TS3A24157: Cross noise

Guru 11785 points

Shu-Cheng LIN

Replies: 3

Views: 1031

Part Number: TS3A24157

Hi Sirs,

We are use 24157 to switching 2 pcs WWAN on voice call function.

We found 24157 couldn't clear to isolation the voice.
Our block:

audio block.pdf

MAX9867 (I2S audio codec) and NAU88L24 (I2S audio codec) are connect 1 pcs I2S interface WWAN module, TS3A24157RSER is used switching ALC888S (HDA audio codec) MIC_IN /LINE_OUT that connect MAX9867 MIC_IN /LINE_OUT, or connect NAU88L24 MIC_IN /LINE_OUT

Now issue is 24157 setting low, the ALC888S MIC_IN/LINE_OUT is connect to MAX9867 MIC_IN/LINE_OUT, but if we send voice from NAU88L24, ALC888S also could hear NAU88L24 voice.
If 24157 setting high, the ALC888S MIC_IN/LINE_OUT is connect to NAU88L24 的MIC_IN/LINE_OUT, if we send voice from MAX9867, ALC888S will not hear MAX9867 voice.
That cross voice phenomenon is if we call to A by the phone, at this time if B call in, the voice will effect A
Could you help us check why and how to fix it?

Our Schematic:

SCHEMATIC1 _ P22 Audio (ALC888S.output-NEW.pdf

SCHEMATIC1 _ P23 MiniPCIE WLAN&.output-New.pdf

• TS3A24157

Guru 11785 points

Shu-Cheng LIN

• Nov 2, 2017 7:46 AM
• Locked
• Cancel

3 Replies

• Intellectual 2900 points

  

  Fan.Wang

  

  • Nov 2, 2017 6:56 PM

  Hi Shu-Cheng,
  
  Thanks for detailed explanation. While I am analyzing the two schematics, would you also capture the waveforms of NC1, NO1, NC2, NO2 and COM1, COM2 for my reference?
  
  Just one thought for your information. All signals are AC coupled to TS3A24157, so I am assuming TS3A24157 is taking negative signals, which can cause the voice coupled from one channel to the other.
  
  TS3A24157 does not have negative signaling capability. By using AC couple circuitry, you may have to use TS5A22362 or TS5A22364.

  Fan Wang

  • Locked
  • Cancel

  Do you have another question?

  We are glad that we were able to resolve this issue, and will now proceed to close this thread.

  If you have further questions related to this thread, you may click "Ask a related question" below. The newly created question will be automatically linked to this question.

  Ask a related question Ask a new question
• Guru 11785 points

  

  Shu-Cheng LIN

  

  • Nov 3, 2017 3:39 AM

  In reply to Fan.Wang:

  Hi Sirs,

  Actually we have try to pull high 1.6v on NO2, the noise is smaller than before but still could heard. The waveform as below that already pull high 1.6v on NO2

  We are open NC1 and NO1, Only to catch NC2/NO2 with COM2 waveform.

  CH1: NO2

  CH2: NC2

  CH3: COM2

  scope_47.png is base on NC2 connected, then call the phone from NO2

  

  scope_48.png is base on NO2 connected, then call the phone from NO2

  

  scope_49.png is base on NC2 connected, then call the phone from NC2

  

  scope_50.png is base on NO2 connected, then call the phone from NC2

  

  We surprise on scope_47.png. We could catch NO2 waveform and hear ALC888S (pass through COM2) phone ring, but we couldn't catch COM2 phone ring waveform.
  I'm not sure it's case from our Oscilloscope or not. Because we can hear phone ring must under the voice is on max state.

   

  • Locked
  • Cancel

  Do you have another question?

  We are glad that we were able to resolve this issue, and will now proceed to close this thread.

  If you have further questions related to this thread, you may click "Ask a related question" below. The newly created question will be automatically linked to this question.

  Ask a related question Ask a new question
• Intellectual 2900 points

  

  Fan.Wang

  

  • Nov 4, 2017 3:34 AM

  In reply to Shu-Cheng LIN:

  Hi Shu-Cheng,
  
  Thanks for the scope captures. Here is my summary.
  
  Circuit configuration:
  MAX9867 NC1/NC2
  NAU88L24 NO1/NO2
  ALC888S COM1/COM2
  
  Experiment:
  1. NC2 == COM2, NO2 has a call in (Scope 47)
  NO2 pass through to ALC888S and it rings ALC888S, which is not supposed to. However, the waveform does not prove it.
  
  2. NO2 == COM2, NO2 has a call in (Scope 48)
  NO2 correctly passes to COM2, and NC2 is intacted. Also the system level also proved this in your first post.
  
  3. NC2 == COM2, NC2 has a call in (Scope 49)
  NC2 correctly passes to COM2, and NO2 is intacted.
  
  4. NO2 == COM2, NC2 has a call in (Scope 50)
  NO2 and COM2 are correctly connected, and NC2 call in does not affect COM2
  
  
  Analysis
  1. Experiment 1 (Scope 47) need to re-visit because COM2 seems quiet while ALC888S rings
  
  2. If Scope 47 is confirmed correct, there must be another path connecting NAU88L24's MINI_SPK to ALC888S' MINI_SPK
  
  3. If Scope 47 is confirmed wrong due to oscilloscope, please experiment biasing both NC2, NO2 to 1.65V. The ringing waveform is +/-0.8V, and without bias, the NC2 and NO2 can go -0.8V, which may cause channel coupling.
  
  4. If you prove the biasing matters, please also bias COM1 since it is the microphone input. If the PCB needs re-work, please consider TS5A22362.

  Fan Wang

  • Locked
  • Cancel

  Do you have another question?

  We are glad that we were able to resolve this issue, and will now proceed to close this thread.

  If you have further questions related to this thread, you may click "Ask a related question" below. The newly created question will be automatically linked to this question.

  Ask a related question Ask a new question