Overview

Passive intermodulation (PIM) has become a hot topic within wireless carriers. While the PIM phenomenon has been known and studied for many decades, wireless carriers have traditionally limited their PIM testing to labs. The more recent introduction of portable test equipment targeted at resolving PIM issues in the field has extended the topic to a much broader audience, including thousands of field operations personnel tasked with maintaining wireless networks.

There are many practical reasons, beyond equipment availability, for an increased interest in PIM and PIM field testing. PIM problems tend to increase as RF transmission hardware ages. Temperature cycling, corrosion, and vibration all negatively impact PIM performance. Carrier networks are aging and the ability to find PIM problems related to this aging is important.

Installation, hardware quality, and hardware design also play an important role in determining PIM performance. The continual roll-out of new technologies and hardware into wireless networks makes PIM testing in the field a perfect candidate for ensuring a high quality of service.

Finally, PIM problems increase as power increases. The increase in popularity of mobile devices requires wireless service providers to use additional power to serve their increasing user base.

This application note endeavors to provide the reader with the fundamentals necessary to understand the causes of PIM and testing necessary to resolve PIM issues in the field. Technical complexities not fundamental to a basic understanding have been avoided in favor of a more practical approach. Readers who desire a more technical treatment of the topic should download the Forward Link Understanding PIM 2.0 application note.

Intermodulation Distortion (IMD)

An Introduction

Passive Intermodulation (PIM) is a special case of intermodulation distortion (IMD), so a discussion of PIM necessarily includes a discussion of IMD.

Technically speaking, IMD occurs when two or more signals at different frequencies are combined within a system that exhibits non-linear behavior. This non-linear behavior distorts the signal, producing signal components not present in the input.

In communications systems, the faithful reproduction of a signal is especially important. In the particular case of a wireless base station, distortion of the output signal results in more dropped calls, failed access attempts, and reduced capacity. To put it more succinctly, unhappy customers and reduced revenue.

Let us assume the input to an amplifier is two signals at different frequencies. Ideally, the amplifier would produce the same two signals, albeit larger, on the output. In this mythical ideal amplifier, the relationship between the output and input would be called linear (i.e. the output is a larger exact replica of the input).

Unfortunately there are no ideal amplifiers and some non-linear behavior will exist. The resulting output, due to IMD, is the presence of additional signals at frequencies other than that of the two input signals.

There are a few things worth noting about the resultant distortion signals. First, the distortion signals (F₃ & F₄) in figure 1 (below) are often called intermods or, in our specific example, 3rd order intermods. While other intermods and harmonics (2nd, 4th, 5th... order) may also be present, the 3rd order intermod is the focus of PIM field testing.

Second, the intermods (F₃ & F₄) are positioned above and below the carriers (F₁ & F₂) at a distance equal to the difference in frequency between the carriers. The frequency of F₄ is the frequency of F₂ plus the difference between F₂ and F₁. Mathematically speaking:

F₄ = F₂ + (F₂ - F₁) = 2F₂ - F₁

Likewise, F₃ is below the carriers the same distance, or:

F₃ = F₁ - (F₂ - F₁) = 2F₁ - F₂

Third, the intermod's magnitude is dependent upon the non-linearity of the amplifier. The greater the non-linearity, the greater the amplitude of the intermods.

Finally, the distortion's magnitude is also dependent on the magnitude of the carriers. Increasing the magnitude of F₁ and/or F₂ increases the magnitude of F₃ and F₄. Stating an intermod's amplitude or power level is meaningless unless the power level of the carriers is also known. A system can be linear at one carrier level and non-linear at another.

For the purpose of this application note, it is sufficient to remember four key pieces of information:

• Intermods appear above and below the main signals (or carriers) at known intervals (2F₂-F₁ & 2F₁-F₂).
• The presence of intermods indicate a nonlinearity in a system.
• The magnitude of the intermod is a measure of the system's linearity.
• The intermod level is dependent on carrier power level. Knowing the carrier power level for a given intermod power level is important.

Passive Intermodulation (PIM)

Definition and Measurement

The prior section's example utilized an amplifier as the non-linear device responsible for producing intermod distortion signals. The amplifier, because it is externally powered, is considered an active source of intermodulation. Passive intermodulation (PIM) is distinguished from normal intermodulation because the non-linear device has no external power supply. Power for the distortion signals produced by PIM comes from the input RF signal itself.

The sources of PIM are varied and certain sources are even exploited to produce a useful system features. For wireless carriers, however, PIM is bad news because it indicates a non-linearity in the antenna/feed-line system that produces unwanted out-of-band signals. Some of the more common sources of PIM are:

• Corrosion in connectors or antennas due to aging or water intrusion
• Dirt or debris in the RF path
• Incorrect connector installation, including overand under-torquing
• Metallic corrosion near antennas like rusty vents, metal fences, or metal framework
• Mechanical failures in the RF path (broken solder joints...)

Although PIM intermod signals are usually low in power relative to the transmit power, they will wreak havoc in a sensitive receiver. With modern receivers capable of demodulating signals below -100 dBm (0.0000000000001 Watts), it takes little power to interfere with a weak mobile signal.

The concept behind PIM testing is relatively straightforward. Two constant frequency carriers are transmitted into the feed-line/antenna system and the power level at the location of the lower intermod's frequency (2F₁-F₂) is measured. If the system is operating in a non-linear fashion, intermods will be generated and their presence reflected in the power measurement. For field testing, the carrier power level is usually set to 43 dBm (20 Watts). A dBm to Watts conversion table is listed here for reference.

The PIM test result is a power measurement of the intermod's magnitude. It is expressed as an absolute power or a power relative to the carrier's power. Relative measurements are expressed in dBc (decibels relative to the carrier). In either case, knowing the carrier power level is important.

Using the example illustrated by figure 2 will help clarify any questions thus far. The test utilizes two carriers set at 43 dBm. The lower intermod produced by these carriers will appear at the following frequency:

2F₁ - F₂ = 2(869) - 891.5 = 846.5

The power level of the lower intermod is -80 dBm or -123 dBc.

Due to the non-linear relationship between carriers and intermods, changes in carrier power are reflected in intermod power at an approximate ratio of 1 dB to 2.5 dB. If the carriers from figure 2 were lowered 1 dB (43 dBm to 42 dBm), the intermod would drop in power approximately 2.5 dB (from -80 dBm to -82.5 dBm). Although the 1:2.5 ratio will vary based on the mechanism causing PIM, there are a number of important implications:

• Carrier power needs to be set accurately to prevent dramatic changes in PIM levels.
• PIM test equipment needs to be checked frequently for accurate power levels.
• Low power PIM testers may not adequately stimulate defects.

PIM Field Testing

Practical Applications

PIM Levels

Determining the PIM level at which repairs must be made depends on a number of factors. The first consideration is that PIM levels near -100 dBm at the receiver will begin to compete with cell phones. Assuming carriers of 43 dBm, the threshold becomes -143 dBm ( -(43+100) ). The threshold should be adjusted for base stations normally transmitting more or less than 43 dBm. New installations should perform at this level.

PIM performance will degrade with time, so setting a secondary level for aging base stations make sense. Because of base station receive diversity, PIM problems on a single receive can be tolerated for approximately 10 dB to 15 dB beyond the -143 dBm threshold (-133 dBm to -128 dBm). Beyond this point, the base station begins to lose receive diversity and call quality suffers.

Arcing and PIM

A common PIM field failure in hardware is not PIM at all, it is arcing. PIM testers are often used to stimulate and detect arcing, so it is commonly lumped into the "PIM failure" category.

Figure 3 (below) shows how arcing impacts the receive channel of a base station. Arching produces a wide-band noise signal that covers a much broader band than intermodulation. Many service providers do not transmit at frequencies that are capable of producing intermods that fall into their own receive bands. Arcing, because it is wide-band, raises the whole receive noise floor. Viewing the receive signal with a spectrum analyzer will help determine if the problem is due to arcing. True intermod problems can sometimes be ignored if they do not impact the receiver. In contrast, arcing generally requires expensive hardware repairs.

PIM Power Level Considerations

Intermod power levels drop at a rate of approximately 2.5 dB for every 1 dB drop in carrier power. Inaccuracy in carrier power levels make PIM problems appear much better or worse than they actually are.

Intermod sensitivity to carrier power level issues are complicated by PIM field test equipment that does not perform internal power calibration. These units, in common use today, vary their output power level with changes in ambient temperature and heating due to their transmitters. The power level of these units must be set frequently and checked regularly with power meters to ensure transmitter accuracy.

A final consideration with regard to power level is feed-line/antenna system loss and return loss. A bad antenna at the end of a run with 3 dB of loss would see a carrier power level of 40 dBm (rather than 43 dBm). The resulting PIM generated would be reduced at the source (the antenna) by approximately 7.5 dBm (1:2.5 ratio or 3 x 2.5). Adding another 3 dB of loss to the PIM signal as it travels back down the feedline results in a reduction of 10.5 dBm. Return loss issues need to be resolved prior to PIM testing. Higher power testing should also be considered for long runs and systems with more inherent loss.

Spectrum Analyzers

The spectrum analyzer is a critical tool for PIM testing and troubleshooting. Because PIM testers simply measure the power level at the lower intermod's frequency, any residual RF power not related to PIM will also be measured. For example, external interference at the intermod frequency will cause false PIM failures. Certain PIM testers will show residual RF, however, they will not provide any indication of the type of problem or provide any tools for troubleshooting it.

Most modern base stations have coupled ports that allow receive channels to be evaluated prior to performing PIM troubleshooting. A spectrum analyzer should be used to validate the presence of a PIM related issue prior to performing expensive repairs. A spectrum analyzer should also be used afterward to validate the repair.