We are all familiar with RG-6 type coaxial cable. Many of us have probably earned blisters installing BNC connectors on these cables at one time or another in our careers. As more facilities are being rebuilt with SMPTE 259M, SMPTE 292M, ASI and perhaps some SMPTE 310M around the microwave or at the transmitter, this is a good time to review the reasons why a particular cable is or is not suited for a particular application.
This month’s column discusses the various sorts of coaxial cable used for studio video applications. Many of the criteria used in deciding what type of coax to use are also important when selecting other types of cable. The column also covers issues that are important when selecting connectors and patch panels.
Cables and connectors are often lost in the shuffle when fancy new equipment arrives on the scene. But cables and their terminations are the backbone of the plant and will typically outlast the equipment they interconnect. Although cable is not particularly expensive, the labor required to install, label and terminate it is. Selecting and using the wrong cable can cause budget and schedule problems. A bit of time and investigation in this area will pay off handsomely over the long term. For example, when wiring an equipment rack, ask yourself what equipment might end up in that rack in the future – high definition, perhaps? If so, it would be nice to simply re-label the existing cable and carry on rather than re-wiring.
When selecting cable, ask yourself the following questions:
What signal will the cable carry, and how far will it carry it?
How much loss can the receiving device tolerate?
In what environment will the cable will be used?
How and where will the cable be installed?
Are there any size, weight or fire restrictions on the cable?
Could the use of the cable change in the future?
Let’s take a look at the signals a new plant might carry. There is NTSC, component analog, SMPTE 259M, DVB-ASI, and possibly SMPTE 310M and SMPTE 292M. The frequencies of these signals range from near DC to well over 1.5GHz. The characteristic impedance specified for all these standards is 75V. That leaves cable loss and shielding as the major issues with which to contend, particularly with digital signals that must travel over any appreciable distance.
Digital signals have the advantage of being tolerant of cable loss because the receiver regenerates the signal. The receiving equipment’s ability to lock onto the stream and decode it sets the maximum cable length for any particular cable. Several different manufacturers make receiver chipsets, and the performance of these chipsets differs, so you must allow headroom in your design. Patch panels and demarcation panels can also reduce the cable’s useable length.
Acceptable losses, as stated in the applicable standards, are: SMPTE 259M, 20- to 30dB at half the clock frequency; SMPTE 292M, up to 20dB at half the clock frequency. DVB-ASI has the same bit rate as SMPTE 259M and the same cable requirements.
Although the rise times of digital waveforms are controlled, harmonics are an important consideration for accurate signal recovery. Cable tilt could be a factor when using smaller, higher-loss cabling. If you are terminating longer runs into a patch panel and using smaller cable from the patch panel to the switcher, you must account for the cascaded losses of both cable runs and the patch panel to ensure adequate headroom. Failing to do so could cause intermittent problems that are difficult to locate.
The cable tilt is defined in the SMPTE 259M standard. The frequency response, in decibels, is approximately equal to 1 divided by the square root of the frequency, from 1MHz up to the clock frequency of the signal being carried.
You can get an excellent indication of the cable’s bandwidth from its velocity of propagation, which is available from the cable manufacturer’s catalog. The higher the velocity, the lower the shunt capacitance and, therefore, the higher the frequency cutoff.
As a general rule, use a larger, lower-loss cable and use the same cable throughout the signal path unless there are substantial constraints on space, weight and distances (such as in a truck). Using the same type of cable also simplifies connectors and connector tooling. The difference in cost between large and small cable is minimal – generally within the “noise floor” of a project’s budget, but the larger cable offers insurance and peace of mind.
A quick look at the table reveals an obvious trend: the higher the data rate, the shorter the typical cable length. For SMPTE 292M, this trend shows that you will require larger-sized cable unless the cable runs are short. To avoid the error cliff (and as a matter of just plain good engineering practice), de-rate the values by 10 percent. To allow for patch panels and different runs of cable, de-rate the numbers in the table by another 10 percent. Margin is cheap; re-installation time and cost isn’t. The author has seen a “good run” go bad when equipment at the receiving end was replaced by that from a different vendor.
There are three shield types available: braid, double braid, and foil and braid, and all are available for different frequency ranges. Considering the fact that digital signals span wide frequency ranges, and the fact that ingress could possibly just nudge the receiver over the error cliff, foil and braid is a good choice. And a bonded foil makes connector installation much less troublesome.
The installation environment is another factor to consider when selecting cable material. Starting with the outer jacket, the primary concern is plenum or non-plenum and UL and/or other agency approval. Generally, the plenum cables have a 20 percent higher loss over the non-plenum version. If the new cables are being pulled into existing conduits or cable trays, jacket friction could affect the existing cable. Some jacket materials in existing cable can develop holes or “burn spots” when new cable is dragged across it during installation.
Generally, the ability of a cable’s outer jacket material to resist chemical corrosion is not a concern in a studio environment. If you are installing the cable in an unusual environment with conditions such as high heat, water or battery fumes, the cable manufacturer can supply information for these applications. The cable’s inner dielectric is an important consideration. Foams have lower loss but are not as crush or abuse tolerant as polyethylene. Some NEC-rated foams have slightly higher loss than polyethylene.
Lastly, some cables are available in a broad range of outer jacket colors for color-coding the signals.
If you carefully select the loss, shielding and jacket characteristics of a cable for the application, the cable failure modes will be limited to manufacturing failures or installation errors.
The manufacture of a coax cable is a dynamic process. It takes place on a production line that can be many hundreds of feet long. Once such a line is set up and the copper-drawing and foaming processes start, it is a major event to shut the line down. Automatic detectors are used throughout the line to measure wire diameter, speed through the dielectric foaming machines, wire temperature and center-conductor centering. When a fault is discovered, the section with the error is marked and physically cut out of the finished cable in a later process. The line keeps running. On rare occasions, errors go undetected, resulting in a reel of cable having a bad section in it.
A type of manufacturing error difficult for the end user to discover is periodicity. This is a small change in the impedance of the cable on a periodic basis. It can be caused by a bad bearing or an eccentric wheel used in the production-line equipment. Such faults can cause minute speed changes in the wire as it travels down the line and can affect the thickness of the wire or the dielectric, resulting in a slight impedance change at that point. If the period is related to the wavelength of the signal, then the cable’s return loss will increase. The higher the frequency carried in the cable, the more this manufacturing error will affect the usability of the cable. The effect is the same as using the wrong length of transmission line at the transmitter site for the TV channel in use. To catch this problem, perform a return-loss sweep on the cable before using it. Unfortunately, this type of incoming quality control is rarely practiced. The end user nearly always depends on the manufacturer’s quality control to catch this problem. Poor installation techniques can also damage a cable in a periodic manner and cause an increase in return loss, although this is generally localized. For example, tie-wrapping a bundle of cable too tightly at regular intervals of the cable run could introduce a periodicity error in the cables located at the bundle’s periphery. Dropping a full roll of cable on a loading dock could also cause a similar periodic deformation of the cable.
Installation errors can cause immediate damage to the shape of the cable or can result in such damage over time. Examples of installation errors are:
using a pulling effort greater than that for which the cable is rated
exceeding the manufacturer’s bend radius
providing inadequate cable support
Kinking a cable and then straightening it out again does not fix an impedance discontinuity, particularly with low-loss foam dielectric. Cable used in high-frequency digital applications must be installed with attention to the manufacturer’s guidelines.
The standards mentioned earlier call for connectors having an impedance of 75V. This can be achieved either by using a true 75V connector or by using a 50V connector on equipment having impedance-compensating circuitry so that the combination yields 75V.
Although not directly related to cable selection, the choice of a connector is important, especially when purchasing a large number of connectors. The characteristics of connectors from different manufacturers differ widely. In particular, center-pin retention is an important consideration. Pin retention in both directions is important. Using connectors that lack sufficient center-pin retention can cause intermittent connections, and result in many hours lost tracking down such problems.
A patch panel can be your best friend or a saboteur. Many engineers may be tempted to use existing old faithful jack fields for digital signals. This could reduce headroom. There are measurable differences in patch panels from different vendors. If you are building a new plant, it is advisable to rent a network analyzer and perform sweeps on vendor samples to determine their impedance and response over the frequencies of interest. Other questions you should address about jack fields include:
Are the connectors standard?
Are the jacks sealed or in some way protected from dust?
Can jacks be changed from the front of a panel?
Is there adequate spacing between jacks on a panel so they do not touch each other now and will not years later due to cable loading or movement?
Lastly, you should sweep terminators with a network analyzer. This process will be a real eye-opener, exposing the differences between vendors. This may not seem very important, but you are putting a lot of time and money into the plant, so it’s best to attend to the details.
When selecting a cable to use in a digital plant, your primary concerns are loss and quality issues. These higher-performance cables, and the signals they carry, are not as forgiving to manufacturing and installation errors as those used in the good old analog days. However, they are easy to install and use. And, if you follow the manufacturer’s guidelines, the cables will provide a long service lifetime.
David Lingenfelter is director of engineering for The Evers Group.