TAS5720L: TAS5720L : Ouput Waveform Issue

Hi Robert,

Thanks for your reply.

I have a doubt that this issue is not a problem of filter design.

OUT_P (Positive) and OUT_N (Negative) have to be differential signal.

My question is that why OUT_P (Positive) and OUT_N (Negative) are same waveform.

Could you explain that?

Anyway, about follow your guide.

To use differential signal between OUT_P (Positive) and OUT_N (Negative), do you means that we have to choose one of below filters?

Could you please send me your final opinion and explain?

Best Regards,

Michael

Hi Michael,

Are you putting an input signal into the amplifier when you are capturing the waveform? If there's no input signal then both sides of the half-bridge will look the same.

Also when looking at the signal on the oscilloscope, please choose a ground that's as close to the test point you are using to avoid getting as much distortion on the signal. Reading a cleaner waveform will allow you to better see the differences. 

For BD modulation I wouldn't choose either filter option. I would choose Hybrid Filter but without the C_BTL capacitor as BD modulation doesn't see much use for that capacitor.

Best Regards, 

Robert Clifton

Hi Robert,

Thanks for your kind comment.

I have a question.

Output is abnormal. In other words, OUT_N and OUT_P are same signal.

And signal level after LC Filter is very low that is about 320mVpp.

Even if speaker is removed, the level is the same.

so we can't hear sound on the speaker.

Could you please review and guide?

* Speaker : 4ohm

* BCLK to LRCLK ratio is 64, so we tied between MCLK to BCLK directly as follow data sheet.

* I2S input waveform (16bit, 44.1khz Music Sound)

* OUT_P, OUT_N pin of TAS5720L

* Both side of speaker after LC Filter

Best Regards,

Michael

Hi Robert,

For your more information, below is register value when we tested it.

This value is same which is sent from me at first time.

You can see "SAIF clock errors are present" at 0x08 register.

What's the problem?

Please review it more detail.

[0x00]Audio Device ID           = 0x01
[0x01]Audio Power Control    = 0xFD
[0x02]Audio Digital Control1  = 0x04
[0x03]Audio Digital Control2  = 0x80
[0x04]Audio Volume Control  = 0xAF
[0x06]Audio Analog Control   = 0x55
[0x08]Audio Status                 = 0x08
[0x10]Audio Digital Clipper2   = 0xFF
[0x11]Audio Digital Clipper1    = 0xFC

Best Regards,

Michael

Hi Robert,

I confirmed that readed value (0X08) of 0x08 register is when I2S signal is not inserted to audio amp.

SAIF clock errors is cleared when I2S signal is into audio amp. 

I'm sorry for the confusion.

Anyway could you please send me your reply?

I'm waiting your quick response.

This is very urgent issue to our customer.

Best Regards,

Michael

Hi Michael, 

Just to confirm the only issue you are still having is you aren't getting any audio signal out of the device? 

Regards,

Robert Clifton

Hi Robert,

Yes, there is no audio signal output from audio amp.

Isn’t there any your review opinion about what I sent you the test result?

Pls help me.

Best Regards

Michael

Hi Michael,

Yes with what you have provided, we can now review the system and hopefully get back to you today with whether we are seeing the issue as well. 

Regards,

Robert Clifton

Hi Robert,

Thanks for your effort and support.

I'm looking forward to receiving your good response.

Best Regards,

Michael

Hi Michael,

I recreated your setup the best I could and found that I was still able to play music. I did find however that if using the settings you chose above and running with a right-justified digital input, the amplifier will not play. Can you verify that you are not running right justified? 

Thanks, 

Robert Clifton

Hi Robert,

First of all, I'd like to let you know that our I2S timing format is no problem.

I don't know well that there is no issue on your side but our side has an issue.

We already tested right-justified, left-justified.

Also audio output is nothing.

If input format is not matched with I2S or right-justified or left-justified, we cannot see PWM signal of OUT_P / OUT_N.

In other words, I think that the audio amp is well aware of the input signal format.

Isn't there any other point what do I have to check for this issue?

Do you have a guidance about PCB layout except data sheet comment?

Pls give me a helpful comment.

Best Regards,

Michael 

Hi Michael,

That's good to hear. 

If you can share the layout that might be able to help us uncover what's going on here. 

Regards,

Robert Clifton

Hi Robert,

I did an experiment by jumping the I2S signal to the TAS5720 EVB.

But, the symptom is same.

So there is no pcb layout issue.

I think that there is problem register setting or I2S waveform.

But, when I checked these 2 items, I couldn't find problem any more.

I don't know what should I do.

Can you give me a guide on how to go back to the beginning and check it out?

Could you please give me your test setup?

1. All register value of TAS5720L when you tested with our setup.

2. Your I2S waveform.

3. OUT_N, OUT_P waveform.

4. Audio signal after LC Filter.

I will double check and retest.

Best Regards,

Michael

Hi Robert,

Our customer's end is Samsung Electronics.

So they have to clear this issue ASAP.

I desperately need your help and support.

Pls refer to the below test result today.

-. Audio source file : customer's wave file (Alarm2.wav)

   

1. Audio output : OK

2. Audio output : Fail

* Question

1. About Audio Data (SDIN) of I2S.

  -. When they send audio data by I2S on the STMicro MCU, The header 44byte portion is discarded and only data of wave file is sent.

     I think that this audio data format is able to be issue.

     In other words, Is there any difference data format (SDIN) of I2S between I2STAS1020 and STMicro MCU?

     How should we send data format through I2S interface? or Could you check data format when change from wave file to I2S signal?

     I'm looking forward to receiving your response by tomorrow.

     If you have any other question, let me know.

  -. As you know, below is wave file format.

Best Regards,

Michael

Hello Michael,

As long time more discussion, i would close this topic. If more feedback, pelease reply on this thread.

Dylan

Hi Luis or Robert,

I'm sorry for the rate reply.

When we compared I2S waveform between TAS1020 EVM and Customer SoC, it was same.

Here is our another test result.

If we connect MAXIM - MAX98357 EVM and I2S output of SoC on customer B'D, audio out is OK.

In other words, I2S output of SoC on customer B'D is no problem.

I'd like to let you know that some of my comments are missing on my post which is Oct 30, 2020 9:24 PM.

About below image, only inserted TAS5720L & Peripheral components on customer's bare PCB for test.

If we use customer B'D which all components are mounted on their PCB, audio out is fail.

Here is my another question.

It seems like that there is another root cause not I2S signal.

 -. If there is some noise or other issue on customer B'D except I2S signal, is it possible that TAS5720L audio out is fail?

Could you please check it and let me know your opinion?

We have to solve this issue.

If we can't fix this issue, our customer will change from TI to MAXIM.

Best Regards,

Michael

Hi Luis or Robert,

I attached image and copy all my comment again.

I'm sorry for the rate reply.

When we compared I2S waveform between TAS1020 EVM and Customer SoC, it was same.

Here is our another test result.

If we connect MAXIM - MAX98357 EVM and I2S output of SoC on customer B'D, audio out is OK.

In other words, I2S output of SoC on customer B'D is no problem.

I'd like to let you know that some of my comments are missing on my post which is Oct 30, 2020 9:24 PM.

About below image, only inserted TAS5720L & Peripheral components on customer's bare PCB for test.

If we use customer B'D which all components are mounted on their PCB, audio out is fail.

Here is my another question.

It seems like that there is another root cause not I2S signal.

 -. If there is some noise or other issue on customer B'D except I2S signal, is it possible that TAS5720L audio out is fail?

Could you please check it and let me know your opinion?

We have to solve this issue.

If we can't fix this issue, our customer will change from TI to MAXIM.

Best Regards,

Michael

Hello Michael,

our supporting team has limited bandwidth today, they will follow up with you tomorrow. sorry for the waitting.

Dylan

Hi Michael,

It is long history. But from your configuration below, i can see setting are correct expect for audio status. But you claimed when I2S (right justified) is available it will be normal.

[0x00]Audio Device ID            = 0x01
[0x01]Audio Power Control    = 0xFD
[0x02]Audio Digital Control1   = 0x04
[0x03]Audio Digital Control2   = 0x80
[0x04]Audio Volume Control   = 0xAF # -16dB
[0x06]Audio Analog Control   = 0x55  
[0x08]Audio Status                 = 0x08
[0x10]Audio Digital Clipper2   = 0xFF
[0x11]Audio Digital Clipper1   = 0xFC

On my side, may you do experiments below?

1. Please just use your PC and TAS5720L EVM board, can you make sound out?
   1. If yes, read out all the register above for my check. (play alarm2.wav file)
   2. If not, please refer to the EVM user's guide to make it work normally.
2. If you can work the EVM board, then re-do the experiments above (PC-TAS1020 I2S --> (disconnect TAS5720 in EVM by modify the resistor to determine the i2c address ) customer's board) --> Isolate I2S from SoC in customer's board --> make sure shared GND --> redo the initialization
   1. to check all i2s is exactly the same as former EVM, then read out all registers for my check
   2. To check whether you can work the customer's board.
3. If you do have the audio precision in your lab, please use the AP as audio source, it would be better.

BR,

Alix Wan.

Hi Dylan,

I couldn't receive your reply.

This issue is very urgent to our customer and me.

Could you please send your review result to me ASAP?

Best Regards,

Michael 

Michael,

Alix is the person to help you. He feedback you above post.

Dylan

Hi Dylan,

A lot of time has passed. When can I get response?

Why is there no reply.

Pls send your review opinion.

Best Regards,

Michael

Hi Michael,

I am responsible for this part. Would you update debug process with questions below?

It is long history. But from your configuration below, i can see setting are correct expect for audio status. But you claimed when I2S (right justified) is available it will be normal.

[0x00]Audio Device ID            = 0x01
[0x01]Audio Power Control    = 0xFD
[0x02]Audio Digital Control1   = 0x04
[0x03]Audio Digital Control2   = 0x80
[0x04]Audio Volume Control   = 0xAF # -16dB
[0x06]Audio Analog Control   = 0x55  
[0x08]Audio Status                 = 0x08
[0x10]Audio Digital Clipper2   = 0xFF
[0x11]Audio Digital Clipper1   = 0xFC

On my side, may you do experiments below?

1. Please just use your PC and TAS5720L EVM board, can you make sound out?
   1. If yes, read out all the register above for my check. (play alarm2.wav file)
   2. If not, please refer to the EVM user's guide to make it work normally.
2. If you can work the EVM board, then re-do the experiments above (PC-TAS1020 I2S --> (disconnect TAS5720 in EVM by modify the resistor to determine the i2c address ) customer's board) --> Isolate I2S from SoC in customer's board --> make sure shared GND --> redo the initialization
   1. to check all i2s is exactly the same as former EVM, then read out all registers for my check
   2. To check whether you can work the customer's board.
3. If you do have the audio precision in your lab, please use the AP as audio source, it would be better.

BR,

Alix Wan.

Hi Alix,

Your response is posted on Nov. 10th from you.

I didn't say when I2S(Right Justified) is normal operating.

I said that both Right Justified and Left Justified are no audio output.

And register value is no problem.

Could you please read my post from start to end again?

And then If you have any request to me, let me know.

If not, let me know your another reply.

Best Regards,

Michael

Hi Alix,

Isn't there any response?

I really need your help.

Pls give me your feedback.

Best Regards,

Michael

Hi Michael,

For your experiment: PC--> USB --> TAS5720 EVB (TAS1020) I2S --> customer B'D, you claim that:

1. OK: only inserted TAS5720L & Peripheral components on customer's bare PCB for test.

2. Fail: If you use customer B'D which all components are mounted on their PCB, audio out is fail
3. You claim I2S from TAS1020 is exactly the same as SoC's

Then the results do not make sense on my side. It seems that other components mounted on PCB in customer's B'D 'affects' the results. Then:

1. With OK, bare PCB for test, please capture one complete MCLK/LRCLK/BCLK/SDIN by playing a SIN wave;
2. With fail, mounted PCB for test, please do captures waveforms above for comparison.

Further, may you confirm your tests on TAS5720 EVM board?

1. Fail: STMicro MCU -->TAS5720 EVM board 
2. OK: TAS1020 --> TAS5720 

For all the tests above, we are confident of i2s from TAS1020, but chances are that board may be damaged. Hence, to sync up our test. It is strongly suggested to use Audio Precision (AP555, AP2700) for the audio source as we believe that should be correct. May you have a try?

BR,

Alix Wan.