SNR in dBV vs dBm

by stevenjones3.1
Tags: None
stevenjones3.1 is offline
Jan6-14, 02:01 AM
P: 28
Hello All

I have a question regarding the SNR of an optical signal

The optical signal was measured with a FFT Spectrum analyzer with the amplitude in dBV
- the "SNR" was observed to be 60dB (peak value in dBV - noise floor value in dBV = "SNR" in dB)

The SNR is typically reported as the signal to noise ratio for power and not voltage. I was having a discussion with my boss about this and he was saying the the SNR in power could be reported as 120dB since there is a factor of two difference between the formula for dBV (w.r.t. V) and dBm (w.r.t. mW).

Although I can see his resoning, this seemed odd to me as other measurement (made in power) suggest that the SNR is not 120dB, however, the other measurements were made with spectrum analyzers that are not as good as the FFT spectrum analyzer (which measures in dBV).

I was just hoping someone could give some insight on this.

Cheers and thanks.
Phys.Org News Partner Physics news on
Information storage for the next generation of plastic computers
Scientists capture ultrafast snapshots of light-driven superconductivity
Progress in the fight against quantum dissipation
maajdl is offline
Jan6-14, 03:14 AM
P: 263
Remember how decibels are defined.
See :

The decibel (dB) is a logarithmic unit used to express the ratio between two values of a physical quantity (usually measured in units of power or intensity).
The decibel symbol is often qualified with a suffix that indicates which reference quantity or frequency weighting function has been used.
It follows that:

dB(Power) = dB(Vē) = 2 db(V)
sophiecentaur is offline
Jan6-14, 07:29 AM
Sci Advisor
PF Gold
sophiecentaur's Avatar
P: 11,341
Strictly, the Bel is a measure of the ratio of Powers (originally it was sound power iirc). A perfect transformer has a true gain of 0B - or 0dB (i.e. None).The Power ratio is 1, despite there may be a step up in volts. It's only when you are considering a system where the impedance is unchanged that you can use dB without making it clear that you are 'breaking the rules'.
In practice, it is often used sloppily and people can easily get twice or half the answer they wanted when they use the factor of 2 wrongly.
i.e 10log(ratio) or 20log(ratio)

When you get down to it, any amplifier / receiver / spectrum analyser is responding to signal Power. What it does with the value it gets and how it displays it will depend upon the specific requirement. The truth is the same. (Unless it's a cheap copy.)

RF equipment tends to use a fixed working impedance throughout and it's normal to use dBm, dBW or dBμ and the voltage problem doesn't arise. Use of the dB in audio is littered with confusion and misunderstandings. Except in studio systems with a standard working impedance (600Ω in the old days - but even there, many line amplifier inputs needed a 600Ω termination to be included or the dreaded 6dB error would creep in.

Register to reply