Post Go back to editing

Why ACE FFT Peak Power is Different from Graph (off by 4dB)

Category: Software
Product Number: ACE
Software Version: v1.28.3252.1429 (x64)

I'm observing fft peak power on ACE being different from when it gets exported & plotted. Follow graph is a 1kHz signal measurement using AD4630-24. 

ACE measures fundamental power -0.491dBFS.

When exported in csv file, the FFT plot (done in matlab) show fundamental power is -4.81dBFS. Then I came back to ACE plot and noticed that dot on fundamental, lying somewhere around 0dB (full scale), whereas the tone below it has a peak at around -4dB, which aligns with MATLAB plot. Why the small gap and difference? 

Thread Notes

Top Replies

  • Hi  ,

    Can you try putting a ground connection between the ground pin from the signal generator to the ground pins of your board? On the other hand, are you using our evaluation board for the AD4630…

  • Hi  ,

    Can you try putting a ground connection between the ground pin from the signal generator to the ground pins of your board? On the other hand, are you using our evaluation board for the AD4630? And what are your inputs to the ADC?

    Warm regards,
    Jo

  • Hi Jo,

    Yes, I can certainly give it a try. There are exposed ground pins on AD4630 eval board, and on the signal source which is the ADI's low latency dev kit (DAC part number is AD3552r). The input comes from the AD3552r DAC output, single-ended, sinusoidal wave generated within the DDS of LLDK system. Various frequencies were available to test, in this one I was testing 1kHz 5Vpp full scale signal.