dBc: Measuring Spurs and Harmonics Against the Carrier

dBc is a level relative to the carrier: P_signal (dBm) = P_carrier (dBm) + level (dBc). A -40 dBc harmonic on a +20 dBm carrier sits at -20 dBm, or 10 µW, which is one ten-thousandth of the carrier power.

Engineer’s note: Regulatory limits like FCC and CE are often written in absolute power (dBm), while chipset vendors usually quote spurs and harmonics in dBc. An amplifier with -40 dBc harmonics running at 30 dBm puts its harmonic at -10 dBm, which can easily break a regulatory limit. Watch the conversion.

For practitioners

What dBc means

dBc means decibels relative to the carrier. It is a ratio, not an absolute level. It tells you how far a signal sits above or below the carrier on the same spectrum:

level (dBc)=Psignal (dBm)−Pcarrier (dBm)\text{level (dBc)} = P_{\text{signal}}\ (\text{dBm}) - P_{\text{carrier}}\ (\text{dBm})

Both terms are in dBm, so the 1 mW reference cancels. What is left is 10 log10 of the power ratio between the two signals. A negative value means the signal is below the carrier, and almost every dBc number you meet in practice is negative.

The unit shows up wherever the question is "how clean is this signal": harmonics at 2f0 and 3f0, discrete spurs from clocks or synthesizers, LO feedthrough, and, as a density, phase noise in dBc/Hz. A relative unit suits these specs because what matters is how much unwanted energy rides along with the wanted signal, not the absolute power of either one.

On a spectrum analyzer, dBc readings have a practical advantage. Errors common to both tones, such as the analyzer's absolute amplitude uncertainty or a flat loss in the test cable, cancel out when you take the difference. Errors that change with frequency do not cancel. A cable that loses more at 2f0 than at f0 makes the harmonic look lower than it really is.

Converting a dBc spec to dBm and watts

Rearranging the definition gives the absolute level of the unwanted signal:

Psignal (dBm)=Pcarrier (dBm)+level (dBc)P_{\text{signal}}\ (\text{dBm}) = P_{\text{carrier}}\ (\text{dBm}) + \text{level (dBc)}

From dBm, convert to watts in the usual way:

P (W)=10PdBm/10×1 mWP\ (\text{W}) = 10^{P_{\text{dBm}}/10} \times 1\ \text{mW}

You can also go straight to the power ratio without passing through dBm:

PsignalPcarrier=10level (dBc)/10\frac{P_{\text{signal}}}{P_{\text{carrier}}} = 10^{\text{level (dBc)}/10}

The first route answers "what will a power meter or an absolute emission limit see". The second answers "what fraction of my output is wasted in this spur". Both come from the same number.

Worked example: harmonics of a +20 dBm carrier

Take a transmitter with a +20 dBm carrier at f0, which is 100 mW. The spectrum shows the second harmonic at 2f0 at -20 dBm and the third harmonic at 3f0 at -30 dBm.

+30+20+100−10−20−30−40−50−60Power (dBm)f02f03f0Frequency+20 dBm = 0 dBcCarrier at f0+20 dBm (100 mW)2nd harmonic−20 dBm (10 µW)3rd harmonic−30 dBm (1 µW)−40 dBc−50 dBclevel (dBc) =P_signal (dBm)− P_carrier (dBm)
Harmonics of a +20 dBm carrier read in dBm against the spectrum axis and in dBc as the drop from the carrier level. Concept drawing.

Relative levels:

  • Second harmonic: -20 - (+20) = -40 dBc
  • Third harmonic: -30 - (+20) = -50 dBc

Absolute power in watts:

  • Second harmonic: 10^(-20/10) mW = 0.01 mW = 10 µW
  • Third harmonic: 10^(-30/10) mW = 0.001 mW = 1 µW

As a fraction of the carrier, the second harmonic is 10 µW / 100 mW = 1/10 000. That matches 10^(-40/10). The third harmonic is 1/100 000 of the carrier. Going the other way, if a requirement says "harmonics at or below -40 dBc" and the carrier is +20 dBm, the absolute ceiling for each harmonic is -20 dBm.

dBc as a power ratio

Because dBc is 10 log10 of a power ratio, every 10 dB is a factor of 10:

LevelFraction of carrier power
0 dBc1
-10 dBc1/10
-20 dBc1/100
-30 dBc1/1000
-40 dBc1/10 000
-60 dBc1/1 000 000

A useful consequence: the power in harmonics and spurs specified at -30 dBc or lower is a negligible part of the total output. A -30 dBc spur takes 0.1 % of the carrier power. dBc limits are about interference to other users and receivers, not about lost output power.

In voltage terms the same ratios are square roots. A -40 dBc tone has 1/100 of the carrier's voltage amplitude across the same impedance, because 20 log10(1/100) = -40 dB. The dB and dBm signal-chain page covers this distinction.

dBc/Hz: phase noise per hertz

Phase noise is spread over frequency, not concentrated in a single line, so it is given as a density. Single-sideband phase noise L(f_m), in dBc/Hz, is the noise power in one sideband, measured in a 1 Hz bandwidth at an offset f_m from the carrier, relative to the total carrier power. This matches the IEEE definition L(f) = S_φ(f)/2 when the phase deviations are small.

If L is roughly flat over a bandwidth B, the single-sideband noise in that bandwidth relative to the carrier is:

NSSB (dBc)=L(fm) (dBc/Hz)+10log⁡10(B/1 Hz)N_{\text{SSB}}\ (\text{dBc}) = L(f_m)\ (\text{dBc/Hz}) + 10\log_{10}(B / 1\ \text{Hz})

Counting both sidebands doubles the power, which adds 3 dB. To get an absolute noise level in dBm, add the carrier power, exactly as for a discrete spur.

Pitfalls in dBc specs

A dBc limit means nothing without the carrier power it applies to. The same -40 dBc harmonic is -20 dBm on a +20 dBm carrier and a different absolute level at any other drive. Always write down the carrier level next to the dBc figure.
  • The carrier level moves the answer. A spur with a fixed absolute level, such as clock coupling, gets worse in dBc as the carrier is turned down. Harmonics caused by nonlinearity usually grow faster than the carrier as drive goes up, so their dBc value also changes with power. Quote dBc at a stated output level.
  • Sign conventions vary. Some documents write "40 dBc" and mean 40 dB below the carrier. Read the context before you add the number to a carrier level.
  • Modulated carriers need a defined reference. For a CW carrier, the carrier's peak on the analyzer is its power. For a modulated carrier, the reference should be the total channel power, measured as described in the power measurement comparison, not the height of one resolution bandwidth bin.
  • Frequency-dependent losses do not cancel. Correct for cable and fixture loss at each harmonic frequency, not only at f0.

In plain words · for students and beginners

dBc in plain words

An everyday analogy, and where it breaks down

Think of a singer on stage (the carrier, the main radio signal you actually want to send) and a faint hum from the speakers (a spur or harmonic, an unwanted extra signal; a harmonic sits at two or three times the carrier's frequency). Saying "the hum is 40 dB below the singer" is a dBc statement. It tells you how quiet the hum is compared to the singer, not how loud it is in the room.

Where the analogy breaks down: decibels are not plain subtraction of power. Every 10 dB down means ten times less power. So 40 dB below is not "40 units less". It is 10 × 10 × 10 × 10 = 10 000 times less power.

The same example in simple steps

dBm is power compared to 1 milliwatt (mW). 0 dBm is 1 mW, and +20 dBm is 100 mW.

  1. The carrier is +20 dBm, which is 100 mW.
  2. The second harmonic is specified at -40 dBc.
  3. Add them: +20 + (-40) = -20 dBm. That is the harmonic's real power level.
  4. Turn -20 dBm into watts: -20 dB means 1/100, so it is 1 mW / 100 = 0.01 mW = 10 µW (microwatts, millionths of a watt).
  5. Check: 100 mW divided by 10 000 is also 0.01 mW. The two routes agree.
Relative level(dBc)Absolute level(dBm)PowerFraction ofcarrier power0 dBc-20 dBc-40 dBc-60 dBc+20 dBm0 dBm-20 dBm-40 dBm100 mW1 mW10 µW0.1 µW11/1001/10 0001/1 000 000carrier2nd harmonicEach step-20 dB= 1/100of the powerP_signal (dBm) =P_carrier (dBm) + level (dBc)
A dBc-to-dBm ruler for a +20 dBm carrier, using P_signal (dBm) = P_carrier (dBm) + level (dBc), with each 20 dB step equal to a factor of 100 in power. Concept drawing.

The ruler works in steps of 20 dB. 0 dBc is the carrier itself at +20 dBm. -20 dBc is 0 dBm (1 mW, 1/100 of the carrier). -40 dBc is -20 dBm (10 µW, 1/10 000). -60 dBc is -40 dBm (0.1 µW, one millionth).

Mistakes beginners make

  • Reading dBc as an absolute level. -40 dBc is not -40 dBm. You need the carrier power to get a real level.
  • Thinking -40 dBc means "40 times smaller". It means 10 000 times smaller in power.
  • Dropping the minus sign. Adding +40 instead of -40 puts the harmonic far above the carrier, which makes no sense for an unwanted signal.
  • Forgetting the carrier can change. If the carrier power changes, the harmonic's dBm value changes too, even when the dBc number stays the same.

Remember

  • dBc = signal (dBm) minus carrier (dBm). It compares a signal to the carrier.
  • To get the real level, add: signal (dBm) = carrier (dBm) + dBc.
  • Every 10 dB is a factor of 10 in power: -30 dBc is 1/1000 and -60 dBc is one millionth.

Frequently asked questions

What does dBc mean in a spurious or harmonic specification?

dBc is the level of a signal in decibels relative to the carrier power. A harmonic at -40 dBc carries 1/10 000 of the carrier power, whatever the carrier's absolute level is.

How do I convert dBc to dBm?

Add the dBc value to the carrier power in dBm: P_signal (dBm) = P_carrier (dBm) + level (dBc). A -40 dBc harmonic on a +20 dBm carrier is at -20 dBm.

How many times smaller than the carrier is -60 dBc?

-60 dBc is one millionth of the carrier power. Each 10 dB step is a factor of 10 in power, so -30 dBc is 1/1000 and -60 dBc is 1/1 000 000.

What is the difference between dBc and dBc/Hz?

dBc compares a discrete signal, such as a harmonic or spur, with the carrier. dBc/Hz is a density: single-sideband phase noise L(f_m) is the noise power in one sideband, in a 1 Hz bandwidth at offset f_m, relative to the total carrier power.

References and further reading