TPA3113D2: Amplifier no audio output issue, getting in to fault mode

Part Number: TPA3113D2


Tool/software:

Hi SIR:

Our device uses a TPA3113D2, with a 12V PVCC input, Audio Input Voltage 0.6 Vrms,a GAIN setting of 20dB, BTL setting ,and a PLIMIT setting of 3.5V. using  1PCS 5W 8ohm when speaker.

After using it for a while, we encountered the problem of no audio output.

Its not all the devices are behaving like this, About 500 tablets, 4 of which had this issue, When this problem occurs, the power must be turned off and then on again to output sound normally.

This problem is not easy to replicate,However, at an ambient temperature of 60°C, this situation may be reproduced,But when the temperature dropped back to 0 degrees, The thermal protection has no automatic reset feature, so the problem is not caused by heat.

When something goes wrong, FAULTZ output is low level ,IC pin3&4 (Left channel positive and negative audio inputs) unbiased 3 V outputs,IC Pin23&26 No oscillator frequency output.

The output and input beads were changed to 5A, but the problem remains.

Is there a solution to this problem?

Thanks.

  • Hi Dream,

    Just to clarify, this fault is happening after a certain amount of time playing audio at room temp? 

    Also, did you try holding the SD pin low to clear the fault? 

    A possible scenario could be a DC detect fault happening which will reported on the fault pin as a low state. This causes the device to shutdown and turn the outputs to Hi-Z. To clear the DC detect fault, you must power cycle PVCC. 

    Here's a snippet from the datasheet that describes how a DC detect fault is triggered. 

    Regards,

    Sebastian 

  • Hi Sebastian

    1. Just to clarify, this fault is happening after a certain amount of time playing audio at room temp? 

    YES,about 1 month.

    2.Also, did you try holding the SD pin low to clear the fault?

    YES,But FAULTZ did not reset.

    3.Regarding Table 3. Does it refer to PVCC or AVCC or Left channel positive and negative audio inputs?

    Thanks.

  • Hi Dream,

    The DC Detect Threshold table refers to the voltage differential between the positive/negative audio inputs. 

    Cycling the SD pin won't work if it is a DC detect fault.

    Could you possibly probe the inputs on the boards where this is is occurring to see if there is a voltage differential equal to or greater than 112mV for 420ms (min duration) ? This would trigger a DC detect fault. 

    Regards,

    Sebastian 

  • Hi Sebastian

    At an ambient temperature of 60°C, 3 PCS were tested for 2 weeks, and 1 PCS had a problem.

    The following is the waveform measured when the problem occurs.

    When measuring OUTPL and OUTNL with a multimeter on a problematic board, there will be a voltage difference of more than 0.3V, while a normal board will not have a voltage difference.

    Is this the problem that causes the DC detection abnormality?

    Thanks.

  • Hi Dream,

    Thanks for providing the scope shots. Yes a voltage differential across the OUTPL and OUTNL pins of 0.3V would certainly trigger the DC detect fault. Below are the thresholds:

    Were any of the passive components connected to the output pins damaged on the problematic board? 

    Regards,

    Sebastian