How to Measure RF Power: Power Meters, Analyzers and Losses

What dBm Means

RF power is quoted in dBm, a logarithmic unit referenced to 1 mW. The conversion is:

So 0 dBm is 1 mW, 30 dBm is 1 W and 40 dBm is 10 W. Every +10 dB is ten times the power, and every +3 dB is about double. The log scale lets one number range from a 10 W transmitter to a very weak signal without a long string of zeros.

Gains and losses are plain dB, and they add. Adding 10 dB to a level in dBm gives a new level in dBm. Subtracting two dBm values gives a ratio in dB. Adding two dBm values together is meaningless, so convert both to milliwatts first. The dBm to watts converter does the conversion when you need watts.

Power Meter or Spectrum Analyzer

A power meter with a sensor measures the average power at the sensor input. It returns one number for everything that reaches the sensor. That makes it the natural choice when you need the total power of one transmitter on a clean connection and you want the best accuracy.

A spectrum analyzer shows power versus frequency. It can read one signal among others: a carrier beside a spur, or one channel in a busy band. The trade-off is accuracy. Its reading is typically less accurate than a power meter reading.

Choose by the question you are asking:

• How much total power does this one source deliver? Use the power meter.
• How much power is at this one frequency, with other signals present? Use the analyzer.
• Where does the power sit in frequency? Only the analyzer can show that.

State which instrument produced a number whenever you report it. A value from an analyzer and a value from a sensor are not equal-quality evidence, even when they agree to a few tenths of a dB.

Why an Attenuator Goes in Front

Every instrument has a specified maximum input level. Exceeding it can damage the sensor or the analyzer front end. Check that limit, and add an attenuator if needed, before you connect a transmitter. Do this first, not after you see a strange reading.

An attenuator lowers the level by a fixed number of dB. The level at the instrument is:

where A is the attenuation in dB. With 30 dB of attenuation, +40 dBm (10 W) becomes +10 dBm (10 mW). That is a level a typical instrument input can accept, though you must still check the actual limit of yours.

The attenuator also has to be accounted for afterward. It is part of the path, and its loss is not part of the source power. If you forget it, the reported power is too low by exactly the attenuation. To choose or build a pad, use the attenuator calculator.

Adding Back Cable and Attenuator Loss

The instrument reads the power at its own input. To get the source power, add back every loss between the source and that input:

All terms are in dB or dBm, so this is simple addition.

Worked example. The meter reads -12.3 dBm. The path has a 10 dB attenuator and 1.5 dB of cable loss.

1. Start with the reading: -12.3 dBm.
2. Add the attenuator: -12.3 + 10 = -2.3 dBm.
3. Add the cable: -2.3 + 1.5 = -0.8 dBm.

The source power is -0.8 dBm. In milliwatts that is 10^(-0.08) mW, about 0.83 mW, which is slightly below 1 mW.

The sign matters. Losses are subtracted on the way to the instrument, so you add them going back. Write the loss of each piece in a table before you start, and measure the cable loss at the test frequency, since cable loss changes with frequency.

What Mismatch Does to the Reading

The source and the meter are never perfectly matched. Part of the wave is reflected at each end and bounces between them. Depending on the phase, the reflections add to or subtract from the power the meter sees. The reading is therefore uncertain, even if the meter itself is perfect.

The reflection coefficient magnitude follows from the VSWR:

With a VSWR of 1.2 at both ends, each reflection coefficient magnitude is 0.2 / 2.2 = 0.091. The worst-case error in dB is:

The product is 0.091 × 0.091 = 0.0083, so the power can be off by up to about ±0.07 dB. That is small here. A poorer match makes it grow quickly, because the product of the two coefficients grows with both of them.

You cannot correct this error with a single number, since the phase is unknown. You can only bound it. Include it in your uncertainty, and improve the match if the bound is too large. The VSWR, return loss and mismatch loss calculator is not needed for this arithmetic, because the two formulas above are enough.

Peak and Average Power

A modulated signal has a peak power above its average power by the peak-to-average ratio:

with all terms in dB or dBm. A power meter with a sensor gives the average power at its input. Say which one your number is. A peak value and an average value of the same signal differ by the PAR, and mixing them is a common source of arguments between test results.

A short checklist before you connect anything:

• Find the instrument's maximum input level.
• Estimate the highest power the source can produce.
• Add an attenuator if the source could exceed the limit.
• Record the attenuator and cable loss in dB.
• Add the losses back to the reading.
• Note the mismatch bound and whether the value is average or peak.

In plain words · for students and beginners

RF power measurement in plain words

An everyday picture

Think of a garden hose and a bucket. The hose is the cable, the tap is the transmitter and the bucket is the meter. The bucket catches less water than the tap releases if the hose leaks. To know what the tap really gave, you add back what leaked. Cable and attenuator loss work the same way: the meter sees what is left, and you add the loss back.

The picture breaks down in one place. A leak only removes water. In RF, a poor match at the ends can also push a little power back and forth, so the meter may read a bit high or a bit low. Water in a hose does not do that.

The same example in simple steps

A meter reads -12.3 dBm. You used a 10 dB attenuator and a cable that loses 1.5 dB.

1. Write down the reading: -12.3 dBm.
2. The attenuator took away 10 dB, so add 10. That gives -2.3 dBm.
3. The cable took away 1.5 dB, so add 1.5. That gives -0.8 dBm.
4. The source power is -0.8 dBm, a little under 1 mW.

dBm means "decibels compared with 1 milliwatt". A decibel is a way to say how many times bigger or smaller a power is, using addition instead of multiplication.

Mistakes beginners make

• Connecting first, checking later. Every instrument has a maximum input level. A transmitter can be far above it. Read the limit and add an attenuator before you connect.
• Forgetting the attenuator. The pad protects the instrument, but it also lowers the reading. If you do not add its loss back, your answer is too low.
• Subtracting the loss instead of adding it. The meter sees less than the source gave. You go back up, so you add.
• Adding dBm values. You can add dB to dBm. You cannot add two dBm values and get a sensible answer.
• Mixing peak and average. A signal that changes in time has a peak power and an average power. They are not the same number. Always say which one you measured.
• Trusting the analyzer like a power meter. A spectrum analyzer is very useful, but its reading is typically less accurate than a power meter reading.

Remember

• dBm is power compared with 1 mW: 0 dBm is 1 mW, 30 dBm is 1 W and every +10 dB is ten times the power.
• Check the maximum input level first, use an attenuator, then add its loss and the cable loss back to the reading.
• Mismatch makes the reading uncertain (about ±0.07 dB at a VSWR of 1.2 at both ends), and you must say whether a number is peak or average.

Related pages on RF View

• dBm to Watts, Volts and dBµV Converter
• Attenuator Calculator: T-Pad and Pi-Pad Resistor Values