INGRESS - SOURCES and SOLUTIONS

by John w. Ward Jr.

Comcast Cablevision of Montgomery County, Inc. Signal ingress is a problem which has external radio signals are especially high, always plagued cable systems. While there even defects that are otherwise insignifi­ has always been potential for problems cant must be found and corrected. from VHF television stations, with the modern cable system encountering the UHF Interference due to ingress can be television band, the susceptibility of classified into two basic forms, either co­ cable systems to ingress interference is channel or discrete carrier. When there are increasing. Immunity from ingress problems one or more local VHF TV stations located can only be achieved by maintaining system near a cable system which uses a channel integrity at levels better than those occupied by one of these VHF stations, required by specification. there will without doubt, sooner or later, be need to correct co-channel interference Interfering signals leak into a cable between the two. Discrete carriers from system not only thru flaws in the cable communications transmitters will cause system but also by way of consumer problems on mid-band and super-band cable equipment, cable ready televisions and channels. As communications transmitters VCRs. The increase in the number of these include everything from car phones and devices as well as other factors lead to a personal pagers to amateurs and the need to be able to efficiently diagnose and National Weather Service, discrete carrier cure signal ingress problems. ingress can occur anywhere and often at random times. Cable systems near the VHF TV WHAT IS INGRESS transmitters of a large city are the systems that will likely suffer from co­ Ingress, as far as the CATV community channel ingress related problems. At two to is concerned, is the entrance into a cable five miles from a full power TV transmitter system of any undesired external radio it is not uncommon to have a field source. Ingress will occasionally be in strength of 35 to 40 dBmv or more, very the form of static or electrical noise, but often 25 dB more than than what is inside it is normally considered to be interfer­ the cable. Levels from TV transmitters as ence from a radio frequency signal. Ingress far away as 30 miles may exceed the average of such signals will result in interference levels of a CATV plant. Beyond that range to cable pictures. Modern cable systems the effects of ingress related co-channel have sufficient isolation from ingress, or interference become less noticeable. shielding, to prevent ingress of carriers in even the noisiest of radio environments. Co-channel type ingress interference But, an awareness of ingress related in its most basic form will appear on a TV problems is required by the cable system picture as a strong beating pattern , when operator to enable repair of inevitable, the cable channel is not phaselocked with natural, flaws in the system. the interfering station. If the cable pro­ graming is phaselocked to but not sync Ingress is the opposite of egress, or locked to the unwanted carrier, a wiping of system radiation. The principle by which the interfering stations' sync bar through both phenomena operate is the same, related the background of the desired picture is by the principle of antenna reciprocity, the first effect noticed. If cable pro­ which is that antennas transmit and receive graming is both phase and sync locked to equally well. That cable systems do radiate the local station, as it is when operation energy is an established fact, that cable is "on channel", the first effect will be systems are susceptible to ingress there­ faint ghosts in the picture, either of text fore follows. As system egress levels are characters with their high energy edges, or reduced to within regulation, effects of of the horizontal sync bar, stablized, but ingress are reduced toward acceptable in the middle of the screen. The difference levels. Still, in order to completely over­ in the time it takes the signal to arrive come the effects of ingress in areas where at the set both through the cable and

1985 NCTA Technical Papers-11 through the air allows the channel to HOW DOES INGRESS GET IN interfere with itself. If the two signals arrived at the same time they would mesh Ingress gets into the cable system by perfectly, all the interference would be way of poor shielding and faulty connec­ hidden. tions. The cable acts as an antenna and will have currents from external radio When operating "on channel", or with fields induced onto it's shield. Electron alternate programing phase-locked and sync flow, or currents, of radio frequency locked to a local VHF transmitter, the energy, happens only on the surface of a cable signal must be at least 50 dB or more conductor. Under normal conditions, the above any ingress from the airwave signal cable signal energy flows on the inside in order to suppress sync bar ghosting of surface of the shield and broadcast radio the picture. Without sync lock, a separa­ signals flow on the outer surface of the tion of 55 dB is a minimum to prevent an shield. A "hole" in the shield will join annoying wiping pattern in the background. the two surfaces, allowing undesired cur­ If the cable channel operates with altern­ rents to flow both out and in. Unbalanced ate programing not phaselocked to the local current flow between the shield and center TV station sharing the same channel, the conductor of the cable will cause the un­ cable video carrier should be 60 dB higher desired signal to be added to the cable than any interference. In extreme cases of signal. co-channel ingress interference, where an off air signal is only 40 or 45 dB down Experience has shown the most common from the cable signal, strong ghosts or point of ingress to be a slightly loose other distortions may be expected, regard­ connector. The connection is normally tight less of the operating mode. Customers will enough not to noticeably effect the cable surely call for service when the interfer­ signals, if not for the ingress problem. ence is this bad. The connection may be just loose enough to permit air molecules to permeate between the threads and, given time, form a layer The effect of ingress due to discrete of corrosion. The improperly made connector carrier interference is similar to second will also permit gasses to corrode the order beat problems in that both will aluminum of the shield itself, forming appear on the customers set as a herring­ aluminum oxide, a poor electrical conduc­ bone pattern or wavy series of vertical tor. Corrosion will create a point of lines. If several unwanted carriers are resistive and/or capacitive nature in the present, as is often the case with channels shield of the cable. This breakdown of the shared by the communications bands, the outside conductor is the unwanted hole in effect may be a soft distortion similar in the shield. symptoms to third order product accumula­ tion. A discrete carrier located near the Theory and experiment show that a color sub-carrier of a cable channel may mismatch on the inner conductor will not cause a beating pattern in picture tint or, permit signals to enter the cable, only the if sufficiently strong, may even drive the signals inside the cable already will be picture into black and white. affected. This can be demonstrated by cut­ ting the center conductor short at a splice in the middle of a section of drop. The Picture distortions due to discrete isolation is as good as the shield in this beats vary depending upon the level of the experiment. It can also be demonstrated beat as well as the position of the beat in that a single crack or hole in the shield, the band of the cable channel. For example, an interfering carrier 30 dB down and 10 khz. from the cable video carrier will 60 probably not cause any noticeable picture low band distortions. Interference located 1/2 mhz so high band --- above the video carrier with a level as low as 40 dB down from it, will cause strong d 40 ..._..,, -- beats in the picture. When the interference B is from a carrier located in the middle of m 30 '~,,, the cable channel, 1-1/2 mhz away from the v ,, video, a rejection of up to 55 to 60 dB is 20 principal area limit necessary to prevent a "busy background" grade "A" signal effect. The amount of immunity required by 10 --'~-~' -- the cable system from ingress beats is the ,, same as and can be compared to FCC specifi­ 0 cations regarding carrier to second order beat ratios. miles 1 2 4 8 16 32'' 64

FIELD STRENGTH OF TV TRANSMITTERS [3]

12-1985 NCTA Technical Papers not completely around the cable, is in and it is next to impossible to find a itself not a significant source of ingress. proper fitting. The fitting must not only However, when the small cracks are spaced pass signals but must properly seal the at regular distances, an effective amount shield from ingress as well as survive of energy is transferred into the cable (as through time. well as out). Other parts of the cable system Improper handling or installation of responsible for ingress are loose amplifier drop cable can cause periodic cracks in covers and tap plates. Although experience several ways. One common way a flaw may is that an amplifier housing must be open arise is when a staple gun, faulty itself and the amp's module cover almost off in or improperly used, causes a severe sharp order to get a significant amount of in­ dent in a drop wire as the staple is fired. gress into the cable itself, amplifier Even though the outer shield is not actual­ covers must not be ruled out. ly pierced by the impact, a small crack might be created. A series of a dozen or Tap plates, however, especially when so of these, regularly spaced 18 to 20 drops are connected to them, are crucial inches apart, can reduce the shielding of a points of shielding breakdown in the feeder drop, down from a nominal 90 dB, to only 50 system • The rf shield around the edge of a or 60 dB of isolation at mid band frequen­ tap plate can only work well when making a cies. Periodic bumps and cracks in drop good pressure connection to the housing cable can also be caused by rough pulling plate. Loose tap plates, with contaminants the wire from boxes and reels, or flexing between the plate and the housing will the cable sharply around corners, although cause a discontinuity to occur between the the observed occurrence of this type of drop shield and the shield of the feeder failure is rare. cable, allowing ingress into the drop and to a lesser extent, into the feeder itself. One way the largest.amount of unwanted Even when the tap plate is tight, corro­ signals can be transferred into the cable sion due to moisture is frequently a prob­ is by a total discontinuity of the shield lem as the rf gasket is located at the due to radial cracks. Faulty connections point of maximum water accumulation as a are similar to these radial cracks. A tap hangs on the feeder line. A very thin radial crack all the way around the cable layer of waterproofing grease will aid in shield will typically reduce the cable preventing this problem. A word of caution signals by about 10 to 12 dB, implying, in needed here, over zealous tightening of tap the worst case, as low as 3dB isolation plate screws will lead to striped housing between the the outside and inside of the threads, clutch type torque drivers are cable. On the other hand, a typical "bad" recommended. connector might reduce the isolation to 40 dB, with less than 1/10 of a dB reduction Studies have indicated that an in cable signals. unterminated tap port will provide greater RF isolation than a terminated tap port.[2] Compared to defects created by bad The terminator itself is a connector and connections, the amount of shield provided hence subject to the inevitable natural by the wire itself is of minor importance corrosion of the connector threads. As the as far as ingress is concerned. With trunk outer shell loses its ground connection, and feeder lines, the shielding is complete the terminator becomes a stub antenna and as possible, with more than 110 dB of iso­ hence a point of ingress into the cable lation often the case. Flexible drop wires system. The port to port isolation of a two with foil shields under a wire braid, the way splitter is normally about 25 dB and a type used by the cable industry, typically are rated with 85 to 100 dB shield iso­ lation.[l) The ability of the cable -10 shielding to physically withstand handling and to survive the elements is of more d -20 Visible Beats importance when selecting drop cable of B this quality then the shield factor itself. -30 D The shielding factor of the drop 0 -40 cable becomes significant when non standard w wire is used, such as a situation in which N -50 a house has been wired by a customer using his own wire. The shield for this wire can -60 be as low as 50 dB for wire with a heavy Invisible Beats braid, 35 dB or so with typical 40% braid ---0 1 2 3 5 6 shield wire. Also, aside from the poor mhz. from band edge to band edge shield, it is almost impossible to make a proper connection to these wires as the dimensions vary greatly from type to type CARRIER/INTERFERENCE RATIO [4)

1985 NCTA Technical Papers-13 stub antenna about an inch and a half long some extreme cases it might be necessary to will pick up as much as 0 dBmv of signal replace game switches located after the near a high band VHF TV station. A strong converter if a local transmitter is operat­ to moderate interference is observed on ing in a channel adjacent to the converter drops connected to adjacent tap ports. output channel. The lower sideband of a broadcast station extends well into the Port to port transference is also lower channel. Suppressed properly, it responsible for mysterious interference nevertheless has enough strength to over­ problems when temporary disconnections of a come the poor shielding of these inexpen­ neighbors drops are made. In one case, when sive, manufacturer supplied switches. a neighbor disconnected the cable drop from his VCR, the center conductor of the drop Of 'all the ailments created by would contact the metallic case. The field subscriber equipment, problems due to the strength of the local TV stations were poor shielding of some cable ready TV sets about +35 dBmv in the area. This resulted are the only truly incurable ones. The in +10 dBmv of interference being back-fed amount of ingress which is introduced by into the other customers' otherwise perfect cable ready televisions varies greatly from drop. And, due to the directional coupler model to model and no brand can be said to characteristics of the tap itself, the rest be best. There are many models of cable of the system was uneffected. This happened ready televisions that exhibit excellent every night for a few hours at a time, shielding while other models of the same greatly reducing the mental stability of brand don't. It is also found that problems the service personal. The solution was with cable ready sets depend upon the simple when the reason was discovered. The location of the set in the room , as well customer with the VCR was given an A/B as the strength of the interfering local switch so he could switch inputs and still transmitter. Cable ready sets are subject maintain system integrity. to the same conditions as the rest of the cable plant and if there is a potential for SUBSCRIBER CAUSED INGRESS ingress, it will enter the system through the poorly shielded television just as easy Perhaps the most perplexing cause of as it would any other part of the distribu­ ingress problems is subscriber owned equip­ tion system. It will sometimes be neces­ ment. It is the one part of the system over sary to tell a customer there is nothing which the cable operator generally has the that can be done, that the set itself is least control. As the consumer becomes more the problem. and more 'video active', the occurrence of ingress problems due to consumer related Cable converters have the critical equipment is sure to increase. The fact portions of the signal path inside a tight that subscribers will loosen drops by metallic box, which constitutes a good simply moving converters as well as by shield. A typical television will have the connecting their own equipment will be a shielding open on one side, or have a cir­ sure ingress problem from now on in any cuit card pass through it, with the shield metropolitan area. In a typical case an only spot soldered to the circuit card, otherwise perfectly good VCR and TV set leaving gaps enough to allow more ingress will be connected with factory included than several loose fittings, plenty enough wiring with easy to use push on fittings. to cause problems. The set will otherwise The customer installs the source and will work perfectly, on channels other than the suffer the effects of ingress. Installation ones occupied by any communications band or of quality wiring with proper connections local television station. The best option will correct the problem. While the solu­ the cable operator has in order to correct tion is easy, it is a service call never a ingress problem directly inside the tele­ the less. vision set, is to place a converter before the cable ready set, allowing the set to Another common occurrence with operate on a clear channel. This sometimes customer installed equipment leading to an upsets the customer who has paid extra for ingress problem is the video game or compu­ the cable readiness, and often complicates ter switch normally supplied with such the instructions of using various remotes equipment • When installed before a cable in order to gain satisfactory operation. ready TV or VCR , or indeed installed any­ where but after a converter, they will Aside from placing a converter before without doubt permit ingress. When video a cable ready set, the only other practical games or computers must be connected to solution to this problem is to raise the cable ready TV sets it is necessary to signal levels into the set to a point where install a well shielded, self terminating the level of the interference becomes in­ CATV grade A/B switch in place of the cus­ significant. If the cable ready TV set has tomers switch. Using these and standard a shielding factor of 40 dB for example, adaptors available at local electronic about the average for the a problem causing dealers, a connection can be made that will set, and the field strength of a local provide a proper amount of isolation. In transmitter is 0 dBmv, an input level of

14-1985 NCTA Technical Papers +10 dBrnv to the set will a provide brute The possibility of ingress into the force solution. Even this solution proves cable system of signals which could inter­ impractical if the set is located only fere with data carriers used by addressable several miles from local TV transmitters, converters and horne security systems is the input levels needed to mask the greatly magnified by FM tuner hook-ups. interference will be more than the highest With FM transmitters often as powerful as level the cable system is allowed to TV stations, combined with a tuner's operate at by law. characteristically poor shielding, back­ feed of unwanted signals into the cable can create serious problems. As the data car­ The degree of the isolation provided rier is already well below the operating by cable ready sets can easily be deter­ level of a cable video carrier, the side­ mined by reading the strength of the known bands of a local FM station can easily source of interference directly out of the distort a data carrier. Data transmis­ back of a TV set and comparing this to sions, when distorted by interference, may dipole readings at the same location. If contain errors resulting in random charac­ one positions the antenna for maximum ters being received between transmissions, receive levels , and does the same with a or worse yet, in severe cases, the corrup­ television receiver, with the TV set off to tion of a desired transmission. To avoid prevent reading RF generated by the set problems with data communications in the FM itself , a direct estimate of the shielding band, you may use a directional coupler factor of the set may be made. between FM tap-off device and the effected equipment to prevent backfeed. The use of regular splitters in place of FM taps Cable ready VCRs present another should be avoided. potential source of ingress, fortunately though they do not generally appear to be as great of a problem as cable ready TVs. TROUBLESHOOTING THE INGRESS PROBLEM Although poor VCR shielding has ocasionally been the source of ingress, the cables used As the number of urban homes wired for and other problems with connections are cable increases and as older drops aproach much more bothersome than the VCRs life expectancy, the amount of service themselves. calls related to ingress is bound to in­ crease also. The problem is compounded by the wider bandwidths of modern cable sys­ A poorly shielded cable ready set can tems as they encounter more communications also present a problem to other sets if the bands and even UHF television. Increases in field strength of a local tv transmitter is the number of communications transmitters moderately strong. It is quite possible for in the spectrum and the huge impact of a set with 30 dB or less shielding factor, consumer equipment will lead to the need about the worst encountered, located in a for quick and sure means of detecting and typical urban environment, to pass ingress correcting ingress problems. Even today it at -10 dBrnv up a drop to a splitter. There is often the case that a cable system in a it will back-feed down the other drop leg urban area will find 10 to 15, and some­ at a level of -35 dBrnv of interference times as much as 30, percent of its service versus 5 dBrnv of signal, more than enough manpower spent finding and correcting in­ to be noticeable. gress related problems. Routine procedures must be taught to service employees so that they can handle the problems efficiently. FM hook-ups create two types of ingress related problems. The first is with interference to FM services provided by the It is often difficult to distinguish cable system, and the second with address­ between faint interference and the symptoms able converters when connected along with a of amplifier distortion. However, if the FM hookup. The typical FM tuner will work interference is strong enough to cause a perfectly well with little or no antenna heavy beat, it may generally be assumed connected to it in an urban environment. that if a amplifier was ernittirng a spurious While the shielding of the tuner is gener­ product strong enough to be clearly visible ally very poor, it is possible to deliver it would probability have other by-products quality FM signals thru a cable system. If on adjacent channels. If you have a problem the signal from a local FM station is with beats on one channel only, it will delivered unshifted in frequency, FM most probability be ingress related, a tuners will generally be unable to distin­ quick check of other channels will provide guish between the two carriers, in the an answer. Also, high signal levels at a cable and off the air. If the cable opera­ customers set will likely indicate ampli­ tor is careful to avoid using FM channels fier distortions as a cause of beats, for within 400 khz of local FM transmitters for as the levels go up, the probability of other, imported or operator generated distortion increases and problems from signals, problems may generally be avoided. ingress decrease.

1985 NCTA Technical Papers-15 In an area where it is expected that any frequency desired when the levels of there will be a sufficient number of in­ the cable signals at that point are known. gress calls to warrant it (an area where By disconnecting different sections of a there are many local TV stations and etc.), drop at a splitter and measuring ingress it may prove very practical to leave the levels from each, a fast and sure trouble­ channel of the station most likely to leak shooting decision can be made. Readings into the system empty. The empty channel from a disconnected section of drop, will provide a convenient way to determine terminated at the other end, will indicate the degree of system integrity. Measure­ if a drop is good or if it must be serviced ments of a local station leaking into the or possibly replaced. A drop cable in cable on a channel not occupied by any service should typically be capable of 70 cable channel, when compared to the levels dB or more isolation , so any detectable of an adjacent cable channel, will provide levels inside the drop would indicate the a good indication of the signal to inter­ necessity for service. ference ratio between other local TV stations and cable signals. If the tele­ As the drop is connected to the field vision set or converter is tuned to the strength meter, first insert only the cen­ channel of the local station not on the ter conductor of the drop in the meter, cable and a noisy but steady picture is then tune to the source of off air inter­ present, a faint ghost or beat on another ference and read the level (often close to channel will almost certainly be caused by or more than what is read with a dipole at ingress. the same location). If you then tighten the drop on the meter and turn the attenuator The point at which the ingress first all the way down a good indication of drop enters a feeder system can also be measured integrity can be made. If an amplifier with quickly by making signal/interference read­ a gain of 20 dB is placed between a good, ings at taps, via the unused channel terminated long length of drop and a method. Customers will still have service meter, in an area where the field strength while troubleshooting is being performed is 40 dB or more from off air transmitters, and this method will physically disturb the it is just barely possible to detect the system as little as possible. This is local transmitter above the noise floor , desirable as a small movement of a slightly indicating a 90 dB shield or better for the loose connector may be enough to correct drop (and the test equipment too). the problem for the present, but it will likely soon start misbehaving again. Also The only other practical way to any disturbance to perfectly good portions detect the source of ingress is to make use of the plant will often lead to them of the various sensitive radiation detec­ becoming less tight, resulting in more tors currently available from several manu­ problems. facturers. Just as ingress gets in cable, signals leak out and can be detected. With If it is necessary to maintain a this equipment, and little or no training signal to interference ratio of 50 dB or beforehand, a technician will almost be more between a local transmitter and a able to walk right up to a defect causing channel used in a cable system, then the ingress. Sensitive equipment capable of levels on the unoccupied cable channel of detecting radiation levels 15 to 20 dB the local TV transmitter should be -40 to below the FCC radiation threshold of 34 -45 dBmv at the input to the set or conver­ dBmv is required. A shield factor of 60 dB ter. This is near or below the lower limit or more should be maintained in drops when a normal field strength meter can read, so both the field strength of the local any deflection of the meter scale with transmitter and the signal level in the attenuation fully down is undesirable. If drop are 5dBmv. the video buzz can not be heard at all, or if system generated beats at the extreme Standard dipoles and meters are range of the meters sensitivity are heard difficult to handle as troubleshooting instead, then the service technician can be aids, and very often unable to detect faint assured the ingress problem is not coming radiation from points which are, neverthe­ from the upstream portion of the drop or less permiting noticeable ingress inter­ feeder. The test becomes more valid as ference. In areas of strong radio inter­ cable signal levels increase, and, as an ference, even the more sensitive equipment aside, provide rough measurements of system is sometimes incapable of finding faint noise. radiation from leaks permitting severe ingress , for example in drops when cable Another practical troubleshooting signals close to 0 dBmv and the local practice is to disconnect the section of a transmitters are above 20 dBmv at that suspected bad drop or feeder leg and point. measure the levels of the local transmit­ ters directly out of the downstream leg. Using equipment accepted as being many This permits a direct comparison of the times more sensitive than what is required signal to interference ratio to be made at to detect the FCC egress threshold, and

16-1985 NCTA Technical Papers capable of being calibrated so, will also 45 dB of system immunity is good enough for permit the technician to determine if a a character generator only, all that can leak legally needs to be reported or not. be expected if cable ready TV sets are to Typically leaks radiate -40 dBmv or less in be connected. The same channel will be more about 75 to 80 % of the service problems, usable under the same conditions in total while leaks above -25 dBmv are found less converter build. than 10% of the time. This severe a leak will normally affect customers severely, It might also be desirable to require and therefore will also be corrected very measurements of the ingress level present quickly. It is often the case, though, that in each feeder leg as it is turned on. With the strongest points of egress are points the input to a section terminated, a read­ where the level in the feeder is the great­ ing should be made at the end of each leg. est, hence the points where the greatest Accept no ingress whatsoever and require immunity from ingress can be expected. So checks to be made with a spectrum by no means does a lack of ingress mean anaylizer, photo's included. that a system is tight and totally within Plan to put a significant amount of it's requirements. energy into ingress related maintenance in an urban build until you have time to cor­ The repair actions taken to correct rect all construction and new drop defects. ingress problems due to cable faults are Even the best of construction methods and straightfoward and direct, tighten it, workmanship will show some minor flaws, and splice it, or replace it. Detection of a flaw can be very minor and still be a ~ngress problems will require training of significant point of ingress. But with a personnel as to the way to quickly distin­ properly equipped staff of technicians, the guish between ingress related beats and ingress problem can be overcome in a those of amplifier distortion. Service routine fashion, with luck. personnel should be able to determine if customer equipment is at fault and how to bypass these problems. Additional training ACKNOWLEDGEMENTS: is required as to how to use signal level meters as a means of locating a point of 1: Belden Corp,CATV Coaxial Cable Catalog ingress. Specialized equipment will make #ELl0-79, Oct. 1979, pp 15-18. the technician more efficient and help assure FCC compliance. The routine soon 2: Reg James, Comcast Corp.Staff Engineer will establish itself if the system is From a report July, 1982 susceptible to ingress problems. 3: Based on FCC Rules and Regulations,Vol NEW BUILDS IN HEAVY INGRESS AREAS III, part 73, ppl89-191, 1972 From "Reference Data for Radio Engineers"; Before any new build is first turned Howard w. Sams & Co., 1979, p 30-12 on, the cable operator should determine the levels of all local off air transmitters at 4: Based on Jerold "CATV Reference Guide" various locations throughout the build. It #RD-14, April 1983, p36 will then be possible to assign character generators and other similar programing to Special thanks to all the service channels likely to have interference from technicians who helped gather the data and ingress. Phase locking and sync locking and who came up with some good fixes, too.

1985 NCTA Technical Papers-17