The final, most boring reference article. There is probably no point in reading it for general development, but when it happens, it will help you a lot.
Contents of the article series
Subscriber's territory
So, your grandmother stopped showing the TV. You bought her a new one, but it turned out that it was not the receiver, so you should take a closer look at the cable. Firstly, often wrap-around connectors that do not require crimping miraculously twist themselves off the cable, which leads to loss of contact with the braid or even the central core. Even if the connector has just been re-crimped, it is worth making sure that none of the hairs of the braid is connected to the center conductor. By the way, the diameter of the central core is usually noticeably thicker than the hole in the socket of the receiver - this is necessary for good contact due to the expanding petals in the connector. However, if you suddenly replaced the connector with one in which the central core does not come out "as is", but goes into a needle (as in the ones I showed in connectors for RG-11), or changed part of the cable and the new core turned out to be thinner, then you may encounter the fact that the tired petals in the socket will not provide good contact with the central core.
When measured by the device, all this can be easily seen from the shape of the slope of the signal spectrum, which I wrote about in . So we can immediately control the signal level (I remind you that, according to GOST, it must be at least 50 dBmkV for a digital signal and 60 for an analog one) and evaluate the attenuation in the low and high frequencies, which will give us hints for further searches for the problem.
Let me remind you: the attenuation of the lower frequencies is usually associated with problems on the central core, and the strong degradation of the upper ones indicates poor contact with the braid, and this is usually associated with crimping (well, or the general poor condition of the cable, including excessive length).
Having examined the cable with a connector at the TV, it is worth tracking it throughout the apartment: since a coaxial cable is not just an electrical conductor, but a waveguide, not only breaks and other mechanical damage are significant for it, but also bends and kinks. It is also worth finding all the signal dividers and calculating their total attenuation: it may turn out that before that everything worked at the limit and a slight degradation of the cable led to complete inoperability. In this case, in order not to re-lay the cable hidden behind the trim, you can more correctly select the values ββ\uXNUMXb\uXNUMXbof the dividers or put a small amplifier at the entrance to the apartment.
If none of this is observed and everything is in order with the cable up to the very low-current shield on the stairs, then it is necessary to measure the signal level that goes into the apartment. If the level and shape of the signal at the tap of the subscriber divider is normal, then it is worth evaluating the difference between the values ββ\uXNUMXb\uXNUMXbof the TV and in the shield and think about where and what we missed. If we see that the attenuation to the TV is some reasonable value, but at the same time we see problems with the signal on the tap, then we should move on.
Riser
Having seen a problem on a subscriber tap, you should make sure that the divider itself is not to blame. It happens that one of the taps immediately or gradually degrades the signal parameters, especially in dividers for a large number of subscribers (more than 4). To do this, you need to measure the signal level on another tap (preferably as far as possible from the problematic one), as well as on the incoming trunk cable. Here again, an understanding of what form and with what level the signal should be is useful. The attenuation value at the subscriber tap indicated on the divider in the marking (for example, 412 - 4 taps of -12 dB each) must be subtracted from what was measured on the trunk. Ideally, we should get the figure that was removed from the subscriber tap. If it differs by more than a couple of dB, then it is better to replace such a divider.
If we see that the signal is already coming along a highway with a strong slope or a low level, then we will either have to familiarize ourselves with the project of the riser, or, by turning on the logic, figure out two things: the riser was built above or below and how far from the nearest branch we are. The first can be understood by where the cable comes from, which is connected to the input of the divider and where the one from the output goes. Usually it is not difficult to track the main cables right in the shield, but if they are not visible, then you can go to the floor above (or below) and see what value the divider is there. From you probably remember that with the distance from the beginning, the denomination should decrease. In the same place, I wrote about dividing the riser into parts (we usually call them "pilasters", I'm not sure if this is generally accepted). Usually one pilaster extends over 5-6 floors and at its beginning there are dividers with nominal values ββof 20-24 dB, and at the end - 8-10. When there is confidence that the problem is outside the floor, it is worth finding the beginning of the pilaster and taking a measurement from the main divider from which it starts. Here the problems are still the same: both the divider itself and the damaged cable and poor-quality crimping can have their effect. It even happens that after moving the connectors, the signal is restored (but more often it completely disappears). In this case, you have to re-compress everything in a row, and it will be just wonderful if the installers, having foreseen this, left a supply of cable. After all, when re-crimping, it has to be shortened. On the RG-11 cable, the problem of incorrect crimping is very common: this is either a failure to comply with the stripping standard, which leaves too long a central core (as a result, the connector does not fit tightly and the cable can jump out of it), or the same thing, but due to too much section A (see figure below).
Separately, it should be said that even a proper stripping will not save you from a mistake if the crimping does not completely upset the connector and the central core does not enter the βneedleβ of the connector. In this case, the needle has mobility, if you shake it with your finger. When the vein has entered well, it is impossible to move it. This must be checked for each unscrewed connector.
Dividers themselves in houses that are more than 10 years old can experience what is known among scale model collectors as "zinc plague".
Photos from the site
Divider housings made of incomprehensible alloys and located in bad climatic conditions can literally crumble in your hands when you try to unscrew the connector, or even just when you move the cables in the shield. And usually this happens when installers work in the shield, extending the Internet to someone, or some other intercoms.
If the divider from which the pilaster starts has not fallen apart in half, and the signal level on it is as bad as in the apartment, then it is worth finding the divider on which the very first branching occurs and measure the signal that comes to us from active equipment from the basement (or attic - as built). Having passed the riser in this way and not solving the problem, you will have to go look for active equipment and take measurements already on it.
Active equipment
First of all, it is worth noting that between the optical receivers and amplifiers there is also a distribution network built on the same principles as the risers, and therefore having the same kind of problems. Therefore, everything that is written above must be checked here, and only then to sin on the serviceability of iron.
So, we are in the basement (attic, main switchboard), in front of the box with amplifiers
It happensβ¦
If there is no signal in the riser at all and there is a suspicion that the amplifier is dead, then it is easiest to determine which one is easiest by touching its temperature. Even in severe frosts in unheated rooms, a working amplifier will be warmer than the environment, and it will blow cold from a burned-out one. If the temperature difference is not noticeable enough, then an autopsy will certainly show that the power indicator is not lit inside the amplifier. Such an amplifier is replaced with a known working one, and subsequently repaired using a conventional soldering station, because almost all failures are associated with banal swollen capacitors. When replacing amplifiers with remote power supply, it is necessary to de-energize the entire network to avoid a short circuit. Although the voltage there is not very large (60 V), but the current of the power supply that I showed you in can give a lot: when you touch the central vein on the body, a sickly firework is provided. And if such amplifiers do not always successfully survive power outages in the house, then with these special effects there is a non-zero chance to disable a few more devices, which then have to be searched throughout the house.
But it also happens that the amplifier is alive, but at the same time it gives a lot of noise to the network, or simply does not swing to the signal level set by the project (usually 110 dBmkV). Here it is worth at the beginning to make sure that the signal does not come already corrupted by measuring the incoming signal. Of the characteristic incurable problems of amplifiers, the following can be distinguished:
- Gain reduction. Due to the degradation of part or all of the amplifying stage, at the output we have the same signal level as at the input (or more, but not enough for normal operation).
- Signal noise. The operation of the amplifier distorts the signal so much that the Carrier / Noise (C / N) parameter taken at the output goes beyond the norm and interferes with the recognition of the signal by receivers.
- Scattering of the digital component of the signal. It happens that the amplifier passes the analog signal satisfactorily, but at the same time it cannot cope with the "digit" at all. Most often, the MER and BER parameters described in go beyond the permissible limits and the constellation turns into a chaotic mess, but it also happens to be funny when, for example, the amplifier forgets about one of the modulation parameters and instead of the constellation draws a ring or circle on the device screen.
If these malfunctions occur, the amplifier must be replaced, but there are troubles that can be eliminated by adjustments. Usually the signal at the output of the amplifier floats down and it is enough to reduce the value of the input attenuator. And sometimes, on the contrary, the amplifier starts to make noise due to the increased level at the input, then we press it down with an attenuator. All adjustments should be made on one problematic amplifier, because if, for example, we reduce the signal coming out of the optical receiver, this will affect other, serviceable amplifiers and they will all have to be manually reconfigured to the changed parameters. Also, due to overamplification, the figure may crumble (with little noise on the analog). I described the amplifier settings in detail in .
You can try to fix the slope with the settings. Often, when commissioning a newly built network, a large initial slope is not required to ensure good parameters at the ends of the backbone. But over time, due to cable degradation, it may be necessary to increase the slope, which, as we remember, increases due to a decrease in the level of low frequencies, which will need to be compensated by an attenuator.
Optical receivers most often also die simply due to nutrition. If it has a sufficient signal level at the input (which one - I wrote in ), then there are usually no problems with the output. Occasionally, the same thing happens - increased noise and insufficient output level, but due to the stinginess of the settings, this usually does not lend itself to treatment. The diagnostics are the same - we check whether it is warm or not, after which we measure the signal from the output.
Separately, I will say about the test connectors: you should not always trust them. The fact is that even if everything is in order, then a signal lowered by 20-30 dB may not have the problems that a βrealβ output has. But it often happens that problems in the path occur after the test tap, and then everything seems to be fine - but in fact, horror. Therefore, for complete certainty, it is always worth checking exactly the exit that we have looking into the highway.
Optical trunk
You can talk a lot about problems and their search on optics, and it's great that this has already been done before me: . I will only briefly say that if we see a signal drop on the optical receiver and this is not connected with something like this:
We have cormorants in St. Petersburg - you know. And they will get optics underground.
then cleaning or replacing the final patch cord can help. Sometimes it happens that the photodetector, or optical amplifier, degrades, here medicine is of course powerless. But in general, without harmful external influences, optics are extremely reliable and problems with it, as a rule, come down to a tractor grazing on the lawn nearby.
Headend
In addition to the obvious problems with power supply and connectivity with sources over IP networks, one of the main factors in the health of the headend is the weather. A strong wind can easily tear off or deploy antennas, and wet snow that has stuck around a satellite dish significantly worsens the quality of reception. It is difficult to deal with this, because the antennas are located as high as possible, where the weather is fierce and even the anti-icing heating of the plates does not always save, so sometimes you even have to clean them manually.
PS This concludes my brief excursion into the world of cable television. I hope these articles have helped broaden my horizons and discover something new in the usual. For those who have to work with this, I recommend to deepen the book "Cable Television Networks", author Volkov S.V., ISBN 5-93517-190-2. It describes everything that is needed in a very accessible language.
Source: habr.com