And Its Effect on Bandwidth
My goal when consulting or training is to help companies work with what they have to achieve the maximum quality bandwidth possible over paired copper. Getting to best quality sounds simple: you need a clean cable pair with good longitudinal balance within the reach for that particular type of bandwidth service from ADSL to G.fast.
Yet there are many obstacles to overcome to achieving maximum bandwidth. The biggest challenges to achieving maximum bandwidth:
• Longitudinal Balance
• Bonding and Grounding
Some troubles affect only one circuit, and other faults affect all of the circuits on the DSLAM or blade.
When considering longitudinal balance, we test a vacant cable pair and look for unacceptable longitudinal balance caused by DC type faults such as:
• Crossed battery from other conductors in the cable. Crossed battery is caused by water in air-core PIC cable, water in a splice, or sheath damage.
• Ring or tip ground. Terminating faults in pedestal or terminals, sheath damage.
• Pair capacitive unbalance. Open one side on a lateral, open one side beyond the customer, crossed with a non-working conductor.
• Split cable pairs. Split cable pairs are splicing errors. Usually in a splice where the tip side of one pair is spliced to the tip of another pair. Ring conductors are spliced correctly. Split cable pairs affect all circuits on the DSLAM or blade.
• Series resistance. Series resistance usually occurs in a splice. Splicing is done using an incorrect tool, but can occur when lightning passes through an insulation displacement connector.
Using the digital multi-meter, a good multifunctional test set can help you identify the problem(s) and then, resistance bridges, open meters, and a time domain reflectometer (TDR) will locate the problem(s).
Bonding and Grounding
You MUST HAVE continuous bonding from the central office or remote to the customer interface to mitigate or reduce noise from interferers such as AM radio, HAM and shortwave radio, and, with the implementation of G.fast, FM radio. The shield of the cables must be grounded to the distribution power neutral to reduce power influence. This process can take 2 or 3 weeks. (What I have found many, many times is that while working through the process the technician is called to a different case of trouble and he never returns to the site because of other demands to complete bonding and grounding issues.)
There are several types of noise issues to consider:
• Unacceptable Power Influence
• Impulse Noise
Unacceptable power influence is a bonding and grounding issue as is interferers. A disturber is caused by having other high frequency bandwidth circuits such as T1 or HDSL in the same binder group as DSL.
Impulse noise is a different type of noise. Impulse noise is normally power associated and runs across the entire frequency range. Impulse noise is typically intermittent and usually affects all circuits in that 25 pair sub-unit. If sync is possible FEC, CRC, and HEC errors are continuous, and often there is constant retraining of the modem.
Impulse noise is caused by florescent bulbs, microwaves, power tools, cracked insulators on phase wires on distribution power systems, failing rectifiers, and a myriad of other causes, but are usually power related.
Most multi-functional test sets have the ability to identify impulse noise, its amplitude, and the number of times that it occurs. None of them can tell you the root cause.
Divide and conquer (cutting the pair in half to see which way the noise is coming from) does not work. When testing along the cable route from the central office or remote to the customer’s network interface any increase in amplitude of impulse noise is pointing the technician in the right direction. A common source of impulse noise is in the residence.
Crosstalk is a significant problem, especially far-end crosstalk steals bins at higher frequencies. Crosstalk may have little or no effect on ADSL, but affects ADSL 2+, VDSL 2, and G.fast.
One last note and this is obvious but should be stated anyway: bridged tap and load coils are obvious detriments to bandwidth, and can easily be identified and removed. A TDR will give the distance to them.
In today’s bandwidth world, reactively changing pairs (cut to clear) is a bad decision. It takes a team of well-trained technicians backed up by skilled, informed managers to provide a quality copper infrastructure for bandwidth. Proactive fault locating is the name of the game. Any comments on the topic of quality bandwidth? Please contact me: email@example.com. And start planning for ISE EXPO (formerly OSP EXPO) in Texas in September: http://www.iseexpo.com.