The cliff effect is a phenomenon that arises in digital as opposed to analog telecommunications, where reception is abruptly lost as signal strength fades or electromagnetic interference increases. The effect can be observed when traveling in a car that is equipped with satellite radio. As the signal source becomes eclipsed by patches of new vegetation such as tree foliage with high water (hydrogen) content in spring, it fades because high-frequency radiation is blocked.
The sudden drop off is not experienced in lower-frequency analog transmission, in which the loss of reception is gradual. The drop off is called the cliff effect because when quality of reception, which means degree of fidelity to the original, is graphed against authentic signal amplitude at the receiver, the line as plotted remains at a nearly constant level until it abruptly becomes vertical and drops to close to zero, the signal disappearing.
When a digital transmission is used to convey audio information, the post-cliff sound generally goes silent with little if any discernible noise. In the digital transmission of video, the flat-screen video image will generally freeze, often with partial or total pixelation. In data transmission, the signal may become corrupted, slow down or be lost altogether, characterized by the abrupt cliff effect.
Sooner or later, it is to be hoped that there will be a reverse cliff effect, as service is restored thanks to improved weather conditions or reduced electromagnetic interference.
The cliff effect is not an intrinsically good or bad thing, although it makes for an enhanced, user-friendly experience, particularly for audiophiles. In video, occasional pixelation is more interesting to look at and preferable to more frequent snow.
The cliff effect can be attributed to the fact that unlike analog signals that weaken in a linear fashion when signal strength decays or electromagnetic interference becomes more significant, a digital signal is essentially an all-or-nothing proposition. It delivers data to the receiving end that is either perfect or non-existent.
To improve digital TV reception, hierarchical modulation is often deployed. As the primary highest-definition signal approaches its cliff, the receiver intentionally scales back to a lower-definition stream to prevent total loss of the broadcast. This, of course, is dependent upon a transmission that supports a two-level signal.
The cliff effect is a familiar phenomenon to mobile phone users, who experience an acceptably clear call suddenly dropping out, especially when one or both phones is in a moving vehicle.
Digital video signals also exhibit the cliff effect when propagated over cable. The abrupt drop off happens when the cable length is excessive. This distance varies depending upon the quality of the cable. For this reason, before designing or installing a digital video transmission line, tests can be conducted that will evaluate cable to ascertain whether it is suitable in regard to quality for the length that is contemplated.
It is important to realize that in digital transmission the bit error rate increases significantly before the cliff location is actually approached. And although the picture will not be lost in this intermediate region, there will be some loss of fidelity. Nevertheless, these defects will not be visible for most viewers because of error correction technology that is incorporated in digital transmission. A certain number or errors per second can be tolerated before failure occurs.
Two frequently used tests for cable quality are the Bit Error Rate (BER) test and the Eye Pattern test. To perform a BER test, a signal generator is connected to one end of the cable segment or device that is to be evaluated, and that becomes the input. Typically, billions of data bits are sent over the cable. Input and output are compared, and the number of error bits per second is displayed. This instrument is often built in as a permanent part of a microwave repeater, or it may be autonomous for ad hoc testing.
The eye test pattern as displayed in an oscilloscope is produced when a digital signal is applied to the input and the instrument is configured so the traced continually overwrites itself in a fixed display. The degree of clarity depends upon adjustments and settings made by the user and upon the capability, particularly bandwidth, of the instrument.
The comments that follow are applicable to the Tektronix MDO 3000 Series oscilloscopes:
To acquire an eye pattern, begin by setting the time base so a single period of the repetitive wave is displayed, and center it in the screen. The position as determined by the horizontal delay should be such that the edge being triggered is not a part of the period that is viewed.
To proceed, press Acquire and Waveform Display, which brings up the Display Menu on the right side of the screen. In this menu, turn Persistence on, which is done by toggling the associated soft key. Then, use Multipurpose Knob a to increase persistence so it is infinite.
This is the procedure for creating an eye pattern from a digital input. It is accomplished when persistence is infinite because successive traces are superimposed with none of them going away. High and low tracks representing 0 and 1 are visible and any variation, as when the cliff is crossed, become manifest as a pronounced thickening of the line, so that the eye appears to squint or partially close. A similar effect occurs with regard to rising and falling edges. In general, the more white space you see, the better is the signal and accordingly the cable is seen to be of better quality for a given length.
BER and eye pattern test equipment as used in a cable design and production facility is enormously expensive, but a good high-bandwidth oscilloscope will provide the information you need to perform go-no go testing on a case-by-case basis.
Less-than-optimum cable for a given application, especially with regard to cable length and frequency of the signal that is propagated, will introduce several types of distortion that is evident in an eye pattern. An important concern is jitter. In the above context, this may be defined as a timing error due to misalignment of rise and fall times, which can result from, among other things, cable imperfections. (Termination faults, while not our focus in this discussion, are frequently a source of jitter. Therefore, in using eye patterns to evaluate cable, it is essential to create impeccable terminations so that faulty ones do not confuse the issue.)
The eye pattern produced by a perfect digital stream that is not compromised by less than ideal cabling or other problems would look like a rectangular box, but due to the sensitivity of the eye pattern test, there is bound to be some thickening of the rails and partial eye closure. The best way to get a sense of what is acceptable is to perform many tests of both good and bad data streams, and to correlate the eye patterns with results from BER tests.
Digital transmission signals are propagated in reference to clock signals. Early or late-rising and falling edges are faults that may be introduced by cabling imperfections. They may cause timing errors that are likely to make receiving equipment confuse 0 and 1 voltage levels. The severity may be great enough to cause poor performance if not equipment failure. In this instance the remedy is to go to a better grade of cable even if it is more costly.
A complete cable-testing regimen will include both tests. If they are performed prior to installation, the correct cable can be chosen at the outset, saving the expense of reworking the system.
Leave a Reply
You must be logged in to post a comment.