Power Meter Parameters

Many of today's wireless communications systems are based on the premise of frequency reuse throughout the system. The signal transmitted from a base station in a particular cell must cover the entire geographic area inside that cell (or cell sector). However, that signal must not interfere with another signal at the same frequency used in a cell across town. A signal set too high will travel beyond the cell's boundary; a signal set too low will not be able to provide the desired coverage. Base-station-transmitter power levels are determined by the system engineers, who take into account the antenna system, the topography and the radius of the cell to be covered.

This article focuses on measuring the power of a base-station transmitter, which assumes high-power signals. Making accurate low-power measurements is more difficult and involves more sophisticated test equipment.

Necessary Accuracy Base-station-transmitter power levels typically have a total accuracy requirement of +/-10%. This has driven many infrastructure vendors and carriers to demand test equipment that can measure power with an accuracy of +/-5%. Test equipment typically is assigned an accuracy requirement that is 10 times better than the requirements for the particular parameter being measured. This is done so test equipment performance does not contribute significant error to the measurement.

Using that rule of thumb for power meters would yield a +/-1% measurement accuracy requirement, which is almost unattainable outside of a measurement standards laboratory. A more practical value is +/-5%, which may not sound like a tough measurement to make until you think in terms of decibels and consider the variables involved.

A decibel (dB) expresses a ratio logarithmically and is helpful when dealing with power measurements and losses in an RF system because it is easy to trace power levels as the signal flows through the system. To convert an accuracy range from percent to decibels, first apply the percentage to a nominal power value to determine the possible min and max values. Next, convert those min and max values to dB (using 10logX), and subtract them. After you go through the math, it turns out that a +/-5% power measurement accuracy corresponds to approximately +/-0.2dB. Never directly apply a percentage to a power value already expressed in dBm as this has no meaning and will yield the wrong result.

Sweat the Details What good is it to have a +/-0.2dB power meter if the inaccuracy in your test setup exceeds it? When making an accurate power measurement, pay attention to the details. Ensure that all components in the setup have been calibrated recently. It is not uncommon for a directional coupler to have +/-0.5dB of level error, and most attenuators are accurate only to about +/-0.3dB. If your setup uses a few of these components, the resulting error only gets larger. Use good-quality test cables and connectors, and check them daily. Minimize the use of adapters. Even high-quality adapters can add up to 0.1dB of loss.

Connect your power meter as close to the antenna as possible. When the system design calls for a particular transmit power, it assumes that that power is being delivered to the antenna. If you have hundreds of feet of feeder cable between the point of measurement and the actual antenna, your quest for a 5% power setting is all but lost.

Different Tools for Different Jobs * Thermal power meters operate by measuring the temperature rise in a resistive element generated by the application of the RF signal under test. They measure true average power and are accurate on almost any type of signal. Because of their slow response rate, they are best used on continuous transmission signals. They can be used on a burst signal (like that in a TDMA system) if the duty cycle is high enough with respect to the response rate of the meter and the appropriate calibration factor is applied to the result.

Multicarrier RF-power measurements are not a problem for a thermal meter, whose rms-power-measurement capability is independent of the number of signals applied -- as long as the total power remains within the measurement range of the meter. Thermal power meters are broadband and will measure the power of the signal you apply as well as any other interfering signal (either in-band or out-of -band) that may be present.

To determine the total measurement accuracy, you may need to dig through the manufacturer's specifications to find all possible sources of error. For example, some thermal-power measurement heads have a +/-1% accuracy specification. However, the total measurement accuracy of the entire meter may not be that good. In the field, the power head must be calibrated to a signal source provided on the meter. That source has an accuracy specification attached to it.

Next, the meter itself must be able to measure and properly display power based on the signal it receives from the measurement head. Another accuracy specification is attached to that. These and other things, such as zeroing the head and VSWR considerations, mean that many of the thermal power meters have total measurement accuracies in the 5% to 6% range in the field.

One drawback to a thermal power meter is that the power heads are fragile. Many power heads have been damaged after being dropped from the top of an equipment rack.

Another drawback is the limited measurement range of the thermal power meter, which means they cannot measure transmitter power directly and must instead be connected to the power amp via a directional coupler. This test setup can introduce considerable measurement error if not properly calibrated.

* Peak-detecting power meters are common in service monitors. They essentially use a diode and a capacitor to detect the envelope of the applied RF signal. The voltage of this detected signal is then measured. Based on its amplitude, an appropriate calibration/scaling factor is used to yield an rms power measurement.

Peak-detecting power meters can be accurate (5% to 6%), provided they are used on the types of signals for which they were designed. Most were designed for measuring a single narrowband FM RF signal. As with the thermal power meter, a peak-detecting power meter is not frequency-selective and can be influenced by other signals present if they are strong enough. Unlike the thermal unit, a peak-detecting power meter has a faster update rate and does not perform well in the presence of multiple RF carriers and in some types of digital systems.

* Similar to a peak-detecting power meter, a true rms square-law power meter uses a diode detector as its measurement element. In this case, the diode is operated in its square-law region, which yields a true rms power reading. To operate the diode in that square-law region, the signal levels at the detector must be carefully controlled, and the surrounding support circuitry is more complicated. Square-law power meters include the many benefits of a thermal power meter (accuracy, multiple carrier capability and digital waveform measurements) along with the measurement speed of a peak-detecting power meter. These meters also tend to have a wider dynamic range.

* A spectrum analyzer can be a useful tool for measuring power, particularly in the case where multiple signals are present, and you want to measure the power of only one of them. In a live-site test environment, the signal being measured may be only one of several that are passing through the power amp. You must have a frequency-selective, power-measurement capability to make the measurement.

You should consider several things when using a spectrum analyzer to make power measurements. The first is accuracy. The best absolute power measurement accuracy that can be obtained in the field with a spectrum analyzer is about +/-1dB to 1.5dB (a 41% error). Spectrum analyzers have excellent relative power-measurement capabilities, but their absolute accuracy is not as good as that of a power meter.

Power-handling capability is another thing to consider. Many spectrum analyzers cannot handle several watts of power without damage. This means that a coupler/attenuator setup is required, which can add further inaccuracy if not properly calibrated.

Measuring power is an important job for the field engineer because it ensures that today's wireless communications systems provide the service the industry has come to expect. Several tools are available; however, none is optimal for all tasks. The challenge is to use the right tool for the particular job at hand.