TPA6139A2: Too high current consumption - /MUTE floating

Hi Jaime,

Welcome to e2e.
It is recommended to set the /MUTE pin to a defined state. This is recommended for any device that has enable or mute controls.
That missing connection could have affected the current consumption. But increased current consumption when load connected and /MUTE to GND is not expected.

Regarding the schematic, I've noticed that VSS is directly connected to GND. That is not correct. You should place a decoupling 1uF cap from VSS to GND. Your current connection of VSS is actually a short from the internal charge pump outpu to GND, this could explain the increased current consumption.
I would recommend to add this capacitor and try again, if possible with a new IC.

Best regards,
-Ivan Salazar
Applications Engineer - Low Power Audio & Actuators

Hello Ivan,

Thanks for the prompt reply. Certainly the lack of capacitor at Vss is a good reason for the issues I was experiencing.

Will give it a try and report when get it solved.

Regards,

Jaime Aranguren

Hi,

Installed the 1uF cap between pin 5 (Vss) and GND using a completely new device, also replaced the 5V to 3V3 LDO; different behavior, but no success.

- The LDO was again totally destroyed.
- Nothing out of the TPA6139A2.

Surely there is something really bad on my design (maybe the headphone driver choice?), next approach is to try on breadboard using a PCB adapter, to have more freedom and faster iterations for experimenting.

Regards,

Jaime Aranguren

Hi Jaime,

Is there a chance that you could share your layout for review? It is weird that the device is apparently connected correctly by the schematic but there must be something wrong.
You can send the design files to ivan.salazar@ti.com if you don't want to upload it to the forum.

Best regards,
-Ivan Salazar
Applications Engineer - Low Power Audio & Actuators

Hi Ivan,

Thanks for the offer, I've just sent you PDFs of the layout and a couple of pictures to your email.

Regards,

Jaime Aranguren

Hello Ivan,

Thanks for your offline review. I should say, I am impressed by how difficult and unstable this part is. A total disaster, I am really disappointed.

I made a prototype using TH components on protoboard. At the beginning it was OK, I could see it working (without load). This means, I had my wiring fine, and I was happy and thinking the problem was solved.

After some minutes in operation, the output signal started to drop spuriously. However, input signal and power were rock solid and noise free at all times... Still the outputs were disappearing for moments. Until they suddenly disappeared completely.

Power cycling, adding 33 Ohm load, discharging the 1uF caps with 2 Ohm resistor (of course with power off), adding 68 Ohm series resistor at the output... nothing helps.

The device did not get hot, nothing exploded, no smoke... Simply does not work!!! I notice, however, that current consumption rose to ~100 mA, whereas initially (when working) it was below 10 mA.

The only thing I can suspect is the specs of the charge pump capacitors. The only thing said is "low ESR" and minimum 1uF which is also the value used on the EVM. How low ESR is low enough?

With TH capacitors it is not easy to find 1uF having low ESR, can you recommend parts?

I am currently using CT4-0805Y105M250A1T. Other possible parts that I can source quickly and locally for experimentation could be:

• CT4-0805Y105M500A1
• CT4-0805Y105M250A1
• 453382
• 330C105M5U5TA+ (Kemet)
• C330C105K5R5TA7317+ (Kemet)

Which one do you recommend and why?

Nevertheless, it is really too risky to design-in such an unstable Op Amp . Or, do you have a suggestion to make it work rock solid?

Thanks and regards.

Hi Jaime,

You can try with the component values from the EVM. And you can also try the EVM itself for a solid application example.

Best regards,
-Ivan Salazar
Applications Engineer - Low Power Audio & Actuators

OK, thanks.