dB, dBm, and Watts Explained

Watts measure power on a linear scale. Two watts is twice one watt, and a milliwatt is one thousandth of a watt. This form is necessary for energy, heat, efficiency, and physical loading calculations. Very large ranges can be awkward to compare linearly, which motivates logarithmic units in acoustics, electronics, communications, and control.

The decibel dB expresses a ratio and has no absolute reference by itself. A gain of 3 dB approximately doubles power, while a loss of 3 dB approximately halves it. Saying a transmitter is 20 dB is incomplete unless 20 dB describes gain, loss, signal-to-noise ratio, or a referenced level with another suffix.

dBm attaches the reference 1 mW. Zero dBm equals 1 mW, 10 dBm equals 10 mW, 20 dBm equals 100 mW, and 30 dBm equals 1 W. Negative dBm values are positive powers below 1 mW, not negative physical power. The minus sign belongs to the logarithmic level relative to the reference.

For two positive powers P2 and P1 measured on compatible bases, ratio in decibels is 10 log10(P2/P1). Reversing numerator and denominator changes the sign. A ratio of one gives 0 dB, a ratio of ten gives 10 dB, and a ratio of one hundred gives 20 dB. Equal decibel steps therefore represent equal multiplicative changes.

To return from dB to a power ratio, use P2/P1 = 10^(dB/10). Adding dB values corresponds to multiplying linear ratios, which simplifies cascaded gains and losses. Subtracting two levels in dBm gives their power ratio in dB, provided both levels use the same reference and represent comparable quantities.

Logarithms require positive power arguments. Zero watts corresponds to a level tending toward negative infinity dBm; it has no finite logarithmic representation. Negative signed power in circuit conventions cannot be inserted directly into the magnitude conversion. Treat direction or flow sign separately from positive power magnitude.

Related in this workflow: dB to Linear Converter, Linear to dB Converter.

Power level in dBm is 10 log10(P_mW/1 mW), commonly written 10 log10(P_mW) when the numerical value is in milliwatts. Converting back gives P_mW = 10^(dBm/10). To obtain watts, divide milliwatts by 1,000, or use P_W = 10^((dBm - 30)/10).

The 30 dB offset follows from 1 W being 1,000 mW and 10 log10(1,000) = 30. Therefore dBW, referenced to one watt, equals dBm minus 30. Reference suffixes matter: dBV, dBu, dBFS, dBc, and dBi represent different electrical, digital, carrier-relative, or antenna quantities and cannot be converted to watts without their definitions and conditions.

Calculators should preserve enough range for very small signals without rounding them to zero. Scientific notation is often clearer for microwatt, nanowatt, or picowatt levels. For example, -90 dBm is 10^-9 mW = 10^-12 W, or one picowatt. A physically meaningful received signal can be far below one milliwatt.

Related in this workflow: dBm to Watts Converter, Watts to dBm Converter.

Electrical power in a resistor is proportional to voltage squared divided by resistance and to current squared times resistance. If the compared impedances are equal, a voltage ratio V2/V1 corresponds to a power ratio squared, producing 20 log10(V2/V1). The same logic applies to current ratios under equal impedance conditions.

The factor 20 is not a universal rule for every voltage measurement. If source and load impedances differ, equal voltage ratios do not imply the same power ratios. RF systems often use specified reference impedances such as 50 ohms, while audio measurements may use voltage-level references that are not direct delivered-power statements. State impedance and whether values are RMS, peak, or peak-to-peak.

Field quantities such as sound pressure and electric-field amplitude can also use 20 log10 because their associated intensity is proportional to amplitude squared under stated conditions. The reference and physical definition remain essential. Selecting 10 or 20 based only on the unit label can produce a factor-of-two error in decibels.

A communication link budget commonly starts with transmitter power in dBm, adds antenna and amplifier gains in dB, subtracts cable, connector, filter, propagation, and mismatch losses in dB, and obtains received power in dBm. Units alternate deliberately: absolute levels remain dBm and relative changes remain dB. Adding two dBm values is not meaningful.

For example, 20 dBm transmitter power plus 6 dB antenna gain minus 2 dB cable loss minus 100 dB path loss gives -76 dBm before additional terms. Converting -76 dBm to watts yields approximately 2.51 x 10^-11 W. Whether that is adequate depends on receiver sensitivity, bandwidth, modulation, coding, noise figure, interference, required signal quality, fading margin, and regulatory limits.

Antenna gain may be specified in dBi relative to an isotropic radiator or dBd relative to a half-wave dipole. Path-loss models have geometry, frequency, distance, environment, and far-field assumptions. A clean arithmetic sum does not validate the physical inputs or establish compliance, coverage, or reliable service.

Convert 2 W to dBm. Because 2 W is 2,000 mW, the level is 10 log10(2,000) = 33.01 dBm. The same result follows from 10 log10(2 W) + 30. A quick check is that 1 W equals 30 dBm and doubling power adds about 3.01 dB.

Convert -20 dBm to watts. P_W = 10^((-20 - 30)/10) = 10^-5 W, or 10 microwatts. The negative level does not indicate reverse power. It indicates a magnitude one hundredth of the 1 mW reference, because -20 dB corresponds to a power ratio of 0.01.

A voltage rises from 0.5 V RMS to 2.0 V RMS across the same resistance. The amplitude ratio is four, so 20 log10(4) = 12.04 dB. Power increases by four squared, or sixteen, and 10 log10(16) gives the same 12.04 dB. If the load resistance changed, that equivalence would need a direct power calculation.

Noise power depends on measurement bandwidth. Thermal noise available from a matched source is commonly represented by kTB, so doubling bandwidth doubles linear noise power and raises its level by about 3 dB. A noise-floor value without bandwidth, filtering, detector, temperature, and impedance context is incomplete.

Spectrum analyzers may display power in a resolution bandwidth, power spectral density per hertz, channel power integrated over bandwidth, or amplitude with detector and averaging choices. Converting the displayed dBm number to watts is arithmetically possible but may not answer total signal power. Read the measurement mode and corrections before interpretation.

Signal-to-noise ratio is a ratio in dB, not normally an absolute dBm level. Received signal in dBm minus noise power in dBm gives SNR in dB when both refer to the same bandwidth and point. Comparing values from different bandwidths or reference planes creates a misleading margin.

The most common errors are adding dBm to dBm, using 20 log10 for power, using 10 log10 for equal-impedance voltage ratio, and forgetting the 30 dB watt-milliwatt offset. Another is treating zero dBm as zero watts. Memorizing the anchors 0 dBm = 1 mW and 30 dBm = 1 W catches many mistakes.

Do not mix peak voltage with RMS power formulas, or assume an impedance that was never specified. Do not treat dBi as dBm; antenna gain is a ratio while transmitter power is a level. Do not remove a minus sign from a dBm value because the resulting linear power seems small. Small received powers are normal in many radio systems.

Rounding in logarithmic cascades should occur at the end. A tenth of a decibel can matter in a tight budget, while component tolerances and propagation uncertainty may be several decibels. Report precision consistent with source data and separate deterministic insertion losses from statistical fading or measurement uncertainty.

These conversions describe scalar power levels and ratios. They do not model impedance mismatch, return loss, standing waves, modulation peaks, crest factor, nonlinear compression, intermodulation, noise figure, antenna radiation pattern, polarization, propagation variability, receiver demodulation, or regulatory spectral masks. Those effects require additional measurements and models.

ScholarTool converters support transparent educational arithmetic, not calibrated RF measurement or communication-system certification. Final link, EMC, safety, exposure, and compliance work requires correct reference planes, calibrated instruments, frequency- and impedance-specific component data, uncertainty and margin analysis, applicable standards, and qualified technical review.