannot use the low bands for RTTY. No one
has claimed that in this thread.
The point is that steam RTTY requires much higher SNR to operate under
propagation conditions that includes echos, than it does under
propagation conditions that includes selective fades (that is why the
experimenters eventually chose more than one tone for RTTY in the
first place -- early amateur RTTY experimenters had used OOK (one on-
off tone), but went to FSK to avoid selective fading).
In case you think no one else has an interest in the low bands, here
is another paragraph from ZL1BPU (http://www.qsl.net/zl1bpu/DOMINO/Index.htm
):
> Other developers have concentrated on getting around the problems of
> conventional MFSK by using wider, faster transmissions with strong
> FEC (for example ALE and Olivia) The FEC is used to mask the
> underlying problems in the modem, but at the expense of wider
> bandwidth, unacceptable latency, or both. Virtually nothing was
> known about the performance of IFK modes on HF until the
> experimental work using DominoF was carried out in 2004. None of the
> existing modes, or even more recent modes such as Olivia, were
> optimized for noisy low HF bands where NVIS conditions predominate.
The last part of that paragraph is very pertinent -- "noisy low HF
bands _where_NVIS_conditions_predominate_".
If you accept the statement that MF propagation is predominantly NVIS,
you can ring out your favorite modem with PathSim and compare the
error probabilities between selective fades and multipath echos. You
will find that when there is echo, you need much higher SNR to print a
signal than the SNR needed to print selective fading. VE3NEA has
shown that for a couple of Windows modems.
The reason echo is a problem with the ITU NVIS profile is that the
particular profile consists of two paths with a 7 ms delay between
them. That is one third of the 22ms symbol period of steam RTTY.
Selective fading is easy to solve. Most TUs since the ST-6 (perhaps
even earlier, but I could not find any reference to the use of ATC
before the ST-6) had ATC controlled slicers. When one tone fades, the
slicer threshold level moves to keep the decoder unbiased. The only
"difficulty" is to make sure that the ATC distinguishes a true
selective fade case from the flutter case (flutter is not often
accompanied by selective fading).
When there is echo, however, the signal from one tone is affecting not
just the second tone, but also affects itself. Mark-only or Space-
only, or an ATC will therefore not solve the problem. The mechanisms
are completely different -- you need echo cancelers to move the
character error rate down.
My experience when monitoring NMC (which transmits at different bands
simultaneously) has shown to me that the lower bands consistently show
more echo than the higher bands. Which is highly unfortunate, since I
can copy NMC's lower frequency transmissions way better than I can
copy their higher frequency transmissions.
When NMC transmits black and white charts, the signal looks pretty
much like a wide shift FSK signal, so the Coast Guard stations are
especially good for observing daily variations of echo levels for
different frequencies as related to RTTY and other digital modes.
Bill, if you have done SNR studies that show that you do not need
better SNR on the lower bands to print an RTTY signal when compared
with the higher bands, or that the Watterson propagation model (used
by the ITU 4187 and CCIR 520 standards) does not apply to lower bands,
I am sure all of us would love to see it.
Superimposing the data points that you have on 160m on top of VE3NEA's
charts would be very useful to understand what the predominant
propagation effect is (for example, do the data that you have take the
trajectory of selective fading or do they look more like the echo
curves).
As I have mentioned earlier, I welcome RTTY activity on 160m so I can
try to record lots of signal to create a useful Watterson propagation
profile. Only a contest will bring up enough stations to make a study
useful.
73
Chen, W7AY
_______________________________________________
RTTY mailing list
RTTY@contesting.com
http://lists.contesting.com/mailman/listinfo/rtty
|