TPA3113D2: Amplifier no audio issue - getting in to fault mode needs power cycle.

Hi Prasan,

Is this something you can consistently replicate?
From your schematic I noticed that the bulky decoupling cap on the supply is located before the ferrite bead, is this at least close to the IC? 
Regarding the ferrite beads, the 3 options mentioned in the data sheet are rated for 5A and higher, perhaps it's worth testing without the ferrite beads or with a higher rated one.

Best regards,
-Ivan Salazar
Applications Engineer

Hi Ivan,

Its not consistent, hence its being so tricky to debug this issue. 

Device will be working fine and all of a sudden audio stops. We have seen such scenario in our lab with field returned unit and then we tried couple of things in OS to understand if the driver reset fix it, but no. only power cycle would fix it. 

The PCB is as per the recommendation, caps are close to the IC. What happens if the output FB are 3A rated? Does it trigger over current? We dont see any fault triggered on Amplifier FAULT# pin during these events. If this was due to over current and DC detect, I am expecting FAULT pin low but its no the case.

Thanks for your time!

-Prasan

Hi Prasan,

I agree the device should trigger FAULT if such issue like OC happens, and since this is not really defined, I'm wondering the low current rating of the inductor, perhaps combined with speaker impedance variations over frequency/temperature/etc could combine to cause some unexpected high current peak that could mess up with the device logic.

If you test without ferrites, please make sure to also remove the capacitors to GND. The theory here is that if ferrite beads go into saturation, they become a short and connect the output of the Class-D directly to the cap to GND, which is not desired.

Best regards,
-Ivan Salazar
Applications Engineer

Hi Ivan,

Thanks for the details.

Could you please share datasheet reference which is recommending 5A ferrite bead in option 3?

In other words if OC is the issue, by reducing the ferrite bead current capacity further low I should be able to reproduce this issue more frequently or consistently. right? Then I can check if this OC triggers #FAULT or not. Currently this issue is totally random and we dont know how to reproduce this during its functionality. Sometime this happened during the device boot up itself, but there is not set procedures.

I will plan to check with both higher and lower current range FB to understand the behavior and keep you posted once done.

Thanks,

Prasan

Hi Prasan,

Sounds like a good test.
Here's an extract from data sheet:

Best regards,
-Ivan Salazar
Applications Engineer

HI Ivan,

Sorry, I did not check those parts specs last time. Thanks for pointing out.

However, I tried testing with FB having 1A current range, and when I turned ON the device it worked perfectly fine. We are using 8ohm 8 watt speaker and with this FB (0603) packaged no trouble observed. I set the volume to full, change the frequency to low and tried all possible way but that did not create any such issue what we have been seeing randomly.

I am almost reaching dead end in this testing with amplifier and not sure if I have anything else to try out. 

Thanks,

Prasan Hegde

Hi Prasan,

Unfortunately, being the failing test not replicable is hard on our end to give further suggestions.

Best regards,
-Ivan Salazar
Applications Engineer