What units of Power/Amplitude Spectrum?

[Abraham_E]

Hello.

I have been reading a lot about noise such and to say the least it is a very confusing topic for me to grasp as there so many ways to define the problem.

I have this spectral measurement from my system. I am wondering what units the Y-axis should represent?

First graph shows an Amplitude Spectrum (RMS) at dB. Is it expressed as dBV? And to define both XY axis, is it dBV/Hz?

Second graphs, instead of amplitude spectrum, the power spectral density. Is the unit dBm?

Can I calculate the total noise of the system by just looking at that graph?

Thanks.

[Abraham_E]

I forgot to mention the input is voltage which is being measured.

[SunshineDesign]

You need to mention what you measured it with. If it was with a standard Spectrum Analyzer the units could be dBm. If the instrument was the National DAQ or High Speed DAQ/Scope, then units are in not dB and it needs to be scaled for the correct system impedance to get to dBm. To get the noise power, you have to convert from dBm to linear terms of Watts then you can integrate to get the total power. So you need to define more for us to help you get an answer.

[Viper]

What you show are log units which are unitless. For example dBV = 20* log(Vmeas/Vref). The volt units cancle out and you are taking the log of a number with units removed. If you want to use units you must convert to linear scale and use the reference unit to get the right units. i.e. Vmeas = Vref *10^(dBv/20) then Vmeas will be in units of volts. To convert to noise you will have to understand spectral density. Wikipedia is a wonderful thing. Please donate if you use it.

Norm

[Abraham_E]

Hello.

Thanks for your response.

SunshineDesign:I measured my circuit with a MCC DAQ USB-1208FS. How do one scale it for the correct system impedance? Do I need to calculate how much impedance my whole circuit has? Then it is a very complex task 🙂

Viper: I know. I have read a lot about noise and there is just way too many terms and definitions to grasp in a short amount of time. And it all seems to be relative, where one measurement cannot be applied to some other measurement.