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.

TLV320ADC3101: Microphone input not loud enough

Other Parts Discussed in Thread: TLV320ADC3101, TLV320ADC3101EVM-K

When using the microphone inputs on the TLV320ADC3101, even with all the gains turned all the way up, the overall volume of recorded sound is very low.

We see this both with the EVM-K using the supplied microphones, and with our own microphones on our PCB. We compared recorded audio captured using the windows recording tool from the EVM-K, with audio recorded by a relatively inexpensive digital camera at the same distance. The camera audio comes through loud and clear, the EVMK audio comes in clear, just not loud. If we turn up the gain all the way, then the voice levels start to come up, but sadly so does all the noise.

Has TI seen this with their EVM-K's? Do they have "Get Great Microphone Audio from a TLV320ADC3101 for dummies"?

 

  • Hi Hedley,

    This depends on various factors, one of them is the microphone being used and another are the features implemented within the part.

    Electret condenser microphones typically have ~60dB SNR, so if gain in increased with an ideal amplifier to have a full-scale of 707mV (e.g. full-scale of ADC3101), then the noise floor would be 60dB below that. This would be greatly audible with typical headphones -- Our hearing has a dynamic range of ~130dB before hitting the threshold of pain. 

    Now, the signal that is output from the microphone can be relatively weak. A tone 3ft from the microphone outputs say, 20mVrms, which is around -31dB relative to 0.707Vrms full-scale of the ADC3101. This means that a gain of 31dB is needed to have a full-scale signal. Factored into that will be: ambient noise, the microphone noise and some amount of amplifier noise (see Figure 7 of the datasheet). Using the analog PGA (MicPGA) of the ADC3101, instead of the ADC PGA (which is digital), helps increase the dynamic range relative to the 0dB PGA SNR, because amplifier noise does not increase in a 1-1 relationship with the applied gain.

    Using the automatic gain control (AGC) feature in combination with the noise gate function helps to keep the nominal amplitude constant while returning the gain to 0dB whenever the amplitude is below a certain threshold (e.g. background noise). See http://www.ti.com/litv/pdf/slaa446 for information on how to configure this feature. There is a tool in the EVM-K software that you can 'see' how the waveform is affected upon changes of the different parameters.

    Let us know if this helps.

    Regards,

    J-

  • Hmm.. before we get too deep into the technical details, I want to ask if we are even barking up the right tree.

    Our objective is  "normal spoken voice at 20 ft is easily distinguishable in a quite environment" where quiet means in a typical office without lots of conversation going on, or in a house.  We've compared the results we can get using the EVM-K with the performance obtained from a canon point-and-shoot camera with video recording capability. The canon easily outperforms our hamfisted register settings of the EVM-K. The volume is louder, the apparent noise is less, it is easier to hear what people are saying.

    Should we expect to be able to achieve our desired level of performance using the TLV320ADCEVM-K?

     

  • Uh.. I meant to say the TLV320ADC3101EVM-K.

    BTW: For reference, the EVM-K BOM lists two possible microphones. One is here: http://www.knowles.com/search/prods_pdf/MD9745APZ-F.pdf


    Thanks,

    Hedley

  • Hedley,

    Most cameras use AGC to control sound levels. This might be the reason the camera is loud/clear.

    Regards,

    J-