• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 32 subscribers
  • Views 416 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

PCM1863: Minimum Energysense Threshold

Adam Lock
Intellectual 305 points

Part Number: PCM1863

Hello,

In the data sheet I see no specification for the minimum Energysense threshold. However, the EVM only allows you to set it to -57dBFS. I want to set it to -72dBFS; however, I notice that i get constant interrupt triggers when i manually set the threshold to 0x00 0x08 0x3B. With some voltage divider changes I can make -60dBFS work in my design and I see proper signal detect functionality on the interrupt line.

Is there a reason that -57dBFS is the minimum available value for Energysense in the EVM? What is the minimum value that can be functionally used?

I am using a standard 24-bit signed level calculation, same as would be used for gain and it appears to match up with the tool generated values.

- Adam

  • +1
    TI__Mastermind 23050 points

    Hi Adam,

    What signal frequency are you using to test the energysense? Keep in mind the secondary ADC does not have a flat frequency response so your input signal may be attenuated prior to the energysense flag and it is likely being attenuated below the noise floor of the secondary ADC. 

    Although the datasheet was updated to change the secondary ADC accuracy spec to 12bit from 10 bit, I believe it is only 10-bit accurate. This would translate to an ideal SNR around 62dB, and the actual SNR would be slightly lower. Thus, an energysense threshold below 60dB would trigger just from the noise of the ADC itself. Also keep in mind the energysense accuracy is only about 3dB.

    Best,

    Zak

  • 0
    Intellectual 305 points

    Zak,

    That makes sense why a -72dB threshold was not possible. I will just be routing a passively attenuated copy of the input signal to another input on the ADC to use for energy sense. The signal is full bandwidth, and 1kHz has been used as the test frequency. I will need to add in some diodes to clip the signal to prevent over voltage on the ADC. We shall see what that does to THD.

    Just my 2 cents but signal detection only at -57dB and up is limiting. I understand the technical limitations, but would love to see an accessible signal detection range of -80dBFS to -32dBFS. 8bits of usable dynamic range. And with sub-woofers you are throwing away another 20dB or so, for me this is my main application for signal detection.

    Thank you,

    Adam

  • 0 in reply to Adam Lock
    TI__Mastermind 23050 points

    Hey Adam,

    Thanks for the feedback. I would also keep in mind that the secondary ADC has an input impedance of ~20kOhms so you may need to factor this into your attenuation as well. As for the subwoofer response, it is possible to use the biquad filters in the secondary ADC to make the response more flat over frequency, but of course you will still be limited by the noise floor of the ADC.

    Best,

    Zak