Steve Lampen examines the issues and challenges of working with Triax cables.
If you look at most professional video cameras, especially those cameras intended for HD-SDI (1080i or 720p), the output options can be very specific. Fixed or robotic cameras can often use single coax cable to carry HD-SDI output with its 750 MHz clock.
SMPTE requires that the camera and associated equipment pass the second harmonic of the clock (750x2=1.5 GHz). Some manufacturers look at the third harmonic (750x2=2.25 GHz) or even the fourth harmonic (750x4=3 GHz) to assure maximum signal strength and data structure.
Many older cables are tested to much lower frequencies, or sometimes not tested at all. Their performance at HD bandwidths is then unknown. The second half of this testing is not just bandwidth, but how perfect, how accurate the impedance of the cable is at these high frequencies, referred to as "return loss".
A return loss of -20 dB means that 99% of the signal reaches the destination device and 1% is reflected (and lost). -20 dB RL (1% reflection) is commonly considered to be the minimum requirement for most digital devices, cables, connectors, patch panels, patch cords and adaptors. SMPTE requires -15 dB at 1.5 GHz (and a 3.16% reflection).
A return loss of -10 dB means that 90% of the signal arrives and 10% is reflected. This is such a huge reflection that many chips cannot deal with it and will shut down, or at least will produce greatly increased bit errors.
On the other hand, a return loss of -30 dB means that 99.9% of the signal arrives and only 0.1% is reflected. Any passive component that can consistently meet -30 dB RL would be excellent.
The whole point is, that anything intended to carry HD-SDI should be tested (and ideally guaranteed) to be no worse than -20 dB RL to 2.25 GHz. But there are other cable formats besides just a coax cable. For cameras, the two other output formats are triax cable, and fibre optic camera cable.
Triax presents a whole set of issues and considerations. First of all, when video cameras first started to go to HD-SDI, the maximum distance on coax (and, presumably on triax also), was around 150 metres (500 feet). I suppose someone compared that with the triax distance in analogue video, many thousands of feet, and made a decision that copper "would not go far enough".
Thus, the SMPTE 311M standard was born and, along with it, fibre optic camera cable. This standard calls out for a cable with two single mode fibres, four power conductors and a twisted pair of wires for control.
If you know anything about fibre optics, you probably know that single mode has no inherent distance limitation. Single mode fibre routinely goes across continents or under entire oceans without additional amplification or regeneration.
The real limiting factor for distance in a fibre-optic+copper cable is the copper DC power pair, limited by Ohm's Law (E=IR) and the voltage drop on those conductors. If the camera is "locally" powered, in a "fibre-distance only" mode, the fibre can go much farther.
But there is a real problem for those users who want to stay with triax. Triax is well-known, easy to use, and easy to repair when it breaks (unlike fibre!).
And, if you look at most manufacturers, they will tell you that they have triax options for their cameras that will go somewhere around 1000 metres (3280 ft.).
But if you examine the 'fine print' on the spec sheet, you will see the bad news: to achieve this extended distance, the camera manufacturer will convert the HD-SDI signal into analogue. It is this analogue signal that runs on the triax. And, at the end of the run, it is converted back to HD-SDI.
Obviously, many of the advantages of digital are lost in this approach. Converting the signal to R-G-B component, often 'wideband component' might improve the quality but does not fix the artifacts caused by conversion into analogue, and the conversion back out at the other end.
Some camera manufacturers convert their HD-SDI into 'Y-C', which can barely be called "component". If you don't believe this, check with the camera manufacturer.
The real question is: what's wrong with 150 metres (500 ft.) of real digital on triax? That distance should cover a lot of applications. This would be enough for virtually ever TV station or studio complex.
It would only be the 18th hole on the golf course, or the ski jump hill, that would require those really long runs. So those would be fibre. All we need is triax tested to 2.25 GHz (or higher) and at least -20 dB return loss.
Steve Lampen is the multimedia technology manager at Belden.
Belden is one of the few cable manufacturers that have one of the largest triax offerings in both European and American sizes. Some are tested to 850 MHz, and others to 3 GHz.
The company claims to be ready to offer a true HD-SDI triax output for those cameras that require this in the future.
Currently, it makes two fibre optic camera cables after the SMPTE 311M standard.
The Belden 7804E is available with PVC, PUR and halogen-free jackets, two single-mode fibres, four 20 AWG power conductors, and a 24 AWG twisted pair for control while the newer 7804C version comes with two 16 AWG power conductors.