ADC sampling for microphone front end signal processing

[cliffj]

ADC sampling

Hi,

I'm recently doing a project about microphone front end signal processing.
My circuit now has a high freq ripple around 10Mhz (compare with the mike signal bandwidth which is 20~ 20khz), and after my amplifier is an ADC. Which's sampling rate is around few khz. And I'd like to know if this would cause any performance degrdation of output digital signal? What happens if we use a low sampling rate ADC to sample a high speed supply ripple ??? Will it degrade the digital outout ??? (~5mV ripple, 250mV output signal swing, and 12 bit ADC)

Cliffj

 

[Vamsi Mocherla]

Re: ADC sampling

Hey in a signal conditioner block, you need to use an anti-aliasing filter to remove the high frequency components over the Nyquist frequency. Aliases will cause your ADC to give wrong results. So, your HF ripple will be filtered as well. If you can go for a higher sampling rate, it should be good because you can have less quantization noise on your ADC.

 

[willyboy19]

Re: ADC sampling

In your system, I assume you have a microphone amplifier preceding the ADC. You can still use the traditional low frequency ADC (a few tens of kHz) to do a nice job, if the Mic amp has a fully differential structure which features in high PSRR. But since the ripple you have on the supply has frequency as high as 10MHz, the Mic Amp may not have sufficient high PSRR at that frequency range. My suggestion is to have a clean regulator to regulate from this noisy supply and use this regulated supply to power your ADC's analog section of the circuitry. Another solution can be using a more clean supply for your ADC.

You can't just use a much higher sampling frequency to deal with the 10MHz supply ripple, as suggested by someone here. For a voice band application, even with sigma-delta oversampling ADC, you can only have sampling frequency as high as two to four Mega Hertz, not even close to the 10MHz ripple noise here.