How to Add Two Signal Powers Given in dBm

Convert each dBm level to milliwatts, add them and convert back, so 0 dBm plus -10 dBm is 1.1 mW, or 0.41 dBm. Two equal uncorrelated powers sum to 3.01 dB above either one, and a 20 dB gap adds only 0.04 dB.

Engineer’s note: Add two signals of different size, say +10 dBm and -10 dBm, and the small one barely moves the big one: the result is still about +10 dBm. A spreadsheet gives the exact number, but when the two differ by 10 dB or more, just call it the larger power (it adds at most 0.4 dB). That rule of thumb speeds up decisions.

For practitioners

You often need the total of two signal levels, and both are usually given in dBm. Examples are a wanted carrier sitting on top of a noise floor, two tones going into a mixer, or two interferers landing in the same receiver channel. The answer is not the sum of the two dBm numbers. This page gives the exact method, a worked example, and a correction table you can use without a calculator.

Why dBm values cannot be added directly

dBm is a logarithm of power referenced to 1 mW: P (dBm) = 10 log10(P / 1 mW). Adding two logarithms gives the logarithm of a product. If you write 0 dBm + (−10 dBm) = −10 dBm, you have multiplied 1 mW by 0.1 mW in ratio terms. That number has no physical meaning as a total power.

Decibel addition is the right tool when one quantity is a gain or loss. A −10 dB pad applied to a 0 dBm signal gives −10 dBm. Combining two powers is a different operation. Power adds in the linear domain, so you have to leave the log scale, add, and come back.

The method for uncorrelated signals

Two signals are uncorrelated when they have no fixed phase relationship over the measurement time. Thermal noise from two sources is the classic case. Tones at different frequencies count too, because their relative phase keeps rotating. The cross term between them averages to zero, so the average powers add directly.

The procedure has three steps:

  1. Convert each level to milliwatts:
PmW=10PdBm/10P_{\text{mW}} = 10^{P_{\text{dBm}}/10}
  1. Add the milliwatt values:
Ptotal=P1+P2P_{\text{total}} = P_1 + P_2
  1. Convert the total back to dBm:
Ptotal, dBm=10log⁡10(Ptotal1 mW)P_{\text{total, dBm}} = 10 \log_{10}\left(\frac{P_{\text{total}}}{1\ \text{mW}}\right)

The same steps work for any number of signals. Convert all of them, sum the milliwatts, and convert once at the end. N equal uncorrelated signals give 10 log10(N) dB above any single one, so two equal signals give 3.01 dB.

Worked example: 0 dBm plus −10 dBm

Take a 0 dBm tone and an uncorrelated −10 dBm tone at another frequency, both arriving at the same power detector.

  • 0 dBm → 10^(0/10) = 1 mW
  • −10 dBm → 10^(−10/10) = 0.1 mW
  • Sum: 1 mW + 0.1 mW = 1.1 mW
  • Back to dBm: 10 log10(1.1 mW / 1 mW) = 0.41 dBm

The total is 0.41 dBm. Adding a signal one-tenth as strong raised the level by less than half a decibel. The wrong answer, −10 dBm, would be 10.41 dB too low. It would even put the total below the stronger signal on its own, which can't happen with uncorrelated signals.

Correction table: what to add to the stronger signal

The ratio of the total to the stronger signal depends only on the level difference D (in dB) between the two signals. If P1 is the stronger, then P_total / P1 = 1 + 10^(−D/10), so the amount added to the stronger signal is:

Δ=10log⁡10(1+10−D/10) dB\Delta = 10 \log_{10}\left(1 + 10^{-D/10}\right)\ \text{dB}

You can work this out once and keep it as a table:

Level difference DLinear sum relative to strongerAdd to stronger signal
0 dB2.0003.01 dB
3 dB1.5011.76 dB
6 dB1.2510.97 dB
10 dB1.10.41 dB
20 dB1.010.04 dB
051015200123Level difference D (dB)Add to stronger signal (dB)3.01 dB1.76 dB0.97 dB0.41 dB0.04 dBΔ = 10 log10(1 + 10^(−D/10))Uncorrelated signals only
Correction added to the stronger of two uncorrelated signals versus their level difference, valid for uncorrelated signals only. Computed from the formula.

The curve starts at 3.01 dB for equal signals and falls quickly. By 10 dB difference the correction is 0.41 dB, and by 20 dB it is 0.04 dB. Two practical readings follow from it:

  • Signal over a noise floor. A tone that reads 10 dB above the displayed noise is overstated by 0.41 dB, because the reading includes the noise power. At 20 dB above the noise, the error falls to 0.04 dB.
  • Dominant interferer. If one interferer is 20 dB stronger than every other contribution, you can use its level as the total. Within 10 dB, you have to do the sum.

To use the table, take the stronger level, find the difference, and add the correction. For example, −50 dBm and −53 dBm (uncorrelated) give −50 + 1.76 = −48.24 dBm.

The table and the curve are valid only for uncorrelated signals. Do not use them for two copies of the same carrier, such as a direct path and a reflection, or the two arms of a combiner fed from one source.

Coherent signals: when power addition fails

Coherent signals at the same frequency keep a fixed phase relationship, so their voltages add. The result depends on that phase. Two equal-amplitude signals in phase double the voltage. That gives 20 log10(2) = 6.02 dB above either one, not 3.01 dB. At 180° they cancel completely. Any phase in between gives a total somewhere between those two limits.

This is why a direct signal and its reflection can cause deep nulls along a cable or in a multipath channel, and why the in-phase or out-of-phase relationship sets the output of a combiner. If you can't be sure the two signals are uncorrelated, the 3.01 dB rule and the table above don't apply.

In plain words · for students and beginners

Adding dBm powers in plain words

An everyday picture

Think of two lamps lighting the same wall. The light from each lamp adds up, and two equal lamps give twice the light. dBm is just a scale for writing power, like a ruler with a stretched spacing where each 10 steps means ten times more power. You can't add the ruler readings. You have to add the actual amounts of light.

The picture breaks down for two copies of the same radio signal. Waves that line up add more strongly. Waves that are opposite (180°, meaning one goes up exactly when the other goes down) can cancel to nothing. Lamps don't do that, but coherent radio signals do. "Coherent" means the two signals are the same frequency and keep step with each other.

The worked example, step by step

We want the total of 0 dBm and −10 dBm, from two different signals.

  1. Turn 0 dBm into milliwatts: 10^(0/10) = 1 mW.
  2. Turn −10 dBm into milliwatts: 10^(−10/10) = 0.1 mW.
  3. Add the milliwatts: 1 + 0.1 = 1.1 mW.
  4. Turn the answer back into dBm: 10 log10(1.1) = 0.41 dBm.
Step 1: convert0 dBm → 1 mWP (mW) = 10^(dBm/10)Step 2: convert−10 dBm → 0.1 mWP (mW) = 10^(dBm/10)Step 3: add mW1 mW + 0.1 mW= 1.1 mWStep 4: back to dBm10 log10(1.1 mW / 1 mW)= 0.41 dBmnot −10 dBm + 0 dBm= −10 dBm
The worked example: two uncorrelated signals at 0 dBm and −10 dBm are converted to milliwatts, added, and converted back to give 0.41 dBm. Concept drawing.

The total is 0.41 dBm. It is a little more than the stronger signal, because the weaker one only had a tenth of the power.

Mistakes beginners make

  • Adding the dBm numbers. 0 + (−10) = −10 dBm is wrong. The total can never be lower than the stronger signal.
  • Assuming equal signals double the dBm number. Two 0 dBm signals do not make 0 dBm, and they do not make "double 0". They make 3.01 dBm, because 1 mW + 1 mW = 2 mW.
  • Using the 3 dB rule for copies of one signal. Two in-phase copies give 6.02 dB more, and opposite copies can cancel. The 3 dB rule is for unrelated signals only.

Remember

  • Convert to milliwatts, add, then convert back to dBm.
  • Two equal unrelated signals give 3.01 dB more than either one, and a 10 dB weaker signal adds only 0.41 dB.
  • Copies of the same signal add as voltages, so they can give up to 6.02 dB more or cancel out.

Frequently asked questions

What is 0 dBm plus 0 dBm?

For uncorrelated signals, 1 mW plus 1 mW is 2 mW, which is 3.01 dBm. That is 3.01 dB above either signal, not 0 dBm and not 0 + 0 in decibels.

What is 0 dBm plus -10 dBm?

Convert to 1 mW and 0.1 mW, add them to get 1.1 mW, and convert back: 10 log10(1.1) = 0.41 dBm. The weaker signal raises the total by only 0.41 dB.

When can I ignore the weaker signal when adding powers?

At a 20 dB difference the weaker signal adds only 0.04 dB to the stronger one, which you can ignore for most work. At 10 dB it still adds 0.41 dB, which usually matters.

Why do two in-phase signals add 6 dB instead of 3 dB?

Coherent signals at the same frequency add as voltages. Two equal voltages in phase double the voltage, which is 20 log10(2) = 6.02 dB. At 180° they cancel completely.

References and further reading