TS3A27518E: TS3A27518E: NC to COM has some issue

Prodigy 120 points

Replies: 7

Views: 149

Part Number: TS3A27518E

Hi All,

We have some issues that need your help.

We found that "NC" to "COM" has a high fail rate. "NO" to "COM" always test pass.

We also had changed input as "NC" connect to I2S5 & "NO" connect to I2S3. It still fail on "NC" to "COM"

 

The structure is a loop test for I2S as below.

7 Replies

  • Hello,

    Thank you for reaching out and providing the block diagram and schematic.

    • Could you describe what behavior is indicating a "fail"?
    • Does this behavior occur on every NC/COM channel path (1 - 6)?
    • What voltages do you read on NC and COM?
    • When NC and COM signal path is "ON", what is the measured voltage on the control pins IN1, IN2?

    Thank you and best regards,
    Kate

  • In reply to Kate Dickson:

    Hi Kate,

    1.Please refer to 2 txt files. It's a test pass & fails log.

    We will output a signal (I2S3 & I2S5) and the CPU input the signal(I2S6). After the CPU receive the signal, the CPU will compare whether the signal is the same or not.

    Pass log:

    I2S5 to I2S6 pass.txt

    Fail log:

    I2S3 to I2S6 fail.txt


    2.We try to change channel 1&2 to channel 5&6.

    NC still has a high fail rate & NO always passes.

    3 & 4 .We have checked the voltage level on each channel & IN1/IN2 when it tests. Please refer attachment

    There is no signal when the test fails.

    The file "voltage level I2S3_DI-I2S6_DO" is a pass signal(low pass rate).

    COM1 →COM5-I2S3_DO-I2S6_DI

    voltage level CLK

    voltage level FS

    voltage level I2S3_DI-I2S6_DO

    voltage level I2S3_DO-I2S6_DI

  • In reply to Max CHANNG:

    Hello,

    Thank you very much for providing this information.

    If you are seeing no signal at NC, this sounds like the this signal path is still high-impedance or not effectively turning ON. Could you point out the exact levels of IN1, IN2, and /EN when you are attempting to use NC?

    All of these scope shots are from NO path, is that correct? 

    Thanks,
    Kate

  • In reply to Kate Dickson:

    Hi Kate,

    Sorry, it is unclear to describe the symptom.

    Both NC & NO channels can see the signal.
    But, the signal connects to NC-COM, will have a high failure rate.
    We check the channel 1~6 and IN1/IN2 level as you mention, it still has the symptom on NC.
    The waveform is my last reply.

  • In reply to Max CHANNG:

    Hello,

    To clarify further, you are able to measure a signal at NC when it is connected to COM. But your test program (which compares the signal at COM to the signal at NC) is giving "fail" result. Is this correct?

    The test logs are a little bit difficult for me to understand, could you show me what in the program log indicates a "fail"? Why are you using this particular test method? What are the parameters/criteria of this test program?

    In your waveforms, it looks to me like the output is following the input, as expected. Keep in mind that the switch is not exactly "ideal", there is some on-resistance and internal parasitics (specified in the datasheet). So, if your test program has very low margin, this may be why it gives "fail" outcome.

    Best regards,
    Kate

  • In reply to Kate Dickson:

    Hi Kate,

    To clarify further, you are able to measure a signal at NC when it is connected to COM. But your test program (which compares the signal at COM to the signal at NC) is giving "fail" result. Is this correct? Ans: Correct

    The test logs are a little bit difficult for me to understand, could you show me what in the program log indicates a "fail"? Why are you using this particular test method? What are the parameters/criteria of this test program?

    Ans: We generate 2KB random test data, send from I2S-3 and receive from I2S-6, but the received data is inconsistent with the sent, so the test fails.

  • In reply to Max CHANNG:

    Hello,

    Thanks for the confirmation. 

    Could you explain more about what you mean by "inconsistent" received data? Do you know if this is inconsistent timing of the data or the voltage level of the data?

    Have you tried using a new device of TS3A27518E? Does the "fail rate" exist with multiple different devices?

    The waveforms you have provided look okay-- the output COM looks like it is sending the input NC unchanged as expected. Waveforms of the input and output of the switch are typically the best way to troubleshoot signal path concerns with the switch.

    Because I do not know how your test program works, I do not know what could be causing a "fail" in your program. You may want to try double-check & debug your testing program.

    Best regards,
    Kate