• State TI Thinks Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 3 answers
  • Subscribers 31 subscribers
  • Views 606 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.

PCM1808: Noise is created at VINL when 4Vpp is applied at VINR

Koyo Asaishi
Intellectual 1710 points
Part Number: PCM1808
Other Parts Discussed in Thread: PCM1803A, PCM1860, PCM1840, PCM1821, PCM1820

Hello team,

When customer apply 4Vpp at VINL (or VINR), noise is created in the PCM1808 at VINR (or VINL).
VCC is 5V.

I referred this E2E post and Ravi recommend 3Vpp.
Why did he recommend 3Vpp?
How to avoid the noise when 4Vpp is applied to VINR or VINL?

Best regards,
Koyo

  • 0
    TI__Guru 62055 points

    Hello,

    With a +5V supply voltage, the full-scale analog input voltage of the PCM1808 is 3Vpp.  Applying more than that will saturate the input and will likely increase cross-coupling as you're experiencing.

    Therefore, keep the analog input less than, or equal to, the full-scale input voltage of 3Vpp to avoid saturation and increased interference to the other channel.

  • 0 in reply to Collin Wells
    Genius 3865 points

    Hello Collin-san,

     

    In current market circumstance, we have several new ADC sockets opportunity which is not require high performance 120dB ADC but 100dB range cost competitive ADC.

    PCM1808 is the most popular, low cost, long sealer ADC, however competitors one has better behavior when overloaded input signal swing. There is no crosstalk at all.

    I am facing same situation at my customer, what is your recommended 100dB low cost, simple HW control ADC without such a behavior?

    PCM1860 has too much function, PCM1803A is high cost, What else do you have?

     

    Regards,

    Mochizuki

  • 0 in reply to Mochi
    TI__Guru 62055 points

    Hi Mochi,

    If the input is a line-input when the overload happens and the source impedance is expected to be lower, would the customer consider an external attenuator before the ADC to limit the input voltage?  

    We haven't directly tested this recently, but could do so next time we're in the lab with the PCM1808 to see if our results match the customers.  Assuming the results match, other than limiting the voltage there may not be much we can offer as it seems to be an internal performance characteristic of the PCM1808.

    You've listed out our other HW controlled 2-channel ADCs.  We have a two performance options for a 2-channel variant of the PCM1840 coming online soon, the PCM1821 (106dB) and PCM1820 (113dB) but it has a similar amount of functionality as the PCM1860 devices.

  • 0 in reply to Collin Wells
    TI__Intellectual 1710 points

    Hello Collin,

    Why is this noise caused and not specify in the Recommended operating conditions?
    The problem is came from analog side or digital side?

    Customer hope to use full scale input voltage to gain DR characteristics.
    At 1kHz sine wave, they don't observe noise with 4Vpp but they observe noise with 4Vpp at 15kHz.

    You recommend PCM1840, PCM1821, PCM1820 and PCM1860 at last post.
    Are you sure those ICs are not problem about this noise?
    Do you have other 1ch device?

    They are looking for other ways using 4Vpp at 5V Vcc.

    Best regards,
    Koyo

  • 0 in reply to Koyo Asaishi
    TI__Intellectual 1710 points

    Hello Collin,

    Let me add measurement result from customer side.

    They have measured PCM1808, PCM1820, PCM1840, PCM1860 and PCM5140.
    The input sine wave is 20Hz/1kHz/15kHz @ Fs=48kHz and 20Hz/40Hz/1kHz/2kHz/15kHz/30kHz @ Fs=96kHz.

    They didn't observe crosstalk and noise at other than PCM1860 but observed at PCM1860.

    Could you help to check root cause asap?

    Best regards,
    Koyo

  • 0 in reply to Koyo Asaishi
    TI__Guru 62055 points

    Hi Koyo,

    This has very high priority and attention right now but realistically we only have one day left before major US holidays start. Please share the data and we'll do our best to take a look but new measurements will not happen until next year.  We apologize for the inconvenience, but we have team members that are unavailable but we'll do our best.

  • 0 in reply to Collin Wells
    TI__Intellectual 1710 points

    Hello Collin,

    How is this progress?

    Best regards,
    Koyo

  • 0 in reply to Koyo Asaishi
    TI__Mastermind 23050 points

    Hi Koyo,

    I have looked at the PCM1808 in the lab and can confirm a few things:

    1) The device and EVM do meet the channel separation spec given in the datasheet, but this requires that the other channel has a low impedance connection. I.e is connected to ground or a source is connected. If the other channel is left floating, you will see more cross-talk in the data output.

    2) When the recommended operating range of the device is exceeded, it looks like the modulator for the second channel also becomes unstable. This is likely due to internal coupling between the modulators when one of them is overflowing as this causes increased current on that channel. 

    As Collin mentioned, the input voltage should be limited to within the recommended 3Vpp operating range to avoid this behavior.

    Best,

    Zak