The Naked Scientists

The Naked Scientists Forum

Author Topic: Is a repeaterless transAtlantic speech cable feasible?  (Read 5867 times)

syhprum

  • Hero Member
  • *****
  • Posts: 3332
    • View Profile
It is well known that such a cable was proposed in 1928 but never got of the ground, I am sure I have seen proposals for a 1939 cable which could have used Vocoder technology and almost certainly worked.
What could we do today with modern data compression, computerised pre distortion, and modern cable dielectrics or even fibre optics.
« Last Edit: 29/12/2011 03:25:01 by chris »

Geezer

  • Neilep Level Member
  • ******
  • Posts: 8416
  • "Vive la résistance!"
    • View Profile
repeater less trans atlantic speech cable
« Reply #1 on: 14/12/2011 20:25:51 »
Are we allowed to use parallel conductors?

CliffordK

  • Neilep Level Member
  • ******
  • Posts: 6409
  • Site Moderator
    • View Profile
repeater less trans atlantic speech cable
« Reply #2 on: 15/12/2011 01:14:42 »
Interesting.

So, the Modern "submarine cables" all have repeaters powered by an electrical circuit from the source of the cable.

Keep in mind that with digital/analog conversion, there really is little difference between voice and data cables now. 

I assume the question on whether or not to use repeaters is based on noise and latency.  One could probably slow down the communication to have more distinct pulses, and perhaps add more redundancy to deal with increased noise. 

But, the benefit would only be if the performance of the slowed communication would exceed that allowed with the repeaters.

I'm seeing distances of about 380 km between repeaters.  One could probably design a system that would do hops across land.  UK-->Iceland-->Greenland-->Canada, but you would undoubtedly need a longer cable.


syhprum

  • Hero Member
  • *****
  • Posts: 3332
    • View Profile
repeater less trans atlantic speech cable
« Reply #3 on: 15/12/2011 06:17:50 »
Such a route was contemplated in the fifties for a VHF TV circuit but was considered too unreliable.
The reason why a repeaterless submarine cable was considered prewar was that underwater repeaters were considered too unreliable but when TAT1 was eventually built they functioned 25 years without a failure

MikeS

  • Hero Member
  • *****
  • Posts: 1044
  • The Devils Advocate
    • View Profile
repeater less trans atlantic speech cable
« Reply #4 on: 15/12/2011 08:16:44 »
As I understand it modern transatlantic cables do not use repeaters.  Instead they use in-line amplifiers that directly amplify and clean up the optical signal without first converting the optical signal into an electrical signal and back into optical as per repeaters.  By coating sections of the fibre with certain rare-earth elements  (erbium, neodymium, praseodymium, or thulium.) these section operate like a laser amplifier cleaning and amplifying the signals passing through the fibre.  This type of in-line amplifier processes all channels passing through the fibre at once and is both cheaper and more reliable than repeaters.  It can also have the advantage of not requiring wires to supply electrical power as is can be powered by light that is distributed through the fibre optic.

see
http://www.rp-photonics.com/fiber_amplifiers.html
« Last Edit: 15/12/2011 09:20:33 by MikeS »

syhprum

  • Hero Member
  • *****
  • Posts: 3332
    • View Profile
repeater less trans atlantic speech cable
« Reply #5 on: 15/12/2011 15:38:57 »
"Figure 1: Schematic setup of a simple erbium-doped fiber amplifier. Two laser diodes (LDs) provide the pump power for the erbium-doped fiber, allowing it to amplify light with wavelength around 1550 nm. Two pig-tailed Faraday isolators strongly reduce the sensitivity of the device to back-reflections".

May I draw your attention to Figure 1 in the link you gave me, I believe about 1 Amp has to be passed thru the cable to energise these laser diodes though thru the erbium-doped amplifying section both the pump light and the signal light flow thru the same section.

MikeS

  • Hero Member
  • *****
  • Posts: 1044
  • The Devils Advocate
    • View Profile
Re: repeater less trans atlantic speech cable
« Reply #6 on: 28/12/2011 10:45:58 »
Quote from the article
"in almost all cases, the pump light propagates through the fiber core together with the signal to be amplified."

This seems to be saying that the lazer diode that operates the pump is powered by light not electricity.  That's how I read it anyway.

syhprum

  • Hero Member
  • *****
  • Posts: 3332
    • View Profile
Re: repeater less trans atlantic speech cable
« Reply #7 on: 28/12/2011 21:58:25 »
This only applies to the local loop where the amplification takes place , the laser pump diodes have to be powered by an electrical current.
look up any standard referances as to how optical amplifiers work

"Fiber Amplifier Modules
 
Some companies offer fiber amplifier modules which can be convenient for OEM system integrators. Input and output are then often attached with the usual fiber connectors. A compact module contains not only the actual fiber amplifier(s), but also the control electronics for the pump diodes, and possibly extras such as an input and/or output power monitor, power stabilization, alarms, gain-flattening filters, etc. Such amplifier modules are available based on erbium-doped fibers, ytterbium-doped fibers, and others, and for various power levels."
« Last Edit: 28/12/2011 22:17:22 by syhprum »

MikeS

  • Hero Member
  • *****
  • Posts: 1044
  • The Devils Advocate
    • View Profile
Re: Is a repeaterless transAtlantic speech cable feasible?
« Reply #8 on: 29/12/2011 07:32:12 »
syphrum

You may well be right.  I was only going by what the article said.  My experience in that field was in the very early days when fibre optic transatlantic cables were no more than a possibility.

If powering the amplifier by light is not yet possible it seems to me theoretically that someday it should be.  Electricity and light are both energy and as such could be used to power an amplifier.  If light powered lazer diodes do not yet exist, no doubt someday they will.  They would simplify cable construction significantly and lead to greater reliability.

The way I read the article I assumed they already existed but as I said you could well be right.
« Last Edit: 29/12/2011 08:04:49 by MikeS »

syhprum

  • Hero Member
  • *****
  • Posts: 3332
    • View Profile
Re: Is a repeaterless transAtlantic speech cable feasible?
« Reply #9 on: 29/12/2011 20:27:37 »
With regards to supplying power to any devices in the cable we have to bear in mind the frequency versus attenuation effect if we pass a direct current thru the cable the frequency is virtually zero so the only power loss is due to the resistance of the conductor and the power that can be drawn of to power the amplifiers is the same at every point in the cable.
if an attempt was made to power the amplifiers with light if sufficient high power could be injected at the start there would be a great excess at the first amplifier but the available power would drop of rapidly as you got towards the centre. 

MikeS

  • Hero Member
  • *****
  • Posts: 1044
  • The Devils Advocate
    • View Profile
Re: Is a repeaterless transAtlantic speech cable feasible?
« Reply #10 on: 30/12/2011 07:34:37 »
I kind of imagine that instead of all amplifiers being in series in the power line they are tapped off in parallel with each one suitably attenuated (much the same as with electrical amplifiers).  I know little of such things but can't see why it should be a major problem.

syhprum

  • Hero Member
  • *****
  • Posts: 3332
    • View Profile
Re: Is a repeaterless transAtlantic speech cable feasible?
« Reply #11 on: 30/12/2011 07:38:08 »
light powered lasers were the first ever made, the original laser consited of a ruby rod with mirrored ends surrounded by a Zenon flash tube such as is used in a camera.

syhprum

  • Hero Member
  • *****
  • Posts: 3332
    • View Profile
Re: Is a repeaterless transAtlantic speech cable feasible?
« Reply #12 on: 30/12/2011 07:43:30 »
No the normal arrangement is for the laser PSU,s all to be in series which ensures the same amount of power is available for each one.
19 centaury street lights operated in the same manner!.

« Last Edit: 31/12/2011 10:37:19 by syhprum »

Geezer

  • Neilep Level Member
  • ******
  • Posts: 8416
  • "Vive la résistance!"
    • View Profile
Re: Is a repeaterless transAtlantic speech cable feasible?
« Reply #13 on: 30/12/2011 08:13:43 »
No the normal arrangment is for the laser PSU,s all to be in series which ensures the same amount of power is available for each one.
19 centuary street lights operated in the same manner!.

If they were all connected in series they would all pass the same current, but I did not know that is how they worked. I thought they were connected in parallel.

syhprum

  • Hero Member
  • *****
  • Posts: 3332
    • View Profile
Re: Is a repeaterless transAtlantic speech cable feasible?
« Reply #14 on: 30/12/2011 11:07:41 »
If a parallel scheme was used it would require a very heavy power cable to avoid the resistive loss of power  whereas with series operation the only problem is the high voltage required 5-10kv for which it is easily possible to provide sufficient insulation.
I suppose a fairly high voltage could be fed into the cable and each laser unit could use a chopped mode PSU but simple Zenner diodes to stabilise the voltage across each series unit makes a more simple arrangement, the insulation of the high voltage between the laser diodes and the doped fibre section of cable where the amplification takes place is provided by a section of fibre optic cable .
« Last Edit: 30/12/2011 11:21:40 by syhprum »

MikeS

  • Hero Member
  • *****
  • Posts: 1044
  • The Devils Advocate
    • View Profile
Re: Is a repeaterless transAtlantic speech cable feasible?
« Reply #15 on: 31/12/2011 08:23:14 »
Clip
19 centaury street lights operated in the same manner!.


But that would mean that if one failed they all failed?

syhprum

  • Hero Member
  • *****
  • Posts: 3332
    • View Profile
Re: Is a repeaterless transAtlantic speech cable feasible?
« Reply #16 on: 31/12/2011 10:35:42 »
The street lights were carbon arc lamps that had to be supplied by a special constant current dc generator, if all the lamps were in parallel each would have required its own current stabilising equipment but by running them all in series one generating set was used for a large number.
Of course there was provision at each lamp to restore the circuit if it failed. 

SeanB

  • Hero Member
  • *****
  • Posts: 1101
    • View Profile
Re: Is a repeaterless transAtlantic speech cable feasible?
« Reply #17 on: 01/01/2012 18:21:59 »
There are still series street lights, though they are rapidly becoming extinct. They are all supplied with a constant current of normally 6.6A, and have a provision to short out faulty lamps when they fail. Simple in that an open cable has a voltage rating that is determined by the insulators, and you can run a very long circuit without worrying that the far end lights are going to be noticeably dimmer, and only need a cable rated for 10A, irrespective of the load. Saves copper in the cable, but needs a single extra transformer at the supply point. They operate on AC, even the carbon arc lamps in the 1900's did so ( hence the current of 6.6A to run them) as this made for even electrode wear and allowed an automatic starter and rod feeder to be built.

As to repeaterless cables, the major limit is the data rate, as if you do not have repeaters to isolate each section then to send signals you needed to charge and discharge the massive capacitance of the cable, whilst being limited by the large inductance of the cable. This limits the data rate, and using higher voltages to speed up the charge and discharge time ( total time is the same, but a high voltage means the far end sees a 1V step sooner if the drive is +- 1000V than if it is +-10V) runs the risk ( as happened on many early cables) of burning out the cable or of shorting it out.

syhprum

  • Hero Member
  • *****
  • Posts: 3332
    • View Profile
Re: Is a repeaterless transAtlantic speech cable feasible?
« Reply #18 on: 01/01/2012 19:00:04 »
with todays technology a repeaterless cable is a ridiculous idea but in 1939 things were diferent, vocoders had just been invented making speech possible in a bandwidth of a few hundred Hertz and polyethelene insulation solved the voltage problem I believe some lengths of cable were produced by Siemens but German technology was (is) never avant gard.
It was an interesting idea that was ovetaken by other developtments (like air ships and ocean liners).
I read too of arc lamps being run on AC but recalled from my boyhood of special DC constant current generators designed to power them .

Geezer

  • Neilep Level Member
  • ******
  • Posts: 8416
  • "Vive la résistance!"
    • View Profile
Re: Is a repeaterless transAtlantic speech cable feasible?
« Reply #19 on: 01/01/2012 20:23:46 »
you needed to charge and discharge the massive capacitance of the cable, whilst being limited by the large inductance of the cable.

Only if you were using very low frequency signals. At moderate frequencies, the energy propagates along the cable.

The big problem is signal attenuation. Because the conductors and dielectric are not perfect, the transmitted energy dissipates with distance.

syhprum

  • Hero Member
  • *****
  • Posts: 3332
    • View Profile
Re: Is a repeaterless transAtlantic speech cable feasible?
« Reply #20 on: 02/01/2012 13:58:07 »
I understand that by 1939 telegraph cables were achieving a bandwidth of 100HZ, and I have just been listening to perfecly intelegable speech at 2.4Kb/sec so I guess the cable was quite possible but I don't know about the economics.
If I could digress to street lights for a moment , When I moved to Maidstone in 1939 there was a small coal fired 20MW DC power station with batteries that supplied the old part of the town and the trolley bus system although the suburbs were on the 50HZ grid.
The sodium discharge street lights were switched on and off by a 400HZ signal -20db superimposed on the 50Hz power.
Have you encounted this system, was it common ?.
« Last Edit: 02/01/2012 14:08:01 by syhprum »

SeanB

  • Hero Member
  • *****
  • Posts: 1101
    • View Profile
Re: Is a repeaterless transAtlantic speech cable feasible?
« Reply #21 on: 02/01/2012 18:10:36 »
Dc mains were mostly a British thing, and were phased out in the early 1960's AFAIK. They had local converters that originally used motor generator sets to provide the 200V DC current, and were arranged so that half of the street was on +200V and the other half was on -200V, so as to reduce the current in the neutral conductor. This survives in the USA with 120V mains, with 2 phase wires and a neutral wire.

Later on the rotary converters were replaced with transformers and mercury rectifiers, as these were lower maintenance ,and did not need daily cleaning and adjustment/oiling. The trolleybus was DC and batteries so it could operate if the mains went out. Many train systems still are DC, but there are no batteries for backup. DC is used so you can have regenerative braking and feed power back into the line for another train to use. The batteries would have done the same.

As to the remote streetlight switching, the superimposed signal method was around for many years, eventually being replaced with photocells as substations became unmanned, and there being no call for staff on duty to adjust voltage, it being done by automatic equipment.

syhprum

  • Hero Member
  • *****
  • Posts: 3332
    • View Profile
Re: Is a repeaterless transAtlantic speech cable feasible?
« Reply #22 on: 02/01/2012 22:16:26 »
When I went to America to install a HELL scanner that ran on three phase 220v delta I was puzzeld by your power system being used to all three phases being the same voltage relative to ground or neutral but I eventualy worked it out.
When I checked out the state of VOCODER development I was surprised how crude and underdeveloped it was I also found a copy of a TV program from 1958 ( which I watched when it was first broadcast ) "the parameters of PAT" and very little progress seemed to have been made, perhaps I was over optamistic about the availability of speech compression in 1939
« Last Edit: 02/01/2012 22:26:12 by syhprum »

Geezer

  • Neilep Level Member
  • ******
  • Posts: 8416
  • "Vive la résistance!"
    • View Profile
Re: Is a repeaterless transAtlantic speech cable feasible?
« Reply #23 on: 02/01/2012 22:55:55 »
When I went to America to install a HELL scanner that ran on three phase 220v delta I was puzzeld by your power system being used to all three phases being the same voltage relative to ground or neutral but I eventualy worked it out.

I had a similar problem when I moved to the US. I was used to 3-phase systems in the UK and I couldn't understand why the phase to phase voltage in my house was exactly twice the phase to neutral voltage.

The penny finally dropped when I moved to the country. There is a three phase transmission line running along the edge of our property with a transformer on a pole to supply our house, and I could see that the transformer was only hooked on to one of the phases.

Doh. It's single phase 220V with a center tap!

SeanB

  • Hero Member
  • *****
  • Posts: 1101
    • View Profile
Re: Is a repeaterless transAtlantic speech cable feasible?
« Reply #24 on: 03/01/2012 18:12:14 »
3 phase power in the USA is funny, where you need 2 separate meters and 2 separate power cables to have both 3 phase and single phase power. Here all you need is a single meter, and then you can load balance to even out the phase currents. Every pole has 3 phase power, you do not need a special connection to the substation and cable to have 3 phase anywhere.

Only exception is rural areas, where you often have a single HV wire pair, allowing you to have a local transformer at the farm. This is because the farmer has to pay for the cost of the pole and the wire from his property boundary, and the cheapest is a single pair of wires on poles. 3 phase needs 4 wires ( one being an earth wire, not a neutral) and the cable is the biggest cost, leading to you finding pump motors of 22kW single phase being common, where most would be 3 phase at that power.

 

SMF 2.0 | SMF © 2011, Simple Machines