Aluminium conductors in data cabling systems may seem like a good idea, but George Georgevits has some test results that prompt another view.

In three decades of running a test and measurements lab I have been asked to examine some odd ideas. Some of them were worthwhile and others could have done with more research. Anyhow, that is what testing is all about – to see if theory matches what happens in the real world.

So, when asked to test some new ‘Cat 6’ data cable of Chinese manufacture that used copper-coated aluminium wire as the conductor, I immediately had some reservations.

(It is worth noting that the Electronic Industries Association (EIA) Standard no longer mandates copper as the conductor material, yet one of its bodies is still called the Subcommittee on Telecommunications Copper Cabling Systems.)

In any case, I like to approach new ideas with an open mind and eagerly went ahead with the required set of transmission tests.

Copper vs aluminium

The use of aluminium as a replacement for copper is not new in electrical work. It is common practice in the power industry.

Many years ago when copper prices first soared, Telstra (or was it the PMG in those days?) trialled aluminium conductor telephone cable, and I think some ended up in service.

My very old copy of the The Cables and Conduits Handbook reveals a listing for APIQL cable – or aluminium conductor, paper insulated quad local. Presumably, if the calculations are right, and handling and jointing procedures are modified to suit, it can be done successfully.

Fundamentally, aluminium as a conductor has several advantages over copper.

For starters, it is substantially cheaper. At the time of writing, the price for aluminium on the London Metal Exchange is about $A1,800 a tonne, compared with just on $A7,000 a tonne for copper.

Now there’s a strong incentive to use aluminium if ever there was one.

In addition, a given volume of copper weighs 3.4 times the same volume of aluminium, so the weight savings can be considerable.

On the down side, aluminium has drawbacks that need to be dealt with, otherwise things may not work as expected.

Aluminium has only 62% of the conductivity of copper, so for the same amount of loss on a cable, thicker conductors must be used.

Mechanical properties

Copper is a soft and malleable metal. It does eventually suffer from work hardening, but a copper wire can be flexed many times before it fatigues.

Aluminium is much harder, and it readily cracks and fatigues when flexed. It also tends to creep when under prolonged stress.

These characteristics are not such an issue with large-diameter power cables, in which there are relatively few joints. Also, the correct jointing procedure for aluminium conductor power cables is well established.

However, it is a different story for small conductor multi-pair communications cables employing punch-down terminations. In the case of data cables, reliable termination is a crucial issue. I am not saying that it can’t be done, but experience shows that much research and testing is required to achieve consistently reliable performance.

Bare aluminium quickly forms a clear oxide when exposed to air. Aluminium oxide is a good insulator, so any form of jointing or termination needs to take this into account.

For the Chinese cable in question, the conductors have a copper coating, but the coating is very thin and easily removed when stripping insulation from a conductor with side cutters (Figure 1).

Jointing and termination

All data cabling is terminated using some form of insulation displacement contact (IDC), be it on a jack or in an RJ45 plug, and all IDCs were mechanically designed to terminate copper conductors.

IDCs rely on maintaining an airtight metal-to-metal contact at the point of termination by using a spring force. The spring contact supplying this force is an integral part of the IDC contact design.

Because aluminium has mechanical properties very different to those of copper (hardness, work hardening and creep in particular), it is unwise to assume that all IDCs will work reliably on aluminium conductor cables.

An IDC places a number of stress points on the conductor material, so I wonder how aluminium wire will behave in the long term, given its propensity to work harden and creep.

The EIA Standard calls for mechanical testing of all cable. In particular 5.3.6 Cold Bend Radius, which states:

“Twisted-pair cables shall withstand a bend radius of 4x cable diameter for UTP constructions and 8x cable diameter for screened constructions, at a temperature of –20°C +/- 1°C, without jacket, insulation, or shield (if applicable) cracking, when tested in accordance with ASTM D4565, Wire and Cable Bending Test.”

Stress at the point of punch-down, combined with the relatively brittle nature of aluminium and its tendency to creep under continued stress could become a source of long-term warranty issues for the installer.

Transmission performance

A set of Standards compliance transmission tests quickly reveals severe shortcomings for this particular cable.

Conductor diameter for the sample under test turned out to be 23AWG, the same as used in most copper conductor Cat 6 cables.

However, as shown in Figure 2, the insertion loss fails by a large margin. Whoever designed this cable probably thought the copper coating would be enough to make it work.

At high frequencies, the current tends to crowd towards the outer edges of a circular conductor (the ‘skin effect’), so putting a copper coating on the aluminium would seem to be a good idea.

However, it gets more complicated. Skin effect is caused by the fact that the inductance of a wire is less at the outer edges than it is near the centre. Inductive reactance is negligible at low frequencies, but it becomes signi cant in the MHz regions, so the current crowds towards the outer edge.

There is a formula for calculating the skin depth – the depth at which the current falls to about one-third of the value at the surface.

At 1MHz, the skin depth for copper is 0.065mm, falling to 0.004mm at 250MHz. For aluminium, it is about 20% more, due to the lower conductivity of aluminium.

However, at least two skin depths are required to make sure most of the current will  ow in the copper part of the wire. At two skin depths, the current is down to about 15% of the value at the surface, which is reasonable.

Given that 23AWG corresponds to ~0.28mm radius wire, and two skin depths corresponds to 0.13mm at 1MHz, the cross section area of wire that needs to be copper is about 70% of the total cross section area!

So if copper plating on aluminium core wire is to have a chance of working, the wire would still need to be at least 70% copper by volume, assuming we stick with 23AWG conductors.

Therefore the copper coating on the sample wire is not nearly enough to have the desired effect. Even with all-copper 23AWG wire, most cables pass insertion loss by only a small margin.

This is a classic example of a quick calculation showing up a bad idea.

In addition, shortfall in insertion loss performance also causes sub-standard results in equal level far-end crosstalk (ELFEXT). (See Figure 3.) And if that wasn’t enough, the return loss performance was below par as well, but this may not have been due to the use of aluminium conductors. It’s hard to say – maybe the pair-twisting machine was not adjusted to accommodate the mechanical properties of aluminium wire.

Conclusion

The above information speaks for itself. Do not use aluminium conductor data cable – it does not do the job.

George Georgevits, BE (Hons), manages his own communications engineering consultancy Power and Digital Instruments, which was established in 1980. PDI specialises in lab and field transmission testing and troubleshooting of cabling systems and components, as well as general electronics and communications engineering. Contact PDI on +61 2 9411 4442.