NavList:

   

Reply

George:

Thanks for the corrections.  I just copied "David Thompson" from a
previous post, didn't check that I had the correct name for Lord Kelvin.

More important, a few seconds after I sent my earlier post I realized
that we had the capacitor charging problem on the cable and it would
delay the receipt of pulses -- not by hours or even minutes but, as you
point out, by some fraction of a second.

For the non-electrical engineers in the crowd, the cable can be thought
of as one of those garden fountains that consist of a series of cups or
pools one above another.  Water flows into the top cup; when it fills,
the overflow flows into the next cup down which in turn fills and
overflows, and so forth.  Think of the electrical pulse as turning on
the water to an empty fountain.  (Oh, and to make the analogy complete,
the cups leak, so the fountain runs dry between "pulses")

Damn, there are some very smart people on this list!!

Lu Abel

George Huxtable wrote:
> I'll have a go at correcting some recent misleading statements about
> telegraphic cables, made under the heading "Time Sights".
>
> Alex said-
>
>
>>I can suggest another reason why time transmission though
>>a transatlantic (or other very long) cable could not be acceptable.
>>In those XIX century cables,
>>the signals were substantially spread in time
>>when transmitted.
>>For example, in the very first transatlantic cable,
>>a short message of few words had to be transmitted
>>for several hours. A sharp impuls you send from one end
>>arrived as a very long wave.
>>So reliable transmission of a time signal could be
>>impossible.
>
>
> It's true, as Alex said, that "a sharp impulse you send from one end
> arrived as a very long wave". But the time-scales that he refers to are
> quite misleading. The first cable, in 1858, only lasted a few days before
> its performance deteriorated so as to make in quite unusable. During that
> period, the receipt of messages was bedevilled by an increasing level of
> intermittent faults, presumably the result of the gradual ingress of
> seawater. With the crude high-voltage source employed at one end and the
> crude signal-detector employed at the other, even the simplest messages
> degenerated into long sequences of "please repeat", "what?" "please send
> slower". The detected signals were just too hard to disentangle from the
> noise. That's how Alex's "several hours" accumulated, just as Lu Abel has
> explained. In 1866 two successful cables were installed, within a few
> weeks, and without those attendant problems.
>
> I don't know how long the interval was between the pressing of the "send"
> key and the detection of the signal at the other end. It depends on the
> voltage and drive-impedance of the source and the sensitivity of the
> detector, and the noise level, but a delay of a few seconds seems to me
> most likely. Such a delay would be amenable to time-correction using the
> method that Paul Hirose has described.
>
> Lu continued-
>
> ""Spreading" of electrical signals takes place when a medium (such as a
> cable) can not transmit the high frequency portion of a signal.  An
> on-off pulse, such as a telegraphic character, theoretically includes
> signal components of very high (indeed, infinite) frequencies.   But the
> speed of transmission is still the speed of light.  Thus a nice sharp
> pulse transmitted on one end of the cable might arrive as a smeared-out,
> possibly unrecognizable mess -- but it would travel across the Atlantic
> (and therefore arrive) in a fraction of a second."
>
> Well, it's true that a coax cable, used in the manner they usually are
> today, transmits signals at the velocity of light. Not the velocity of
> light in vacuum, of 3 x 10-to-the-8 metres per second, but a bit less,
> relating to the speed of electromagnetic waves in its insulation material.
> In polyethylene, as an example, its roughly 2 x 10-to-the-8. (It's similar
> to the way light is slowed when it passes through a medium such as glass or
> water). But such usage requires that the cable is driven at one end by its
> "characteristic impedance" and is connected to a detector at the other end
> with that same impedance. For most coax cables that impedance is of the
> order of a few tens of ohms, such as TV downlead, which is usually 50-ohm
> coax. That characteristic impedance doesn't depend on the length, or
> overall resistance, of the cable. If such a cable could be "terminated" in
> that way, then a transatlantic cable of say 4000 km would have a signal
> delay of about 20 milliseconds.
>
> But it just wasn't possible for a transatlantic cable to be operated in
> that manner. Instead, the rise of the signal at the receiving end was
> dependent on the gradual charging of the total capacitance of the whole
> length of the cable, through the total resistance of the whole length of
> the cable. That "CR time constant" is what dominated the transmission times
> and the sending rates, and increased considerably with length of cable.
>
> Finally, Lu states-
>
> "David Thompson was the chief engineer (or scientist) on the second
> cable. "
>
> No! It was William Thomson, later Lord Kelvin, about the greatest PRACTICAL
> scientist there's ever been, the inventor of the Kelvin (Admiralty pattern)
> compass and a host of other marine devices, and the first man to properly
> understand how an iron vessel's compass should be corrected. In fact, he
> was scientist aboard the first, unsuccessful, attempts in 1857 and 58, as
> well, but on those occasions he had to implement another man's faulty
> designs.
>
> My information on cable matters comes from "Voice across the Sea", by
> Arthur C Clarke, a well-known name now, but perhaps not when he wrote the
> first edition in 1958.
>
> Fred Hebard asked-
>
> "Is there any mention of this in the older texts, such as Chauvenet,
> where time sights were done at geographic markers set by a professional
> astronomer?"
>
> I think the American Coastal Survey, under Gould, had these telegraphic /
> astronomic techniques rather well sewn up long before the transatlantic
> telegraph arrived, and well ahead of the rest of the world. Chauvenet
> refers to Gould's work in 1853 to establish the time-difference (and
> therefore longitude-difference) between Seaton Station (Washington) and
> Raleigh Station in North Carolina, using star transits, as being 6 min
> 32.99 sec, with an error of ? 0.02 seconds, which I estimate to be to a
> precision of about 10 metres! Pretty good going, in 1853...
>
> George.
>
> ================================================================
> contact George Huxtable by email at george@huxtable.u-net.com, by phone at
> 01865 820222 (from outside UK, +44 1865 820222), or by mail at 1 Sandy
> Lane, Southmoor, Abingdon, Oxon OX13 5HX, UK.
> ================================================================
>
>

   

Reply