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.

What is the right ADC?!

Other Parts Discussed in Thread: ADS8881, ADS8883, ADS8885, ADS8861, ADS8863, ADS8865, ADS8860, ADS8862, ADS8864, ADS8353, ADS8350, ADS8328, ADS8330, ADS8354

Hello guys,


I am looking for an ADC, but as I am new to this materia I'd like to ask you for an advice.

The ADC will be used for converting an audio signal which is burried in noise (~ -50dB to -70dB) and later recoverd with lock in technique.

I am thinking about a 16bit to 24bit converter with up to 350kHz sampling rate. This is pretty easy to find in the offers of TI.

The next point is the "low noise" property and here I am not quite sure what fits. So what are good values (SNR, SINAD) for such ADCs??

Wanted properties:

  • 16-24 Bits
  • 350kHz samplerate (minimum)
  • Serial interface (SPI if possible)
  • low noise!
  • 2 Channel (if possible, but 1 would be enough)
  • Features welcomed but not necessary: AA filter, DC filter

I hope you can give me some advice.

Here are 2 ADC I found, maybe you can tell me what you think about them:

AD S8865 | AD S8864


Kind regards

Stefan

 

  • Stefan,


    There isn't anything in the delta-sigma ADC that goes up to that high of a data rate. I'll see if one of the SAR ADC engineers has any comments about the devices you mention.

    In the meantime, since you are sampling an audio signal, is there a reason that you don't try using an audio ADC? Even though the data rate is much lower that what you're describing, the signal is likely to be band limited to audio frequencies.


    Joseph Wu
  • Hey Joseph,

    thanks for your answer. The microphone I am using is for detecting photo acoustic signals. (laser excites gas at certain frequency, gas expands and compresses with the same frequency --> microphone measures this "tone".) Up to now the tone is at around 4kHz so I use 48kHz sampling rate. But there will be measurements with tones up to 16kHz or even 35kHz. As the amplitudes of these tones are so small, the converted voltage signal from the microphone is burried in noise. To recover this signal we want Fs = 10*Ftone.

    Thats why we need low noise ADC with high accuracy and also high resolution as the SNR can be around -70dB. Making the input range of the adc much smaller to gain more resolution with less Bits can lead to clipping if loud external sounds occur near the microphone.

    Still hoping you can help me.

    Kind regards

    Stefan

  • Hi Stefan

    In order to meet your requirements, you can consider the following single channel devices from the SAR ADC portfolio:
    1) ADS8881 - 18bit, 1Msps, SPI, Fully Diff
    2) ADS8883 - 18bit, 680ksps, SPI, Fully Diff
    3) ADS8885 - 18bit, 400ksps, SPI, Fully Diff
    4) ADS8861 - 16bit, 1Msps, SPI, Fully Diff
    5) ADS8863 - 16bit, 680ksps, SPI, Fully Diff
    6) ADS8865 - 16bit, 400ksps, SPI, Fully Diff
    7) ADS8860 - 16bit, 1Msps, SPI, Single Ended
    8) ADS8862 - 16bit, 680ksps, SPI, Single Ended
    9) ADS8864 - 16bit, 400ksps, SPI, Single Ended

    If you are interested in dual channel ADCs, you can also look at ADS8328, ADS8330, ADS8353, ADS8350 and ADS8354.

    In terms of noise performance, these devices are Nyquist type converters where the overall noise is a root-sum-square of quantization noise and internal circuit noise. In an ideal case, assuming the internal circuit is noise free, the SNR for the ADC is equal to SNR = 6.02*N + 1.76 dB, where N is the ADC resolution. By this formula, the SNR for an ideal 16-bit ADC should be 98.1 dB. In reality, all ADCs have internal noise so the SNR spec is less than this ideal value but you can get an idea about the ADC precision performance by comparing the specified SNR number with the ideal number.

    SNR only accounts for circuit noise whereas SINAD includes the effect of noise as well as distortion so it is slightly less than the SNR spec. Hence, SINAD will be a more realistic number to look at if you are trying to assess the device performance.

    In addition to this, the overall SNR/SINAD performance of the signal chain also depends heavily on the noise and distortion performance of the front end signal conditioning circuit.

    For your particular application, we can make further recommendations if you can provide a little more details on the signal levels that you are trying to resolve.

    Thanks