TPA2013D1: Occasional circuit protection for large signal output

Hello Jose,

Thank you for your quick reply.
According to your advice, I additionally assembled 47kOhm and 470kOhm as R13 and R14 respectively, but unfortunately it had no effect on the trouble.
As I wrote in the previous post, circuit protection or breakdown was caused by beep (1kHz or 400Hz) at full or near-full volume. The output pin (VO+ or VO-) of fatally broken TPA2013 turned out to be short-circuited to PGND.
Here are measurements.
Figure 1 shows signals at the inputs of TPA2013, pin #7 and pin #8 at full volume. When beep sounds, the input swings 1.4Vpp, which obserbes absolute maximam ratings of TPA2013. Beep sounds for 200ms long.
Figure 2 shows a normal signal at the output to the transformer (at C13) at full volume . The peak-to-peak is about 4.5Vpp because R6 is changed to 4.7kOhm to avoid circuit protection.
Figure 3 shows a signal at C13 when circuits protection occurs at full volume. The peak-to-peak is 7.5Vpp with R6 of 1kOhm (original value).

If you need other measurements, do not hesitate to provide direction.

Ryoji

If you need other measurements, do not hesitate to provide direction.

Ryoji

Dear Jose,

Sorry for my long blank. Unfortunately the breakdown problem of TPA2013 has not been basically solved yet.
I tentateively changed R6 from 1k Ohm to 10k Ohm to increase loss of the beep. The countermeasure seems successful to prevent beakdown so far, but the root cause of the breakdown remains unclear.

I marshaled the measurement results as follows.
Figure 1 to 3 show input and output signal of TPA2013 using R6 of 10k Ohm, while Figure 4 and 5 show input and output signal of TPA2013 using R6 of 1k Ohm.
All measurements are done with the volume maximum, but tempolary or parmanent breakdown did not occur during the measurements.
Figure 1 and 2 both show beep at in+ and in- of TPA2013, with R6 of 10k Ohm. In Figure 1 the signals are DC coupled, while in Figure 2 AC coupled.
Figure 4 shows beep at in+ and in- of TPA2013 in AC coupling with R6 of 1k Ohm. These figure shows signals do not exeed absolute maximum ratings of TPA2013.
Figure 3 and 5 show beep at out+ and out- of TPA2013, measured at the terminals of T1. My concern is that amplitude of out+ and out- are not equal and balanced in both figures.

We measured V-I characteristics of the broken chip. It proved that shortcircuit occured between PGND (9,10pin) and VOUT- (11,12pin). But we could not find any visible evidence of the shortcircuit in the micrographs of the chip of which plastic mold was removed.

About your last questions, I have not add decoupling caps at Vdd pin yet. The only decoupling cap of 10 uF is placed at the 5.3V input of the PCB, 10 cm away from TPA5013. I will add some caps near TPA5013 and report the results. About the power dissipation, I never measured it but the output audio power of TPA5013 is about 1.5W.

It would be very helpful if you provide me possible couse of the breakdown.

Thank you for your help.

Ryoji