EMC reports dB

[Freshman]

I have a question on EMC reports.

When there's a graph, as shown in this link (https://emcfastpass.com/emc-testing-beginners-guide/emissions/), just above CE limits.

Which are the signals that are being compared to arrive at a dB value? As I understand that, dB has no units. Just a number.

So, which are the 2 signals that are being compared to give a dB peak?

[Siwastaja]

As you say, dB is used for ratio. But if you look closely, the plots do not say dB. For example the image behind your link says dBµV/m on the y-scale. This is basically V/m scaled and plotted on a logarithmic scale.

dB alone can be used to describe the difference between the measured value, and the limiting value, e.g. by saying we are 3dB below the legal limit.

[radiolistener]

the link that you posted has picture which clearly state units dBµV/m

[Freshman]

I was asking like, which 2 signals are in question while measuring?

Suppose, I see a peak at, say 50MHz. Obviously, the peak signal at 50MHz is from the PCB in question. With respect to what signal is this 50MHz being measured and given in dB?

[radiolistener]

I was asking like, which 2 signals are in question while measuring?

Suppose, I see a peak at, say 50MHz. Obviously, the peak signal at 50MHz is from the PCB in question. With respect to what signal is this 50MHz being measured and given in dB?

dBµV/m (decibel-microvolts per meter)  is electric field strength in the environment (usually air or vacuum) represented in voltage per meter units.

dBµV (decibel-microvolts) is a voltage represented in dB units relative to 1 µVrms.

You can convert V/m units to dBµV/m the same as V to dBµV - with simple formula:

[dBµV/m] = 20*log10( [V/m] * 1000000 )

For example: 0.1 [V/m] = 20*log10( 0.1 * 1000000) = 20*log10( 100000 ) = 100 [dBµV/m]

Note: 0.1 [V/m] in this example is RMS voltage, do not confuse it with peak voltage.

Technically it is more correct to write [Vrms/m] or [dBµVrms/m] to distinguish it from peak values [Vpk/m] or [dBµVpk/m]

[Berni]

Suppose, I see a peak at, say 50MHz. Obviously, the peak signal at 50MHz is from the PCB in question. With respect to what signal is this 50MHz being measured and given in dB?

In respect to 1 uV/m just like the units on the graph say.

Do you also treat dB and dBm as the same thing?

[radiolistener]

it worth to mention that there is also exists dBu (decibel-unloaded) unit which is used in audio and it has different meaning than dBµV (decibel-microvolt), so dBu and dBµV units should not be confused, because sometimes dBµV can be written as dBuV.

[PGPG]

With respect to what signal is this 50MHz being measured and given in dB?

If you measure voltage you write how many times the measured value is higher than 1V. So 1V is 1 times higher then 1V and 10V is 10 times higher then 1V.
The same is when you want to write the result in dBV - you take measured value, compare it to 1V reference - how many times it is higher and then convert into dB, but the unit of result contains not only dB, but also the value you compared your value with.
So:
1 V = 20log(1V/1V) dBV = 0 dBV
10 V = 20log(10V/1V) dBV = 20dBV.

If you use dBuV then the reference you compare to is 1uV.
If you use dBuV/m then the reference you compare to i 1uV/m.

Consider also that basic unit is Bel, but all industry decided to use dB (1dB = 0.1B - like 1dm = 0.1m, and 1cm=0.01m)

And to not mix you too much I will not go into detail, but it is worth to remember that to convert into dB you not always use 20log, but sometimes 10log.

[MarkT]

when converting a power, use 10 log10(), when converting a voltage or current or other similar quantity use 20 log10().
This is because power is proportional to voltage squared (for a given impedance), and also current squared. 

In radio work free-space has a fixed impedance of 376 ohms, given by sqrt(µ0 / e0), so you can interconvert V/m to A/m if you want, everyone uses V/m though