Sydney-Melbourne Co-axial Cable

The Sydney-Melbourne Co-axial Cable

When you drive along the Hume Highway between Sydney and Melbourne, or along the Federal Highway between Goulburn and Canberra, if you pay attention to the scenery, every few kilometres you might just notice a non-descript concrete structure sitting in a paddock just off the highway, looking like nothing quite so much as a misplaced bunker.  They are easy to miss if you aren’t looking for them, but in fact they are one of the few remaining signs of a remarkable achievement in Australian telecommunications engineering – the Sydney-Melbourne Co-axial Cable. 

The co-ax repeater hut at Wollogorang, south of Goulburn.

In the words of the editorial in the February 1962 edition of The Telecommunications Journal of Australia:

It is difficult to convey the vastness of the project to those who were not associated directly with the work.  Telecommunications engineering works are not spectacular in nature, and now that the Sydney-Melbourne cable has been laid, there is little visible evidence that the work was undertaken.  Even at the height of the cable construction, the effort extended over such a long length of the route that its magnitude was not obvious.

Nevertheless it involved the laying of 600 miles of cable across difficult terrain…  Some one million tons of rock and soil were excavated for the trench in which the cable was laid, and this represents a major civil engineering effort equivalent to the excavation required in the construction of a large dam…. It was, both in size and complexity, by far the largest work of this nature ever undertaken in Australia.  [TJA, February 1962, p166]

Co-ax repeater hut, just north of Goulburn.

In the various commemorations at the time of the fiftieth anniversary of the completion of the cable, it was common to see it described as the “NBN of its day”.  The comparison fails in a number of ways.  In terms of cost, the Sydney-Melbourne co-ax cable cost a mere £6.89m – equivalent to approximately $175m in 2015 terms.  The project was completed and the first telephone circuits across the 1000kms of cable were brought into service within 33 months of the signing of the contracts for the work.  But probably what people are trying to say when they throw around comparisons to the NBN is that it was a momentous milestone in the nation’s telecommunications evolution, requiring vision, a certain degree of audacity, and large measures of plain hard work and disciplined implementation.

Co-ax repeater hut, just south of Goulburn.

In this blog I am going to try provide an overview of the Sydney-Melbourne co-ax cable, the issues it was designed to address and the project to build it, accompanied by historical images from its construction and present-day photos of some of the remnants.

Why build a new Sydney-Melbourne cable?

At a superficial level, it may be self-evident that as demand grows, telecommunications systems need to be upgraded and enlarged, but it is worthwhile to dig a little further into the background to the circumstances that led to the decision to build the Sydney-Melbourne Co-ax.

Detail from a map inside Wollogorang repeater station

The first telephone circuit between Sydney and Melbourne had been opened on 14 June 1907, running over an open-wire system (ie traditional aerial wires strung from poles).  With developments in technology, this route had been upgraded with the addition of ten 12-channel systems – apparently offering the ability to run upto 120 simultaneous calls.   After the end of the Second World War, a large volume of suppressed commercial demand was released, which was in turn reflected in greater demand for telephony.  Accordingly, a further open-wire route was commissioned, adding a further set of ten 12-channel systems.  By 1947, the route was carrying 48 Sydney-Melbourne circuits, 12 Sydney-Canberra circuits and 22 Melbourne-Canberra circuits, in addition to other trunk circuits to other state capitals.

Detail from a map inside Wollogorang repeater station

Despite this investment and construction, demand for telephony continued to outstrip capacity, and each channel was so valuable that it had to be operated to secure absolute maximum usage.  At least three operators were involved in the handling of each call.  In the first place, there would be a “booking telephonist” who recorded the request for the call, in what was called a “reverted” call system, meaning that you placed your request for a line and the operators would “revert” to you when there was capacity for the call to be made.  When the time came for the call, an operator at each end would contact the parties and bring them to the telephone to wait several minutes for the next available circuit.  At times, it was not uncommon to have to wait 2 to 3 hours between requesting a call and being able to actually make it.  In that environment, things like the transmission of television broadcasts were out of the question.  Significantly more capacity was required.

Detail from a map inside Wollogorang repeater station

So why co-axial?

Which begs the question for most people – what is co-axial cable?  A single co-axial cable has an inner conductor or wire, surrounded by an insulator, which in turn is surrounded by a further tubular conductor that shields the cable.  When a signal is carried on the main inner conductor, the shielding prevents leakage of the signal to neighbouring objects (such as other nearby cables), but also protects the signal from external interference, allowing more faithful transmission of high-frequency signals over a distance.  The “co-axial” refers to the fact that the inner wire and the outer circular shield have the same central axis.

Detail from a log book inside Wollogorang repeater station.
The handwritten title is "Repeater Log Book
Wollogorang GC2".  The log book was empty.
Any pages that may have had writing have long since
been eaten by silverfish.

The range of options available to the Post Office when choosing the technology for the new trunk links between Sydney and Melbourne included radio, co-axial cable, “quad carrier” cable, open-wire systems, or a hybrid combination of these.

Of these options, co-axial cable and microwave systems offered the ability to relay television programmes.  Open wire systems, basically continuing to add to the existing technology, had limited capacity and high maintenance costs, being exposed to the elements.

Diagram showing the structure of the co-ax cable.
From J.F.Sinatt, "Design of the Cable Plant"
Telecommunications Journal of Australia
Volume 13, No. 3
February 1962

A microwave system would have involved a network of hill-top radio stations along the route, each within line-of-sight of the next.  This would have been feasible enough, but in order to deliver service to the towns along the route it would need to be combined with additional radio networks, cables or open-wire systems, resulting in significant complexity and much higher costs.

Ultimately, a co-axial cable system offered the right combination of functionality, scope for further development, capital cost and ongoing maintenance costs.

The cable that was laid actually comprised six separate co-axial tubes, with each inner conductor having a diameter of 2.6mm, and an inner diameter for the outer conductor of 9.5mm (the metric measurements in Imperial-system Australia attributable to the German manufacture of the cable – more on that below).  In addition to the six co-axial tubes, inside the cable were an additional sixteen “quad carrier” wires, providing 32 wire-pairs for control and supervising circuits, gas-pressure alarms and short-haul carrier systems.  The granite memorial plaques that were laid at Casula and Melbourne to commemorate the completion of the cable record the total diameter of the sheath as "2 inches", but that seems likely to be just a rough approximation.

Stationery supplies inside Wollogorang repeater station.
Each station was given a short-hand designation, to avoid
the need for long and unfamiliar place-names.
The designation referred to the order of the station,
running from Sydney to Melbourne, and the first initials
of the main repeater sites at each end of the relevant
section.  Hence "GC2" is the second repeater hut in the
section between Goulburn and Canberra.

Repeater stations

There were a total of 118 stations housing transmission equipment along the route from Sydney to Melbourne.  Of these 103 are unattended minor stations (the “misplaced bunkers” mentioned in the introduction), there was a terminal station at each of Sydney South Exchange and Melbourne West Exchange, and then thirteen manned main-repeater stations located at Campbelltown, Bowral, Goulburn, Canberra, Yass, Gundagai, Wagga, Culcairn, Albury, Wangaratta, Benalla, Euroa and Seymour.

Plan of an unattended repeater station.
From J.V Dunn and M. Fizelle
"Buildings"
Telecommunications Journal of Australia
Volume 13, No. 3
February 1962

It appears that a known pilot frequency of 4,092kHz was transmitted on one channel of the cable, and then equipment at each repeater station compared the pilot signal to the known frequency, and any corrections and amplifications were applied to the signals before they were re-transmitted to the next station.

Detail from the amplification equipment
Wollogorang repeater station.

Instruction chart. Wollogorang repeater station.

Instruction chart, Wollogorang repeater station.

Manufacturer's logo on repeater equipment
Wollogorang repeater station.
"FGF"stands for Felten and Guilleaume
Fenmeldeanlagen, the West German manufacturer.

Repeater equipment, Wollogorang repeater station.

Detail from the repeater equipment
Wollogorang repeater station.
Note the "temperature equalizer"function,
presumably designed to calibrate the repeater
equipment for the typical operating temperatures.
"S.-M" stands for "Sydney-Melbourne", indicating
that this equipment was working on the Sydney-Melbourne
co-ax tube.

Detail from the repeater equipment
Wollogorang repeater station.
"M.-S."stands for "Melbourne-Sydney", indicating that
this equipment was working on the Melbourne-Sydney
co-ax tube.  Note the reference under the meter to "4092kc/s".
It appears that the meter displayed the signal-strength for the
reference signal of 4092kHz.

Detail from the repeater equipment
Wollogorang repeater station.

Detail from the repeater equipment
Wollogorang repeater station

Detail from the repeater equipment, showing the logo
of the manufacturer, Felten and Guilleaume.
The logo is an "FG"superimposed over the top of a trident.
The word at the bottom is "Neptun".

Temperature affected the signal propagation and the performance of the repeater equipment, so the repeater stations were designed to minimise temperature fluctuations.  Walls and the roof of the repeater huts were double-layered, with an air-gap between the layers for insulation, and entry was through a short lobby with two heavy steel-clad doors, one at 
either end, to minimise air-flow during staff entry.  With external heat the main concern, the two doors were specifically designed with a ventilator grille at the top of the outside door, but at the bottom of the inside door.  The huts are all oriented to face south or south of east, to minimise sunlight entering through the open doors.

The repeater huts were all pre-fabricated from concrete, with outside dimensions of 14 feet by 12 feet (approximately 4.3m by 3.7m), and an internal ceiling height of 8 foot (about 2.4m).  In areas subject to flooding, the huts were constructed with raised floors and external stairs.  

Today, the remaining repeater stations are in a range of different conditions.  In more built-up areas, some of the sites have been sold-off and the repeater huts demolished.  Some huts have been stripped and reused to house equipment for subsequent fibre-optic cables that serve the same Sydney-Melbourne route.  In rural areas, some of the huts have been left unsecured and have been used by humans to camp in, livestock to shelter in, or sometimes just ransacked.  Even where huts have been left secured, as with any unattended shelter in rural areas, they can be colonised by rats who nest in amongst the equipment and scatter droppings and nesting material.

Route

A preliminary route for the cable had been determined in 1950, and this preliminary route had been used for issuing the request for tenders to potential suppliers.  Along this preliminary route, sites for the 13 main repeater stations were selected and locked-in, largely based on issues such as the location of appropriate service centres, and population centres that were intended to be served by the cable.

Map showing the route of the Sydney-Melbourne Co-ax
From A.H. Kaye, "Main Features of the Project"
Telecommunications Journal of Australia
Volume 13, No.3
February 1962

That still left a requirement to determine the route between these main repeaters.  This was done by progressively working down from a number of options chosen from detailed maps and aerial photos, and then ultimately walking the routes using compass bearings, to select the most suitable route.  Generally speaking, the factors influencing route section were a preference for the shortest route, combined with considerations of access, practicability, and security from hazards such as floods, landslips and erosion.

From there, an engineer made the detailed route selection for a preliminary pegging party which placed pegs at thousand-foot intervals and at angles in the route.  This enabled the preparation of basic survey plans, allowing repeater sites to be selected so that the process could be started to acquire the sites.  Once the sites had been selected, a detailed final survey could be conducted, with pegging at 100-yard intervals.   The cable was laid from the Sydney end, and weather issues on the New South Wales section gave rise to a need for non-sequential laying – skipping sections that were water-logged or even flooded, and then returning to the difficult sections later when they had dried-out.  This meant that when it came time for detailed surveying and pegging of the Victorian sections, this was done in much greater detail, to allow identification of joints so that cable-lengths could be accurately specified for sections that were being laid out-of-sequence.

A concrete marker post above the route of the
Sydney-Melbourne Co-ax, in a paddock just south of
Goulburn.

Once the cable was laid, the route was marked with seven-foot concrete posts.  Each one has a metal plate on the top indicating the identity of the cable, the depth and the direction of the cable-run on either side of the post.  From each marker-post it should be possible to see the next post or the repeater station.

Plate on a concrete marker post above the route of the
Sydney-Melbourne Co-ax.
It seems unlikely that the "Nearest Postmaster" would
be able or interested to provide too much
"further information".

Sighting from the top of one concrete marker post
to the next in the series.  Note that the line of the
cable is still visible on the hillside behind, more than fifty years after the cable was laid.  Presumably this is
a result of soil compaction from the trenching.

The plate on the top of a concrete marker post.  The marker shows that the cable is six foot to this side of the marker, at a depth of 7'9" below the top of the marker, and that the marker is placed above a joint.  In the top left you can see the names of the cables in this location.  "T1001" is the Sydney-Melbourne co-ax.  It is unclear whether T328 is a separate cable, or the designation of one of the subsidiary systems inside the Sydney-Melbourne co-ax sheath.

Cable laying

Once the detailed cable route had been determined, the cable laying process involved clearing any trees for a distance on either side of the route, ripping a trench along the line of the cable, drilling and blasting any rock material that couldn't be ripped, removing the ripped or blasted rock from the trench, laying the cable on top of an underbedding layer, covering the cable with an overbedding layer, timbering the overbedding, and then completing the backfilling.  In areas where erosion was a concern, earth banks were used to direct run-off away from the cable route.

Some 17 rivers and about 120 creeks and gullies were crossed, with the cable being laid under the bed in most cases.  On three major river crossings, such as the Goulburn River at Seymour, the cable was installed in conduit on road or railway bridges, but generally above-ground installation was avoided.

The initial plan for laying the cable was that ripping would proceed significantly in advance of the cable route, ripping the cable trench to the required depth of 4 feet.  It was intended that this would enable early identification of any significant rocks in the cable path that might require power-tools or explosives to clear. This approach was taken for the first section from Sydney to Goulburn, until it was discovered that with either rain or soft soil, this sort of pre-ripping led to collapse of the soil for a significant distance on either side of the trench, with the result that the cable-laying or trenching machinery would become bogged, and occasionally the cable route had to be moved to one side to find more stable soil.  South of Goulburn a different approach was taken, with pre-ripping only being done where it was known that it would provide assistance, and then only immediately in advance of the trenching.

A box of spare fuses from inside Wollogorang repeater station.

A marker-post and pit in a paddock north of Goulburn.

A pit in a paddock north of Goulburn.  It appears that this is one of the pits containing
sub-repeaters,  which were added subsequent to the original construction to upgrde the capacity from the original 4.5MHz / 960 channels to  6MHz / 1260 channels.    The pit also housed a compressed-air cannister
for pressurising the cable.  At the time of visiting the pit it was full of water.

Old PMG-branded Eveready batteries, for powering the
handset on the magneto phone to allow staff visiting the repeater
hut to communicate with a main repeater station.

Old magneto phone in the lobby of an unmanned repeater station.
The batteries to power the handset were held in the metal box
underneath the phone.  The magneto was cranked to generate a signal
voltage to attract the attention of an operator back at the exchange.

A "Commonwealth of Australia" branded dustpan and broom,
inside the lobby of an unmanned repeater station.

Suppliers

For the Sydney to Canberra portion of the project, the cable itself was manufactured by Felten and Guilleaume Carlswerk A.G of West Germany.  For the portion of the project south of Canberra, the cable was manufactured in Melbourne by Olympic Cables Pty Ltd as sub-contractor to Felten and Guileaume.  Presumably this arrangement allowed the project to commence without delay, while Olympic Cables equipped their facilities in Melbourne for production.

The transmission equipment was manufactured primarily by Felten and Guilleaume Fernmeldeanlagen G.m.b.H. in West Germany.

Cable Pressurisation

Gas pressure gauge inside Wollogorang repeater station.
The electrical contacts on the gauge suggest that it was remotely
alarmed.  This would have been the gauge used to determine when
the compressed-air cannister was running out.

The Sydney-Melbourne co-ax cable is pressurised using a continuous flow system.  This comprises compressed-air canisters housed at repeater stations, and connected to the co-ax tubes through regulators to maintain a constant pressure.  The purpose of the pressurisation is two-fold.  Firstly, by monitoring the pressure both at the repeater stations and inside the cable, it is possible to detect damage to the cable-sheath, allowing remedial action to be taken before a fault develops.  Secondly, the positive pressure within the cable means that, in the event of a breach of the cable sheath, water ingress is less likely.

Originally, the cable was designed to have mercury U-tube manometers installed at 1000-yard intervals, and connected to electrical contacts that would register the air-pressure falling below a pre-determined level.  These mercury manometers were installed for roughly the first 350 miles of the cable from Sydney before testing was able to be undertaken on the first section.  This testing revealed that six out of the ten tested manometers failed.  Investigations revealed that the failures were due, at least in part, to impurities in the mercury being used, resulting in significant innaccuracy in operation.

The mercury manometers were subsequently replaced with mechanical bellows contactors, which made a connection between two wires within the co-ax sheath when the air pressure dropped below the pre-determined level.  The bellows contactors were connected with different resistors to assist in identification of which contactor had been triggered, to assist in locating the approximate location of cable-breaks.  At the repeater stations the air-pressurisation equipment gave the ability to inject tracer gas into the cable to assist in identifying the precise location of the breakage.  It appears that the tracer gas would be either radon or freon.  Apparently the freon would be coloured with a dye for visual identification of the cable-break.  Presumably radon was used in association with a geiger counter.

Thermometer inside Wollogorang repeater station.
This appears to be a later ad hoc addition.  It is unclear whether
its purpose is connected with the air pressurisation or the operation
of the repeater equipment.

Air pressurisation equipment inside Wollogorang repeater station.
A pair of compressed-air cannisters would have been strapped to
the wall using the leather straps at the bottom.

Detail from the air-pressurisation equipment inside
Wollogorang repeater station.  The numbered tubes at the
top appear to go to the six co-ax tubes.  The golden-coloured
tube at the right is the air-drying device - presumably to remove
moisture from the compressed air being fed into the co-ax tubes.

Pressure gauge on the air-pressurisation equipment inside
Wollogorang repeater station.  This would allow regulation of
the air-pressure being fed into the co-ax tubes, to assist in the
location of any cable-breaks.

Air-pressurisation equipment inside Wollogorang repeater station.
The structures on the top right and left will be feeding compressed air
the cable in the two different directions (towards Sydney and towards Canberra).
There is a vacant inlet at the bottom-right of the green stop-cocks,
This would have been used for injecting a gas with dye added, to
assist in locating a cable-break.  When the dye was injected, a linesman
inspecting the line of the cable would see the dye emerging from the ground
at the location of the break.

The air-pressurisation cabinet with the door closed.
Wollogorang repeater station.

Labelling on the air-pressurisation equipment, Wollogorang repeater station.

Customer impact

Once the cable was in full operation, it was possible for the first time to transmit a live television broadcast between Sydney and Melbourne.  The additional voice capacity delivered not only better quality voice calls, but ultimately the ability to implement subscriber trunk dialling (STD) between the two state capitals, as there was now enough capacity available to meet demand on-demand, rather than requiring customers to book in advance.  Finally, many population centres along the cable route were now able to receive significantly increased capacity and automation of their previously-manual exchanges.





References

1. AH Kaye, "Main Features of the Project", The Telecommunications Journal of Australia, Vol 13 No 3, February 1962.
2. IS McDuffie, "The Telecommunications Aspects", The Telecommunications Journal of Australia, Vol 13 No 3, February 1962.
3. JF Sinnatt, "Design of the Cable Plant", The Telecommunications Journal of Australia, Vol 13 No 3, February 1962.
4. DF Barrie and CH Hosking, "Installation of the Cable", The Telecommunications Journal of Australia, Vol 13 No 3, February 1962.
5. FJ Harding, "The Gas Pressure Alarm System", The Telecommunications Journal of Australia, Vol 13 No 3, February 1962.
6. AL Fisher, RJ Clark, RA Colins, "Transport of Cable and Other Materials", The Telecommunications Journal of Australia, Vol 13 No 3, February 1962.
7. JV Dunn and M Fizelle, "Buildings", The Telecommunications Journal of Australia, Vol 13 No 3, February 1962.
8. JR Walklate, "Design of Transmission Equipment", The Telecommunications Journal of Australia, Vol 13 No 3, February 1962.
9. JW Pollard, "The Application of Gas Pressure Alarm Systems to Coaxial Cables", The Telecommunications Journal of Australia, Vol 13 No 1, June 1961.

Acknowledgments

Many thanks to the following people for generously giving their time, knowledge, resources and wisdom:

- Janko Radocaj
- Stef Nowak
- David Piltz
- Phill Sporton

And thanks and acknowledgment as always to Brian Mullins and the team of dedicated volunteers at the Telstra Museum Bankstown.


A similar photo-blog on the old Port Kembla Exchange can be found here:   http://portkemblaexchange.blogspot.com/