Site noise measurements isolate reception problems Simple equipment configurations and test procedures make it possible to measure receiver site noise

With the ever-growing congestion and development of telecommunications sites, radio frequency (RF) noise levels at sites are higher than ever. The following information describes a method of measuring the effect of noise and desensitization on receivers. The “lossy T” coupler is described, along with a method of measuring its insertion loss.

Site noise sources and symptoms Noise at telecommunications sites comes from a number of sources, man-made and natural. Natural noise is caused by lightning, precipitation static and cosmic sources such as solar flares. Cosmic noise, sometimes referred to as white noise, peaks around 50MHz. This noise is heard as a hissing sound when a receiver is unsquelched and no signal is present. Man-made noise is caused by power lines with arcing insulators, vehicular ignition systems, arcing brushes in electric motors and by other radio equipment. Transmitters produce broadband noise, as do receiver local oscillators in close proximity. All of these factors contribute to the background noise floor at a site. In rural areas with only a few radio systems, the noise floor is much lower, with a large percentage of the noise coming from natural sources. In more densely populated areas, man-made noise accounts for a higher percentage of the noise level. Noise is much worse in industrial areas, near major highways and at crowded telecommunications sites. Measuring the effect of noise on a particular receiver may help to identify problems such as poor sensitivity and may lead to ideas for reducing or eliminating some of the noise. It is beyond the scope of this article to propose solutions to noise problems. Noise manifests itself not only audibly but sometimes much more subtly, and with hidden effects. For example, I once was responsible for a repeater that worked well late at night and on weekends. It was located atop a 20-story downtown building. Its receiver would become noisy and desensitized during the day. The problem turned out to be an elevated noise floor. The site noise level increased during weekday daytime hours when large numbers of paging and other two- way radio systems were at peak use. The technique used to determine that the problem was noise is described below.

‘Lossy T’ coupler Performing noise measurements requires a few pieces of test equipment, a dummy load and a “lossy T.” Figures 1 and 2 below show typical construction for both type N and UHF lossy T connectors. Before performing the noise tests described, first characterize the lossy T. After modifying the N or UHF T connector, connect the T in line with a calibrated signal generator, a dummy load and a receiver. Connect a SINAD meter to the receiver’s audio output. Most service monitors contain the SINAD meter and signal generator in one package. The RF signal generator should be connected with any duplexers and cavities still in line with the receiver at the point where the antenna transmission line connects to the radio equipment. This configuration makes it possible to determine the system sensitivity under ideal conditions with all related site equipment in place and operating. First measure the 12dB SINAD of the receiver with the output of the signal generator connected directly to the receiver antenna input and on the receiver frequency. Note this level. Next disconnect the signal generator and connect the receiver, dummy load, SINAD meter, signal generator and lossy T as shown in Figure 3 to the left. Increase the signal generator level until the SINAD meter reads the same level as previously was noted. The loss of the T-coupler is the difference between the first and second RF levels of the signal generator. For example, let’s say that the receiver reads a 12dB SINAD level of 0.35 microvolts or -116dBm when connected directly to the receiver’s antenna input. Next, with the receiver connected as shown in Figure 3, the level reads 11.32 microvolts or -86dBm. The loss of the T is computed as

20 log 11.23 microvolts /0.35 microvolts = 30dB

or, if using dBm,

-116dBm – (-86dBm) = 30dB

Now that we know the loss of our T coupler at the receiver’s frequency, we can proceed to measure the site noise level.

Effective receiver sensitivity with site noise With the antenna disconnected and a dummy load connected in its place as shown in Figure 3, again note the level required for 12dB SINAD. Reconnect the antenna with the Lossy T in line, and use the test setup shown in Figure 4 above. Note the new level on the RF signal generator required to achieve 12dB SINAD. The noise level degradation at the antenna system can be calculated as follows: (RF level required for 12dB SINAD with antenna connected) – (RF level required for 12dB SINAD with dummy load connected) This calculation yields a decibel value if the RF levels on the signal generator measurements are made in dBm; or in microvolts if the generator’s voltage scale is used. To compute decibels from a microvolt level, use the following formula:

20 log V2/V1 = dB.

An example helps to clarify the procedure. Connecting the receiver as shown in Figure 3 yields a reading of -86dBm on the RF generator for 12dB SINAD. Next, connect the setup as shown in Figure 4. The new level of RF required for 12dB SINAD will be -75dBm, and -75dBm – (-86dBm) = 11dB. Thus, site noise is degrading receiver sensitivity by 11dB. For a mobile installation, the procedure should be performed twice, once with the engine off and the antenna connected, and again with the engine running and the antenna connected. The difference between these readings is the additional noise caused by the vehicle’s ignition and electrical systems. Although not necessary to perform the site noise test, the computations involved in determining the insertion loss of the lossy T can be used to translate relative receiver sensitivity with site noise into an absolute value. Using the values given above, we know that site noise produces 11dB of desense. Knowing also that the loss of the T is 30dB, we can calculate effective receiver sensitivity with the antenna connected. Using the above values shows that the receiver sensitivity with the antenna connected is -105dBm. This amount is computed as follows: (-75dBm sensitivity through lossy T coupler) – (30dB T loss) = -105dBm.

Repeater desense measurements Another benefit of using the lossy T is to compute transmitter-induced noise in a repeater system. Again, with the setup shown in Figure 3, measure the receiver sensitivity with the repeater transmitter disabled. Note this level. Now, while leaving the setup the same, turn on the transmitter. If the system requires continuous tone-coded squelch system (CTCSS) coding, then modulate the RF generator with the appropriate tone code to activate the repeater transmitter, or manually key the transmitter. Note the new RF level required for 12dB SINAD with the transmitter keyed. The difference in readings is the amount of desense caused by the repeater’s transmitter. Ideally, this amount should be less than 2dB.

Conclusion For the cost of a few dollars, a lossy T can be a great asset in troubleshooting and in determining whether you have high levels of RF site noise or desense, whether a duplexer may need retuning or whether a defective power amplifier is generating excessive noise in a repeater. A lossy T also facilitates connections to spectrum analyzers, frequency counters and other test equipment that protect them from high RF levels that would cause damage were they connected directly to a transmitter. Lossy Ts are so inexpensive and easy to make, every technician or field engineer should have one on hand.