RadioBanter - View Single Post - Echos from the past, code a hinderence to a ticket

August 9th 05, 02:42 PM

wrote:
Mike Coslo wrote:

wrote:

Mike Coslo wrote:

wrote:

Mike Coslo wrote:

an old friend wrote:

Dave Heil wrote:

Mike Coslo wrote:

Dave Heil wrote:

Of
course the patience of Job will be needed
for all the error correction needed

Slow...error prone...not used by other services...

Where'd I hear that before? ;^)

HF will never be the place for high speed
digital transmission. There
is too much noise and signals are subject
to the vagaries of wave
propagation phenomena.

A lot depends on what you mean by "high speed", and what
resources you can use.

If you're used to 45 baud RTTY, 1200 baud is high speed,
isn't it?

For example, if you're allowed to use wide-enough bandwidths,
all sorts of stuff is possible.

Sure. I think some people are believing that I am saying this is
impossible. It is not impossible. But not very practical.
I'd have to
work it out, but I think there are some ham segments where
we'd need more bandwidth than is alloted

The difference is that unlike, say, the US Army in 1952,
there is no
High Command that determines who gets what frequencies
for what path at what time.

If you're allowed to use very high power and high gain
antennas, all sorts of stuff is possible.

A strong signal mode? 8^)

Sure - that's predictable by Shannon's. It's not hard to
understand and explain, if you really understand what
Shannon is all about. (Some folks make a lot of noise
about it, acting as if it's some big deal, but it's really
quite easy to understand.)

Consider this situation:

Plain old on-off keying lets you send one unit of information
per unit time. Either a 1 or a 0, on or off. Digital and
all that.

But suppose we have a system (transmitter, receiver and connection
between them)that has four states - 100% on,
66% on, 33% on, and off (0% on). We can then apply a meaning to those
four states - say, 11, 10, 01, and 00, respectively - and send two
units of information per unit time in the *same* bandwidth.

We've just doubled the data rate without changing the modulation
method, bandwidth, or basic keying rate. But for it to work,
the system must have good enough signal-to-noise ratio to
determine the four states reliably.

That principle can be continued as far as our signal-to-noise ratio
allows. For example, a system of 128 states could be done,
if the system S/N is good enough, allowing the transmission of
seven units of information per unit time in the same bandwidth as
was previously used for one unit.

IIRC we had this conversation before, only it was about a PSK system
with many more states, rather than an AMK system.

And there's no reason multiple carriers can't be used, as well as
multiple ways of modulating the same carrier - say phase and amplitude
modulation at the same time.

Of course the entire system has to have adequate s/n to deal with
the tiny variations of phase and amplitude.

I can see the needed s/n ratio going up with each addition.

Telephone line modems make use of the predictable stability and
characteristics of the telephone lines. HF radio is a bit less
stable.

And yet even here, I can remember paying extra for a modem line, which
was quieter than a regular phone line. I don't know if there are such
things any more - its been a long time since I've used dial-up.
Otherwise you slow to a crawl, as error checking struggles to keep up
with line noise s/n.

Of course if the taxpayers are footing the bill, all sorts of
things can be done.

Sure! Look at what the military folks did. Big rhombic and
curtain
arrays, etc. Find a suitable site, take it over as necessary for the
military purpose, put up whatever is needed - on the
taxpayer's
dollars. If you need more power, just get it! Receivers like the
R-390, costing thousands of 1950s dollars? How many racks of them are needed? Etc.

Completely different from what most hams deal with.

If you can separate the transmit and receive sites and/or frequencies
so that full duplex is achieved, all sorts of
things are possible.

OY!

Very common military and commercial practice.

If your setup has adaptive features so that it evaluates
the path characteristics and adapts the modulation and
frequencies used to conditions, all sorts of things are
possible.

Doubly Oy!

It's what ALE is all about.

Just not too applicable for our purposes.

My Elecraft K2 has programmable memories that can be set for each band
and mode. Other rigs have similar features.

It would not be difficult to have it step through them (using the
RS-232 port and computer control) comparing each frequency with the
others. With suitable time synchronization between, say, you and I, the
rigs could step through the various preprogrammed QRGs, looking for the
spot with the best propagation and no QRM.

Of course we'd have to set it all up beforehand so we'd - or rather the
rigs - would know precisely where to send and listen.

The K2's ATU also remembers antenna tuner settings per memory.

Of course most of the above is simply not practical for
the average ham, and/or is incompatible with current US
regulations.

really? gee it falls into the same catagory as
when it was said that we
ham had been banished to "useless frequencies"
everything above 200M

Hams were never banished to everything above 200 meters.

What happened was that amateur stations were required to use
only wavelengths of 200 meters and below. But every station had
a specified wavelength. If a ham wanted to use, say, 159 meters,
s/he needed a station license that said "159 meters".

In 1912 there wasn't much known about how HF propagation actually
worked. The 'useless' idea came from extrapolation of what happened on
longer wavelengths. The ionosphere's role was not
even guessed at by "professionals in radio".

There are some pretty darn good reasons why high-speed digital
HF won't
work well. And they aren't related to early "knowledge" that
caused hams
to be relegated to those higher frequencies at the time.

Radio is a fairly mature field, and digital is getting there.
Many
people have a pretty good idea what will likely work, based on education
and experience. And HF is an unruly beast, given to noisy and incredibly
variable conditions. We don't have to be rocket scientists to gain that
knowledge.

Just as an exercise, how much information can be carried by a 1.8 MHz
signal?

As much as the S/N allows! See above.

Of course! As a practical matter though, we have to assume amateur
radio conditions. A real limitation there.

One of the things that I wonder about with the need for huge s/n
handling ratios is that we obviously want a quiet reciever, with a
impressive noise floor. This means all of the "oomph" must be on the
strong signal handling side. We need an exquisitely sensitive and quiet
reciever, with exceptional strong signal handling capacity.

How much error correction will be needed during the summer, and

how much during the winter? Why is there a difference? Why would a
wireless digital transmission system use UHF and above for data
transmission?

All very good questions!

when was that Jim

A long, long time ago. When almost nothing was known about propagation.

Jim might note that they do some bandwidth tricks in similar manner
as he proposed per our conversation in here earlier. Not exact, but
along the same lines
Hopefully we will see an article from those who know the right way
to HF digital soon. 8^)

I don't have my CQ handy, but it took them a fair amount of time
(measured in minutes IIRC, to transmit some heavily compressed (beyond
maximum jpeg compression), and therefore really poor quality (by almost
everyones standard) pictures.

Didn't I make a challenge with some of the HF high-speed digital
believers in here to do a sked? I think the "answer" was that I was
going to steal the technology.

You can't steal vaporware.

Not that that is likely, but how about say some of the
believers among
themselves, do a proof of performance of the technology?

Or is this just one of those Wondertenna type ideas that crop
up from
time to time, only to be found lacking when introduced
into the real world?

Heck, Mike, you want *practical* stuff?

You betchya!

I have to note that we speak of actual systems, real world devices -
some of which are yet to be made of course, but practicalities (granted
you much more than myself - but I am an RF neophyte, having come from
the digital world)

Others seem to be more in the vein of "so there! or someone is going to
come along and prove you naysayers wrong!" mode.

Bingo.

Faith based electronics.

Yes - those who dare to even like modes such as Morse Code
are cursed as infidels who do not understand the Word that
"newer is always better" and "the PROFESSIONALS know best"

If newer is always better, I wonder why these vanguards of the brave
new world aren't typing to us stodgy mortals in leet? After all, it's
newer.

One of the biggest flaws of many people is that they have no
appreciation of mature technology and people. They choose to concentrate
on the limitations and foibles of both, not realizing that they are
themselves full of their own foibles and limitations. They create for
themselves great effort and even harm in their rush to discard the old,
simply to reinvent it.

I wonder how that tremendous antenna from the UofD is coming along? You
know, the one that is going to revolutionize radio? HF antennas a few
feet long that outperform anything we have today.
The one that was so "efficient" that it melted when the inventor
powered it with 100 watts. And I'm not the one who used the word
efficient, *they* did.

It *is* efficient, Mike! It's very efficient at turning HF RF energy
into heat.

I've seen better! ;^)

Don't expect it from "John Smith", Len, "b.b.", or even "an
old friend".
You won't get it.

And I'm starting to think that some of them might be
duplicates anyhow.

Quitefine and Darkguard (nee Blackguard) indeed!

Doesn't matter - they have nothing practical to offer.

They're not about actually *doing* ham radio,
just arguing about it.

Too bad they cant do a better job.

Indeed

Quite! - Mike KB3EIA -