Question on FFT of DC value

[vkristof]

Background: I've been using the AD2/Waveforms to measure/analyze low-level/low frequency signals that have a DC component that we care about.

 

The signal is in the feedback loop of a system, and the AC & DC components are in ~ 1mV range. We have a 200 Hz LPF with a gain of 10 (offset trimmed) before we feed into the AD2 (and a Keysight Benchtop DMM & a Tek TDS3034C). The attached .png shows the standard "Measurements" of the 1V DC as a reference.

 

My question is related in conceptualizing the FFT of a DC signal, and specifically, why do the O Hz and  and next bin up read 3.01x and 2.70x dBv (in other words greater than 0 dBv) when I feed 1 V DC into CH1.

Any attempt to educate me in understanding this seeming discrepancy is appreciated. Please note that I checked the FFT portion with sinusoids from the AWG previous to this DC check, and the FFT bin contents (dbV values, whatever) were what I expected.

In this specific case the AWG is the source of the 1 VDC and the AD2 Scope/measurement reading are 1V (along with the Tek & Keysight readings).

Copy and paste from the excel of the FFT below. Bins go out to 4K, scope sample rate is 8k (as expected)

 

#Digilent WaveForms Scope FFT

#Device Name: Discovery2

#Serial Number: SN:210321B21C1B

#Date Time: 2021-08-25 17:12:26.733

NOTE: this FFT is available in it's scope capture form as .csv/.xlsx and .png. FFT is of a +1.0 V DC from the AD2 AWG and AD2 Measurements reports 1.000V DC RMS, 1.0008 V Average, etc. 

Frequency (Hz)Channel 1 (dBV)Phase (deg)

0                            3.017895119      0

0.488281              2.708513213     0.010973

0.976563             -0.79016818      0.022399

1.464844             -11.22347699    0.034118

[attila]

Hi @vkristof

The DC of FFT shows double of the actual value. This is the result of the algorithm.