Basic transmission impairments
Communications field technicians often maintain equipment that interfaces with a multiplex (mux) circuit or Telephone Company (TELCO) leased line. When circuit-related problems arise, the first step toward trouble resolution is a basic understanding of transmission impairments.
Transmission impairments are typically placed in one of two categories: steady-state (which are constant in nature), and transient. Since transient impairments are intermittent, instantaneous phenomena, associated tests register an accumulation of impairing incidents or hits occurring over a specified period of time.
The most common steady-state impairment measurements include:
Level, Attenuation Distortion and Slope
Noise and Noise Related Tests
Envelope Delay Distortion (EDD)
Return Loss and Related Measurements
Peak-to-Average Ratio (P/AR)
Intermodulation Distortion (IMD)
Transient impairment measurements include:
Circuit testing can be accomplished with a Transmission Impairment Measuring Set (TIMS). This device can simultaneously send and receive test signals used to measure circuit impairments. It can also perform non-disruptive monitoring of live traffic. In addition to understanding how to operate the test equipment, the technician should know the circuit specifications offered by the service provider for each parameter under test.
The TIMS is most often connected to the circuit at the service provider's point of demarcation. This is a specific physical connection point in the circuit that defines where the service provider's equipment (and consequently, their responsibility) stops and where the customer's starts. With a TELCO lease, this point is usually the “customer” connections on a device referred to as a Network Interface Unit (NIU). The NIU serves as an interface between the lease and the customer equipment, and often allows adjustment of gain, loss, equalization, impedance and operation of a remote loopback function.
Most impairment tests are accomplished with a technician on each end of the circuit. The “customer” side of the NIU is connected to the “send” and “receive” ports of the TIMS. This allows transmission of a test signal into the NIU (and consequently into the lease) while receiving and evaluating the signal coming out. When the NIU offers a remote loopback function, a single technician on one end of the lease can remotely loop back the other end (by sending a momentary 2713Hz tone) to accomplish round-trip measurements (Applying the same momentary 2713Hz tone will drop the loop).
However, loopback tests can lead to ambiguities with regard to which circuit direction has the problem.
Many impairment tests are accomplished by transmission of a test or holding tone. When the tone is received at the far-end of the circuit, it is utilized to evaluate the parameter of interest.
Originally, a tone frequency of 1000Hz was used. This was later changed to 1004Hz to avoid problems resulting from the tone having a direct sub-harmonic relationship with the 8kHz-sampling rate used in PCM systems. A holding tone frequency within the 1002Hz to 1022Hz range is considered acceptable, with 1020Hz (+2Hz, -7Hz) recommended (IEEE 743-1995).
Most manufacturers of TIMS equipment for North America are still using 1004Hz for impairment tests.
Level, Attenuation Distortion and Slope
Measuring circuit signal levels is probably the most common and basic of impairment tests. It can uncover a range of problems: from poor wiring connections, to misalignment or malfunction of circuit components. To accomplish this test, the technicians involved must first know the proper levels to send as well as what they should receive. The technician on each end of the lease transmits a holding tone while measuring the level from the other end. Results are expressed either in Decibels “above” (+dB) or “below” (-dB) the expected level, or as an absolute level in dBm.
Attenuation distortion is a parameter that describes the variation in attenuation between frequencies under test and that of a reference (usually the holding tone). Plotting the changes as the tone frequency is varied across the bandpass of the circuit produces a curve. This curve represents the frequency response of the circuit. Tones near the upper and lower limits of the bandpass usually experience more attenuation. This can be caused by many contributors such as loaded cables acting as a lowpass filters, transformers and series capacitors as high pass filters, circuit bandpass filtering and other influences of inductive and capacitive reactance.
The Three-Tone Slope (gain slope) test evaluates the attenuation characteristics of a circuit at 404Hz and 2804Hz relative to that of a 1004Hz reference. All three tones are transmitted at the same level. Because of variations in circuit frequency response, the recovered levels of the 404Hz and 2804Hz tones will be different than that of the 1004Hz tone. As with attenuation distortion, test results are expressed using the convention of +dB meaning more attenuation than the reference, and -dB meaning less.
Frequency translation error is primarily an issue relating to oscillator drift or misalignment of translation carrier generation equipment in older analog mux (FDM) systems. Although advances in analog circuit design and migration to digital transport systems have all but eliminated this problem, the frequency of the recovered holding tone should be checked to verify it is the same as that being sent.
Noise level measurements are often expressed using the term dBrn (Decibels above reference noise). This term is a measure of power expressed in dB, referenced to -90dBm (-90dBm = 0dBrn). Since noise levels greater than -90dBm are of significance, the term dBrn offers a means of expressing power levels using strictly positive numbers.
Many network measurements, particularly those relating to noise, are accomplished in conjunction with specific filters incorporated within the TIMS. Each filter has a unique attenuation characteristic, or weighting.
Measurements are accomplished using the filter that best suits the impairment test and circuit application. For example, noise measurements in a voice circuit are often accomplished with a C filter to best quantify what the human ear would hear. However, since a C filter attenuates frequencies below 300Hz, a 3kHz flat or 3.4kHz flat filter would be a better choice for evaluating low frequency hum. A D filter offers flat response between 300Hz and 3.4kHz for hum rejection and measurement of voice-band data signals. 15kHz flat, E, F, and G filters also exist, each for impairment measurement of specific analog and digital applications.
Message Circuit Noise is a parameter that describes the power of idle circuit noise through the local loop. For this test, the TIMS does not transmit a holding tone. Instead, it is put into a quiet termination (quiet term) mode that places a resistive termination across the “send” port. The TIMS on the opposite-end of the circuit measures the power level of idle circuit noise through the appropriate weighting (filter). Voice circuits typically use a C filter offering results in dBrnC, while voice-band data applications might use a D filter (dBrnD) or a 3.4kHz flat filter (dBrn).
To ensure noise measurements on digital carriers (PCM systems) are valid, a holding tone must be transmitted to operate the compander and quantizing circuitry in its normal operating range. Since no holding tone is transmitted during the message circuit noise measurement, it is only valid for the local loop.
The Noise with Tone, or Notched Noise test measures circuit noise, including noise and distortion resulting from the application of a holding tone. The receiving TIMS utilizes an internal notch filter to remove the holding tone and measure the remaining circuit noise and distortion through an appropriately weighted filter. Tone-induced byproducts may involve quantizing noise, harmonic distortion and any amplitude or phase modulation sidebands falling outside the reject limits of the notch filter.
The Signal-to-Noise ratio (S/N) is expressed in dB and is the ratio of power levels between the holding tone and the appropriately weighted, notched noise.
When both levels are expressed in the same dB-relative term (such as dBrnC), it is simply the difference between the two levels, expressed in dB. The larger the S/N ratio, the quieter the circuit.
The Noise-to-Ground test measures longitudinal or common-mode noise. This type of noise is present on both conductors of a two-wire pair referenced to ground, and is caused by poor power grounding techniques or ground loops in the facility power system. Measurements are accomplished by connecting the tip and ring of the circuit together inside the TIMS, then measuring the noise level, referenced to ground.
Results can be displayed in a dB-relative unit of measure such as dBrn or dBm.
Jeff Ashley is a communications technician and writer based in Ventura, Calif. This is the first of two parts — to be concluded next month.