Spectrum block shows incorrect dbm value

Dear all,

In the attached model NCO generates a sine wave of 32MHz and spectrum is set to RBW of 1MHz, 50Ohm resistance. We can see in the Spectrum, that the peak value in dbm is 6.9823, which corresponds to a peak-to-peak voltage of 1.413V. However, in the scope block we can see, that the signal amplitude is 2V peak-to-peak, which corresponds to 10dbm (dbm = 10*log10(1000*Vrms^2/R)). Why so?

The model is attached.

Thank you!

[Sergei]

Here is the model

[Mathieu NOE]

funny

the difference you mention is exactly 3 dB (as if you swap Vpeak vs a Vrms value = sqrt(2) factor for a sine wave)

[Sergei]

@Mathieu NOE do you mean that is funny because the spectrum is showing this result or because I have made some mistake in my conclusions and cannot see it? I hope the second because this would help me a bit)

And you are correct, if we apply the folloiwing formula - 10*log10(500*Vrms^2/R) - we get exactly the result from the spectrum. That is interesting, would be glad to understand why.

[Mathieu NOE]

hello

"funny" was maybe not the most appropriate word, I was just suggesting that there is somewhere a mistake as somewhere we use Vrms instead of Vpeak and that's why we see a 3 dB difference

but as I can't open your slx file (because I run R2020a release) I can't say for sure where the confusion may come from

Answer 1

[David Goodmanson]

Hi Sergei,

The plot shows peaks at both positive and negative frequencies. Each of those equal size peaks contains half of the power, so each peak will be down by a factor of 2, and

10dBm - 10*log10(2) = 6.9897dBm

[Sergei]

@David Goodmanson, thank you for pointing that out! This is absolutely logical and makes sense, I might have thought about it before. Thank you again!