• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 46 subscribers
  • Views 741 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.

ADC16V130 SNR improvement

Alex Savitskiy
Expert 1230 points
Other Parts Discussed in Thread: ADC16V130

Hello.

We are designing DDC receiver (decided to use ADC16V130) .
We must provide sensitivity above -120dBm (10 dB SNR / bandwidth 12kHz) [First Nyquist zone]. To simplify the calculations I take unmodulated carrier. 


My preliminary calculations are:
ADC16V130 SNR = 77 dBFS


After digital down conversion:
SNR Improvement (dB) = 10 x log (Sampling Rate/2x Signal Bandwidth)
SNR Improvement (dB) = 10 x log (122.880e6 / 2x 12e3) = 37dB


SNR summary = ADC16V130 SNR + SNR Improvement= 77 + 37 = 114dB
But for sensitivity -120dBm we must have SNR= 120dB + 10dB= 130dB


1) Tell me please, my calculates is correct?
2) How I can provide the sensitivity -120dBm on ADC16V130?


Thanks.

  • 0
    TI__Expert 4340 points

    Alex,

    You numbers look mostly correct. I assume that the 10dB factor in the sensitivity calculation converts 120dBm into units of dBFS. But, if I were to assume a 100ohm termination at the ADC and the 2.4Vpp full scale range of the ADC, the conversion factor would not be 10dB, it would be 10*log10((2.4/2)^2/2/100*1000) = 8.6dB. This means that the ADC's sensitivity in 12kHz would be -114 dBFS + 8.6dB = -105.4 dBm. Still, this is a long way from -120dBm. 

    For a given sampling rate and intermediate frequency (IF) and channel bandwidth, the ADC16V130 has a mosty fixed SNR, so there is little that you can do about the sensitivity and dynamic range of just the ADC itself. To reduce the integrated noise in a channel, here are things that you can do:

    1) Reduce the channel bandwidth. (This reduces the noise integration range)

    2) Increase the sampling rate of the ADC. (This decreases the sampled noise spectral density of the spectrum by spreading the noise)

    3) Decrease the IF. (This reduces the impact of broadband jitter on the noise spectral density of the spectrum)

    4) Minimize the amount of broadband noise on the ADC's clock (This also reduces the NSD via the jitter)

    A system solution to improve the sensitivty of the system is to put an amplifier in front of the ADC which improves the sensitivity of the combined subsystem proportionally, although it does not improve the dynamic range (in the presence of a large blocking tone). If you put an ideal amp in front of the ADC with ~15dB of gain, the sensitivity would improve from -105.4dBm to roughly -120.4dBm.

    Your task to achieve -120dBm with just an ADC is very challenging. Assuming the 122.88MSPS sampling range, this would require roughly 90dBFS SNR but there are no devices at this speed grade that can achieve this performance. You need a change to your system solution.

    For 12kHz channel and high SNR, you may be more interested in a device with delta-signal architecture like ADS127L01 (115dB SNR in 256kHz).

    Regards, Josh

  • 0 in reply to Joshua Carnes
    Expert 1230 points

    Josh,

    Thanks for your detailed explain.

    Is your formula { 10*log10((2.4/2)^2/2/100*1000) } from document  slyy068  "Direct RF conversion: From vision to reality" ?

    Please tell me, why you wrote 2 instead 8 (multiplier for Zin) ?

    Thanks.

  • 0 in reply to Alex Savitskiy
    TI__Expert 4340 points

    The equations are the same, but I put the a 1/2 inside the (.)^2 instead of bringing it out and combining with the 1/2 out side the squared parenthesis

    Regards, Josh