Noise may be defined as the combination of unwanted interfering signal sources whether it comes from crosstalk, radio frequency interference, distortion, or random signals created by thermal energy. Noise impairs the detection of the smallest analog levels which may be resolved within the demodulator. The noise level along with the maximum clip level of an analog signal path set the available amplitude dynamic range.

The maximum data rate of a modem is limited by the available frequency range (bandwidth) and signal-to-noise ratio (SNR) which is amplitude dynamic range. If more of either is available, more bits may be transferred per second. The information carrying limit was discussed theoretically by Claude Shannon and is known as Shannon’s limit, or information theory.

Because modems run close to Shannon’s limit today, no further advances will be made to traditional telephone line modems other than incremental improvement of V.90. The frequency range of the audio channel is very limited at about 4 kHz. V.34+ modems are limited to a maximum data rate of 33.6Kb/s by an SNR of about 36 dB caused mostly by network PCM quantization noise. While V.90 improves the SNR by utilizing the network PCM levels directly, it is still subject to Shannon’s limit.

xDSL modems take advantage of the spectrum above the telephone audio channel. While operating with somewhat less amplitude dynamic range they increase data rates by greatly increasing the frequency range of the communication signal (from about 10 kHz to over 1.0mHz). To do this they require the installation of special equipment at the central office and customer premise.

Modulation is a prescribed method of encoding digital (or analog) signals onto a waveform (the carrier signal). Once encoded, the original signal may be recovered by an inverse process called demodulation. Modulation is performed to adapt the signal to a different frequency range than that of the original signal. Here’s how it flows:

bits -> modulator -> audio -> phone network -> audio -> demodulator -> bits

Hence the name MODEM short for modulator/demodulator. The modem is necessary because the phone network transmits audio, not data bits. The modem is for compatibility with existing equipment.

Attenuation is signal loss due to the diminishing availability of signal energy, or signal power. As a analog or digital signal traverses across a medium, it fades. High attenuation may lead to the inability to recover the signal on the far end. Signal repeaters may be used on the transmission path to periodically boost the signal strength. Baseband transmission is extremely limited to attenuation. Broadband much less so. In addition, wireless communications is much less susceptible to attenuation that is wireline communications such as xDSL or cable modems.

Crosstalk refers to the interference between channels. In the xDSL world, the interference between nearby cables can have a negative impact on the performance of the affected cable(s). Have you ever been on the phone and heard some other conversation, not yours, in the background? If so, you have experienced the effect of crosstalk.

Near-end crosstalk (NEXT) occurs when the transmitter sends a signal and a nearby transceiver at the same end of link, through capacitive and inductive coupling, “hears” the signal.

Far-end crosstalk (FEXT) occurs when the transmitter sends a signal and a transceiver at the far end of the link, through capacitive and inductive coupling, “hears” the signal. FEXT will be of more concern in an asymmetrical system such as ADSL than symmetrical systems like HDSL. This is because strong signals originating from the near end, can interfere with the weaker signals originating at the far end.

A tight buffer cable can be terminated directly. THe 900 micron coating on the fiber is rugged enough to allow the connnector to be connected directly and if there is a 3 mm jacket, it is crimped to the connector for strength.

A loose tube cable has 250 micron buffer on the fiber in it and is too fragile to attach a connector directly. It has be be used with a breakout kit that sleeves the fiber in a protective tube before termination.

Multimode 100 Mb/s (FDDI or Fast Ethernet) is limited to 2 km (1.2 miles) by the dispersion of the fiber and the chromatic dispersion limits of the LED in the fiber. This holds for both 62.5/125 micron fiber and 50/125 which exists in some older installations (and is being used for gigabit networks now, perhaps in error.) Singlemode links are avaialble for over 20 km (12 miles) or more. All singlemode fiber for 1310 nm is the same, with a core diameter of about 8-9 microns and an cladding diameter of 125 microns.

If you splice it, you will get directional losses. Transmitting from 50 to 62.5 fiber, you’ll get virtually no losses but from 62.5 to 50, you will get a minimum of 1.6-1.9 dB loss due to the size and NA mismatch. (50 micron fiber has a lower numerical aperture (NA) than 62.5.)

Currently there is a great deal of confusion among Ethernet cable buyers concerning whether to purchase Cat5e, or to use Cat6. Most of this confusion comes from a misunderstanding by the buyer that buying Cat6 cable will give them an “all gigabit” network. This is not the case. Unless every single component in the network is gigabit rated, then you will never have a gigabit network, because your network will always run at the speed of your slowest device. Cat5e cable of good quality can run near or at gigabit speeds, it just cannot be “certified” for this use.

By comparison, Cat6 is designed especially for gigabit use, and is certified to operate at said speed. It becomes a matter of whether or not you want to pay all that extra money, for little or no noticeable improvement in the performance of you network. In most cases, it makes more sense to go with Cat5e. It is for this reason that most of your new installations in the private sector are going with Cat5e. It is more economical, performs well, and is readily available in many colors. Many IT professionals when asked about why they specified Cat6 for a specific job, often responded stated that they “wanted the best they could get.” This is the line of thought behind many purchases of cable. The average consumer often times is not aware that there is no real benefit to them to use Cat6, so they let someone talk them in to buying it. CAT 5 Cable Company is committed to helping people make good decisions about cable purchases

Often abbreviated STP, a type of copper telephone wiring in which each of the two copper wires that are twisted together are coated with an insulating coating that functions as a ground for the wires. The extra covering in shielded twisted pair wiring protects the transmission line from electromagnetic interference leaking into or out of the cable. STP cabling often is used in Ethernet networks, especially fast data rate Ethernets.

In the world of structured cabling systems the cryptic number 568 refers to the order in which the individual wires inside a CAT 5 cable are terminated. The termination could come at either the user’s end socket, the patch panel or termination frame or even the individual leads that connect a computer to the wall socket. There are currently two different specifications with respect to the order these cables should be terminated contained in the international standards document (ISO/IEC 11801:1995) as previously mentioned there is no indication as to which of these standards is preferred.

Non-metallic conduit for all outdoor use shall be coloured white or contain an indelible or durable continuous white stripe which is incorporated as part of the manufacturing process. Conduit must also be legibly and durably marked ‘COMMUNICATIONS’.

Hazardous conduit is not to be painted white and then used by cabling providers for installing outdoor customer cabling.