• State
  • Locked Locked
  • Replies 6 replies
  • Subscribers 30 subscribers
  • Views 1844 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.

TLV320AIC3268: Class D Speaker Over-Current Issue

Durka88
Prodigy 140 points
Part Number: TLV320AIC3268

I was hoping someone could help me undestand how the short-circuit protection feature works for the Class D speaker driver on the TLV320AIC3268. I have a custom board that keeps having the over-current flag being set for the SPK driver. I've tried to recreate this by using the same register configurations on a TLV320AIC3268-EVM and short the SPKP and SPKM by connecting a wire from TP41 to TP40. Is there something else that could create this over-current flag to get set? Also, why can't i re-create this on the EVM board? 

B0_P0_R44 [Sticky Flag Register 2] = 0x80

B0_P1_R69 [over current flags] = 0x20

  • 0
    TI__Guru 57800 points

    Hi, Durka,

    The short-circuit flag will be ON when a short circuit condition is present in one of the speaker outputs, I was able to get the flag ON in the EVM by tying one of the outputs to GND. The flag was cleared once the register is read. Can you please verify there is no short to GND in either the pins or in the speaker output signals of your board?.

    Over-current flag is only applicable for receiver driver, not the class-D speaker output, are you using the Receiver driver of the 'AIC3268?.

    Thanks and Best Regards,

      -Diego Meléndez López
       Audio Applications Engineer

  • 0 in reply to Diego Melendez
    Prodigy 140 points

    I'm not following you about the over-current flag not being applicable for the speaker driver output. It's right in the datasheet and there's a register bit dedicated for it, please see those snipets below. I was able to recreate the same over-current flags on the EVM by tying either the SPKP/SPKM to GND. I tested our implementation and didn't see a short between the SPKP/SPKM to GND or to each-other. However, I'm curious if some caps that we have might create an issue, depending on how the over-current test is implemented on the AIC3268.

    Do you think C89 and C90 might be the issue that's triggering the over-current condition? 

    Here is the snippets from the datasheet talking about the over-current for the class d speaker output driver.

  • 0 in reply to Durka88
    TI__Guru 57800 points

    Hi, Durka,

    Sorry for the confusion, you are right, overcurrent flag is applicable to speaker output as well,  I was referring to overcurrent protection as the device capability to limit the amount of current, while short circuit protection is referred to the device capability to shut-down the output once a high current (or short circuit) is detected at the output.

    Regarding your question, I don't think the caps are contributing to the overcurrent issue, but one thing to mention is that the LC filter used is not well designed as it has a cutoff frequency of about 700KHz (8 ohm load). This device should be able to operate without the LC filter with a ferrite bead only.

    What is the value of the speaker used?. A possible issue is that the speaker amplifier power supply doesn't have a high frequency 0.1µF decoupling capacitor. Typically with Class-D amplifiers, this capacitor is crucial to remove high frequency peaks on the output switching PWM. These overshooting peaks can cause high current peaks at the output stage.  You can try changing the C83 cap to 0.1µF. Also, the location of this cap should be as close as possible to the SVDD pin, can you confirm the location of the decoupling caps of SVDD in your board?.

    Best Regards,

      -Diego Meléndez López
       Audio Applications Engineer

  • 0 in reply to Diego Melendez
    Prodigy 140 points
    Thanks Diego, currently I'm trying to use a 8ohm/4W speaker. The location of the decoupling caps is closest to the SVDD as possible. But even if I don't have a speaker attached I still get the over-current issue. Could you explain when in the power up sequence is the overcurrent condition checked?
  • 0 in reply to Durka88
    Prodigy 140 points

    FYI removing the caps on the SPKP and SPKM line solved the problem. 

  • 0 in reply to Durka88
    TI__Guru 57800 points

    Thanks for the feedback Durka,

    As mentioned, the LC filter was not properly designed, so it is possible that the overcurrent condition was triggered by excessive currents at very high frequencies. The inductor can be removed as well, being replaced by a ferrite bead if is suitable for your application.

    Best Regards,

      -Diego Meléndez López
       Audio Applications Engineer