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.

TAS5825M: POP noise at transition between PLAY and Hi-Z

Part Number: TAS5825M

Hi,

Our customer asks us how to reduce POP noise at transition between PLAY and Hi-Z.
The output waveform of TAS5825M that I observed is shown below.

PVDD=24V,
Process Flow: FIR 48k 2.0 (All filters pass thru)
RL: 6-ohm loudspeaker
Yellow: OUT2-
Cyan: OUT2+
Red: (OUT2-)-(OUT2+)

1. Hi-Z to PLAY
Used script on I2C monitor is as below.
w 98 00 00
w 98 7f 00
w 98 03 03

POP noise was audible. 


2. PLAY to Hi-Z
Used script on I2C monitor is as below.
w 98 00 00
w 98 7f 00
w 98 03 02

POP noise was audible even at this time.

Best regards,

Akio Ito

  • Hi Ito,

    My colleague Matthew is on travel today, he will respond you when he is back to office.

    Dylan

  • Hi Ito,

    Normally output LC mismatch will also worse hiz -> play pop noise. Did you have chance to test with EVM board?

    What's more, TAS5825M can make sure the idle switching DC offset is less than 5mV, and the hiz -> play ITU box measurement result is less than 15mV. It needs to be very near to the speaker to hear the pop.

    Rdgs,

    Matthew

  • Matthew-san,

    Thank you for your reply.
    The customer have recognized POP noise by their prototype board, and the attached waveform is I measured a TAS5825MEVM.
    Today, I measured DC offset of TAS5825MEVM. The left channel offset was almost 0 mV, but the right channel was 9mV.
    As one division of the waveform voltage is 2 V, 9 mV is too small to distinguish.

    POP at the time PLAY to Hi-Z transition disappears with AC coupling by non-polar connected two 470 μF electrolytic capacitors.
    As 9 mV offset corresponds to about a 2^(-12) FS step signal, it is enough to be heard.
    I suppose they do not want AC coupling. Is there any way to reduce the POP except AC coupling?

    Best regards,

    Akio Ito

  • Hi Akio,

    Do you have any evidence to prove ‘ 9 mV offset corresponds to about a 2^(-12) FS step signal, it is enough to be heard'? What's the relationship btw idle DC offset and digital FS signal?

    What's more, could you show us the schematic of 'non-polar connected two 470 μF electrolytic capacitors'?

    It's suggested that you can use the same speaker with EVM same configuration to double check the pop again, and finalize the measurement method with customer. For example, 1 meter SPL level or ITU filter out electrical signal level. You attached ~2V signal includes lots high frequency content, which is not audible for speaker.

    Maybe you can send us email for more detaled discussion. (matthew-miao@ti.com)

    Rdgs,
    Matthew

    Rdgs,
    Matthew

  • Matthew-san,

    Thank you for your reply. I will send an e-mail to you soon.

    Best regards,

    Akio Ito