TPA3251: TPA3251 Issue

li darren

Part Number: TPA3251
Other Parts Discussed in Thread: TPA3250

Hi,

My product used TPA3251, PBTL output, power supply, 34V, load is 4 ohm, it has mass production, but it was found output issue, 1 pcs. will be found about each 100 pcs. it is ok when the board was changed new device.

What's the reason?

over 6 years ago

0 Shawn Zheng over 6 years ago

TI__Mastermind 24835 points

Hi Darren,
We are looking into this and will get back to you later.

0 Justin Bohr over 6 years ago in reply to Shawn Zheng

TI__Mastermind 27850 points

Hi Li,

I have not seen an issue like this before with the TPA3251. Does it happen on every positive side of the sine wave? Have you looked at the PWM output to see what the PWM behavior is during this anomaly?

Have you physically examined the solder joints of the pins to make sure there were no issues?

To me it seems more like something that can happen from a DAC, when the information or clocks are incorrect on the sampling. What is the audio source for the amplifier?

Justin

0 li darren over 6 years ago in reply to Justin Bohr

Intellectual 520 points

Hi Justin,

all issue board happen on every positive side of the sine wave.

There are not problem about solder,there are I have took down the issue device and solder onto good board, it was same.

Audio source is analog signal.

0 li darren over 6 years ago in reply to li darren

Intellectual 520 points

Hi Justin,

replace R40(10K),R41(10K) to 0ohm, output wave is oK. but EVM schematic has these resistance, datasheet output C,D connect to GND, What would that do?

0 Justin Bohr over 6 years ago in reply to li darren

TI__Mastermind 27850 points

Hi Li,

R40 and R41 were on early EVMs, but we have since removed these resistors. They cause enough noise on the inputs to affect the operation of the TPA3251 in PBTL mode. It is recommended to tie Input C and Input D to gnd directly when in PBTL mode.

Does this mean the artifact in the output is gone?

Justin

0 li darren over 6 years ago in reply to Justin Bohr

Intellectual 520 points

Ok, We're aging tests.

0 li darren over 6 years ago in reply to Justin Bohr

Intellectual 520 points

Hi Justin,

Thanks for your reply, but there is a problem.

Is TPA3250 the same connection method? What effect are there?

There is a similar situation in TPA3250, so I put input C,D to GND, but it sounds different.

I find out remove PVDD decoupling capacitors or change 105 to 103 and issue disappear.

0 Adam Sidelsky over 6 years ago in reply to li darren

TI__Guru 61360 points

Li,

What issue are you having on the TPA3250?

Regards,

-Adam

0 li darren over 6 years ago in reply to Adam Sidelsky

Intellectual 520 points

Hi Adam,

TPA3250, PBTL out.

If put a heat sink on TPA3250 in some PCB, output is bad, and the schematic is as above. connect input C,D to GND direct, or remove PVDD decoupling capacitor(105), or change decoupling capacitor to 103, output is good. but I feel it sounds is different. What are the implications of these practices?

0 Adam Sidelsky over 6 years ago in reply to li darren

TI__Guru 61360 points

Li,

Adding a heatsink will not cause issue unless you are electrically connecting to the device pins or mechanically compressing the device too much. Have you looked at your PVDD close to the device when changing the PVDD cap? Do you see oscillation or artifacts? Can you measure the THD or anyway objectively measure the sound change?

Regards,

-Adam

0 li darren over 6 years ago in reply to Adam Sidelsky

Intellectual 520 points

Hi Adam,

Thanks for you reply.

Output is good when Remove PVDD cap proved heat sink did not connect to the device pin or compress the device too much.

1,it sounds different when input C,D directly connected to GND ， it is tone， feeling is insufficient bass. So, what effect when change C28,C27 to 0 ohm.

2,Output is ok when Remove PVDD cap or change cap to 103, it sounds no different, So, Can I do this？

0 Adam Sidelsky over 6 years ago in reply to li darren

TI__Guru 61360 points

Li,

Do you have R40 and R41 as 0 ohm? If not, they should be 0 ohm.

I don't see a problem with making C27 and C28 0 ohm.

I would suggest some capacitor on PVDD but the value is up to you and depends on your system. As long as 103 on PVDD is good for you, then it should be fine.

Regards,

-Adam

0 Adam Sidelsky over 5 years ago in reply to Adam Sidelsky

TI__Guru 61360 points

Li,

I am closing this thread. Please make a new one if you need more help.

Regards,

-Adam

Audio

Audio forum

TPA3251: TPA3251 Issue