• State Resolved
  • Locked Locked
  • Replies 1 reply
  • Subscribers 46 subscribers
  • Views 1070 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

How should ADC input signal spectrum look like

Dhaval Patel46
Prodigy 40 points
Other Parts Discussed in Thread: ADC16V130, ADC16DV160

Hi,

      I am using ADC16v130 in my design. I have given provision to see spectrum at the input of ADC. I am using ADC clock of Fs=40.5 MHz. When I see the spectrum at the ADC input, clock is present at multiple of 40.5 MHz. Screen shot is attached. M1 =4*Fs,  D1= M1 - 5*Fs

     When I am given input to signal of 70 MHz pure sinusoidal signal, and I see the spectrum at the ADC input, 70 MHz is present in each nyquist zone. Screen shot is attach. Marker M1 = input signal, M2 = 4*Fs, M3 = 5*Fs, M4 = input singal in 9th Nyquist zone, M5= input singal in 10th Nyquist zone 

   Ideally, adc input signal is present in each nyquist zone after ADC, i.e if we plot fft of ADC data, signal will be present in each Nyquist zone but i am seeing this at the input of ADC.

   Is it normal or something issue with the my Layout ?

Regards,

  • 0
    TI__Expert 4340 points

    Dhaval,

    From this observation, I do not see a design issue. The ADC16DV160 has a switched-capacitor input network and periodic charge transients kick charge back out of the ADC input pins resulting in additional frequency content that you are observing on your spectrum analyzer. The intuitive explanation for the frequency content is the ADC sampling clock is modulating the input signal, leading to a component of the input signal observed in all the Nyquist zones.

    If the input network is design properly with some small capacitors to absorb the kick-back charge, then the transient should settle before the next sampling instant.

    The charge transients corrupt the signal at the ADC inputs *in-between sampling instants*. The true ADC performance should only be measured at its output with the samples.

    Regards, Josh