Manufacturing and system design engineers tell us to expect a 5 to 6 dB reduction in system coverage performance after converting our analog FM systems to narrowband digital technologies. On the uplink or talk-in side, this would equate to trading your 4-watt portable radio for a 1-watt unit. Probably not the performance you were hoping for from your shiny new state-of-the-art communications system.

Discussion of the actual numbers and reasons for this reduced performance we’ll leave to other articles on the subject. The purpose of this article is to identify areas in which improvements can be made to help recover these losses and result in equal or better performance. It’s likely that a number of dBs of potential improvement are hiding in your existing system. All we need to do is find them and put them back to work.

Receiver Multicoupler

Static system sensitivity is overall sensitivity as measured at the input to the receiver multicoupler or tower-mounted preamplifier, if you have one. Depending on the age of your system and how it was installed and optimized, some improvement to system sensitivity may be available.

Let’s start by assuming you have only a receiver multicoupler with no tower top amplifier. The multicoupler’s amplifier truly represents the input stage of your receiver system and as such establishes the system’s Noise Figure, which, in turn, determines the system’s static sensitivity.

If the multicoupler system is old, the amplifier design may be outdated. Older amplifiers had Noise Figures in the 3–4 dB range while newer state-of-the-art units (see Figure 1 above) can have Noise Figures as low as 1 dB. To take advantage of the low noise amplifier, the system gain has to be set properly. In the past, many multicouplers were set to provide unity gain between the multicoupler input and the receiver. It was thought that the purpose of a receiver multicoupler was to overcome the losses in the power dividers and interconnecting cables, generally called the distribution losses. Although it’s true that one purpose of the multicoupler is to overcome the distribution losses, a properly designed and optimized multicoupler can also improve system performance. If a multicoupler system is set for unity gain or just a few dBs above unity gain, the system sensitivity will be no better than the receiver’s chassis sensitivity, and, in some cases, it will be slightly worse. However, if the multicoupler is set to have about 10 dB of gain, the system sensitivity could be as much as 6 dB better than the receiver’s static or chassis sensitivity.

The mathematics behind this is based on Noise Figure. The typical base station receiver has a Noise Figure of approximately 10 dB. This usually results in a static sensitivity of around -118 dBm for a 5% bit error rate. If an amplifier or multicoupler with a 3 dB Noise Figure (noise factor of 2) and a net gain of 0 dB (gain of 1) is placed ahead of the receiver that has a 10 dB Noise Figure (noise factor of 10), the resulting system Noise Figure will be approximately 10.4 dB. The system sensitivity would actually be degraded by 0.4 dB compared with the receiver’s static sensitivity.

Figure 2 (left) shows the calculations we use to determine the system Noise Figure. Note that the equations are based on noise factor and linear gain not the logarithmic values of Noise Figure and dB gain.

The calculations in Figure 3 (right) show that if we take the same multi­coupler and configure it to have a net gain of 10 dB, with all other parameters the same, the system Noise Figure improves to 4.7 dB. This represents an improvement of 5.3 dB in system static sensitivity.

The graph in Figure 4 (below) shows the relationship between multicoupler net gain and the improvement in receiver sensitivity. The graph is based on a receiver with a 10 dB Noise Figure and a multicoupler amplifier with a 3 dB Noise Figure. If your multicoupler has an amplifier with a lower Noise Figure, the improvement will be greater.

Be careful about taking this too far. You’ll also notice that the curve begins to flatten as we pass 10 dB of net gain, and after 15 dB, there is little to no additional improvement. Running excessive amounts of net gain will only result in an increased probability of interference due to receiver overload. I personally feel that 10 dB is a good target in today’s crowded and noisy RF environment. At locations with a large amount of high-level RF activity, you may have to reduce the gain to avoid receiver intermodulation and other forms of receiver overload.

The net result: If you’re running your multicoupler with little or no gain, you’re not getting all of the available performance.

Transmission Lines

So far, we’ve been looking at the system from the input of the receiver multicoupler. However, the real input to your receiver system is at the base of the receive antenna. The best signal to noise (S/N) ratio in your system is at the output of the receive antenna. The purpose of the rest of the system is to preserve that ratio as much as possible until the receiver can demodulate it into audio or data.

Transmission line losses between the receive antenna and the input to the receiver or multicoupler will reduce the system’s overall static sensitivity. In technical terms, we say: “All losses ahead of the first active stage add directly to the system Noise Figure.”

If you have a long transmission line from the receive antenna to the equipment room, you could be losing several dB of system receive performance.

Table 1 shows the average line loss for two common sizes of transmission line at VHF, UHF, 800 MHz and 900 MHz. The values are based on the manufacturer’s published specifications. The loss at 700 MHz is essentially the same as at 800 MHz. Depending on your frequency band and the configuration of your system, your transmission line loss could be severely degrading your system performance.

In addition to the published specifications, you also have to consider the age and condition of the line. In the never ending battle with water, the water always wins—eventually. If you’re considering reusing your existing lines and antennas with your new narrowband system, having the lines swept with a Site Analyzer would be a good investment. If you’re replacing the lines, you should consider the line loss versus the cost and tower loading of the larger lines. There is, however, a diminishing return issue that occurs with regard to transmission line size. Generally, bigger is better, but only to a point. Once, on a whim, I compared using 4-inch rigid line that the broadcasters use to 1-5/8-inch line at 900 MHz. The cost would have been tremendous, and the improvement small—not to mention that it would have probably overloaded the tower.

Tower Top Amplifiers

If you’re operating at 450 MHz and higher and the distance from your equipment to the receiving antenna is long enough, you might want to consider using a tower top amplifier (TTA). As a basic rule of thumb, if the line loss ahead of your receiver multicoupler is over 2 dB then your system would benefit from a TTA.

When we use a TTA, the input to the receive system is essentially moved to the top of the tower and the transmission line losses are eliminated. From a sensitivity or Noise Figure perspective, if your receive system has a 5 dB Noise Figure as we computed previously, and you have 3 dB of loss in the receive transmission line, your overall system Noise Figure is now 8 dB when computed at the base of the receive antenna. Placing the preamplifier at the top of the tower will bring down the system Noise Figure to 5 dB or less depending on the quality of the amplifier. This represents another 3 dB of system improvement and 7 dB improvement over the receiver’s static sensitivity. Again, if your amplifiers are better than our example, there will be even more improvement (see Figure 5, right).

The Environment

All of the benefits we’ve discussed so far can be negated by noise and interference. We live in a world that’s becoming increasingly noisy, not only to our ears, but also to our radio systems. In the past 30 years, I’ve seen the environmental noise increase to the point at which we now have environmental noise degradation in metro areas on the 150 MHz band—that’s as bad as it was on the 30 MHz band in 1976. The good news: As you go higher in frequency, the levels of environmental noise decrease. But that doesn’t mean that at 450 MHz and above there are no problems.

Any receiver on any band is subject to being degraded by transmitter noise, intermodulation and spurious emissions from other nearby transmitters, as well as co-channel and adjacent channel interference.

You have to live with generic environmental noise. Co-channel and adjacent channel issues usually must be solved by proper frequency coordination and mutual cooperation. Transmitter noise, intermodulation and spurious emissions (sometimes called out-of-band emissions), can often be reduced or eliminated if one is willing to put forth the time and effort to locate the source and work with the owner.

Effective Receiver Sensitivity

Whether you have an aging existing system or a new narrowband digital system, the first step is to determine the system’s Effective Receiver Sensitivity (ERS). This is the system’s static sensitivity plus any degradation due to external noise or interference. Every system should have the ERS checked on every frequency at every infrastructure site at least once a year—more often if you have, or suspect you may have, coverage problems. This will give you a good benchmark of how well your receivers are working and allow you to accurately track any changes. Checking ERS should be as much a part of your preventative maintenance schedule as checking transmitter power output.

Figure 6 (bottom) shows the basic test configuration for performing an ERS test. The test consists of three steps:

1. Measure the system static sensitivity at the multicoupler or tower top amplifier input;
2. Measure the system static sensitivity through an iso-T with the antenna port terminated in a 50 ohm load; and
3. Measure the system static sensitivity through the iso-T with the antenna connected. Be sure to use the same criteria for all three measurements, 12 dB SINAD, 5% bit error rate etc.    

When you connect the antenna in step three, in all probability, you’ll hear or see the sensitivity drop as the external noise is injected into the system. Simply subtract the Step 3 value from the Step 2 value to obtain the degradation. Then subtract the degradation value from the original static sensitivity measured in Step 1.

If you have a tower top amplifier, remember that the input to your receiving system is the TTA input at the top of the tower. Most amplifier systems provide a test port to allow you to inject a signal at the amplifier input and controls allow you to switch between the antenna and a 50 ohm load. See your equipment manuals for specific details or contact your manufacturer.

If you feel the amount of degradation your system is experiencing is excessive, then efforts can be made to track down the source of the problem. If you don’t see any degradation, then I submit you have room to improve your system sensitivity through lower nose amplifiers, adding a TTA, adjusting the net gain or changing the transmission lines—to name just a few possibilities. When I optimize a system, I like to see a good system static sensitivity and 1–2 dB of external noise degradation when doing the ERS test. These readings tell me that my system static sensitivity is as good as it needs to be. For any further improvement, we would have to go outside the system and clean up the environment.

Interference Sources

There are many potential sources of interference, and as the world becomes more wireless, the list grows longer every day. Interference is a growth industry. Most interference can be grouped into a few categories: interference created in or by transmitters, interference created in or by receivers and external interference.

Transmitter-based interference consists of transmitter noise, spurious emissions, and transmitter intermodulation. Receiver interference consists of receiver desensitization, spurious responses, images, receiver blocking and receiver intermodulation. External interference consists of co-channel and adjacent channel interference, as well as external intermodulation. One of the worst sources of external interference is external intermodulation. This is a form of passive intermodulation involving the mixing of signals on the tower or rooftop external from the system’s equipment and antennas. RF energy can mix in almost any non-linear environment, including, but not limited to, dissimilar metals; consumer electronics, such as TV antennas and preamps; poor electrical connections; loose fittings and connectors; and even within certain metals.

It was somewhat of a surprise a few years back to realize that “rusty bolt” intermodulation—intermodulation associated with old rusty towers—mounts and other site hardware isn’t necessarily caused by the rust—although I’m sure the rust makes the problem worse. The real problem is ferrous metal. Ferrous metal is any metal a magnet will stick to. Without getting too technical, we’re dealing with a phenomenon called “magnetic hysteresis.” The short story is that when RF flows in a ferrous metal, the magnetic field is distorted, and distortion is the primary building block of intermodulation.

The solution is to plate the ferrous metal with something conductive, but not magnetic. The ideal candidate is zinc, which is galvanizing. All hardware at an antenna site should be non-magnetic, such as stainless steel or galvanized steel. Sorry people, painting won’t help. On a rooftop, anything that’s painted steel could be a source of intermodulation if the RF fields are strong enough. This includes air-conditioning equipment, solar panels, steel doors and door frames, electrical boxes and conduit. In one case, it was the building itself. There was steel rebar in the structure, and the building was creating and radiating a strong intermodulation product on the customer’s frequency.

Solutions

Many interference problems can be resolved. But we must remember the following axiom: “All interference must be resolved at the source.”Example: On-channel interference, such as co-channel or spurious signals from another transmitter, including transmitter intermodulation, cannot be resolved by filtering at the receiver. On-channel energy is just that, energy on your channel. You can’t filter it out without also filtering out the desired signals (your subscribers).

On the other hand, problems that occur in the receiver must be solved at the receiver. Such things as receiver overload, receiver images and receiver intermodulation can’t be resolved by filtering at the transmitter. The best place to start is a simple test that you can do in conjunction with the ERS test. Key all your system’s transmitters, and measure the system ERS on each channel. This will immediately alert you to any intra-system interference, such as passive intermodulation, transmitter noise or receiver desensitization, you may have. After you know the source, action can be taken to reduce or eliminate the problems. Such problems as transmitter noise, spurious or out-of-band emissions or transmitter intermodulation can usually be helped by installing band pass cavities and isolators on the offending transmitter.

Problems at the receiver, such as desensitization, intermodulation, images and other spurious responses, can often be resolved by installing additional filtering on the receiver input. Receiver multicouplers and TTAs usually will have filters that pass some or all of the receive band. Normally, this doesn’t create problems for the TTA or the receiver multicoupler, but high-level signals can get through to the receiver and cause desensitization or receiver intermodulation. This can be resolved by narrowing the amount of spectrum reaching the receiver. Either install a narrower filter just ahead of the receiver multicoupler or a band pass cavity or crystal filter between the receiver multicoupler and the receiver. It’s rarely necessary to change the filter in the TTA. The nice thing about this type of downstream filtering is that the insertion loss of the additional filter doesn’t affect the system sensitivity due to the net gain of the TTA and multicoupler.

Some basic guidelines to prevent or resolve interference: 1) All transmitters at multi-user sites must have a minimum of a band pass cavity and a dual stage isolator. 2) Multicouplers and TTAs should be adjusted for approximately 10 dB of net gain, but keep an eye out for high-level signals that could reach the receiver and cause intermodulation or overload. Reduce the gain if necessary. 3) Properly size your receiver multicoupler preselectors, especially for VHF and UHF. The less spectrum that reaches the receiver, the less probability of interference. Use filtering between the multicoupler and the receiver if necessary. 4) Never put frequencies into a transmitter combiner that can produce third order intermodulation on a site receiver’s frequency. 5) Try to avoid putting any frequencies at a site that can produce harmful 3rd, 5th or 7th order intermodulation. The lower the order of the intermodulation, the stronger the resulting product will be and the harder it will be to eliminate. 6) If you have to accommodate known intermodulation products on site, provide as much antenna isolation between the transmitters and the potential victim receiver as possible. 7) Always use galvanized or non-ferrous metals for all site hardware.

Summary

“Plan for the worst, hope for the best.” I’ve always thought that was good engineering advice. In hoping for the best, we have our optimization of the receiver sensitivity. We should design, install and optimize the receive system to provide the best possible performance the hardware is capable of achieving. In planning for the worst, we should test for and resolve any interference to the maximum degree possible, understanding that we can’t achieve perfection.

If you follow these concepts you’ll find enough dBs hiding in your system to offset any degradation due to narrowbanding.

Alfred T. Yerger II is an RF engineering specialist for Bird Technologies Group. Contact him via e-mail at [email protected].

Originally published in Public Safety Communications, 76(1):29-32, January 2010.