• State Resolved
  • Locked Locked
  • Replies 19 replies
  • Answers 3 answers
  • Subscribers 46 subscribers
  • Views 382 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

ADS8568: ADS8568 output data is shifted to about 60 % of correct data.

Kazuya Nakai59
Guru 20540 points
Part Number: ADS8568

Hello guys,

One of my customers uses ADS8568 for their current mass-production.

Recently, the following phenomenon was observed on a specific ADS8568.
They have the questions as the follows.
Could you please give me your reply?

[Phenomenon]
At a specific ADS8568, when the board is turned off and then turned on again, the AD converter value becomes abnormal.
The AD converter itself is working, but the data conversion value is approximately 60% of the expected value.
It looks like the ADC internal reference setting (internal DAC output?) is abnormal, or the ADC internal conversion gain is abnormal.

This phenomenon is reproducible. If the power to the board turned off and turn it on again with AVDD (5V) remaining at 0.5V or more, the same phenomenon will occur with this same device.
If I wait until AVDD is below 0.5V, it works fine.

[ADS8568 setting]
1. /HW/SW terminal = L
2. /PAR/SER terminal = H
3. RANGE terminal = H
4. REFEN terminal = H
VREF is internal reference (2.5V)

[Questions]
1. Is there a possibility that this phenomenon will occur? Or is this device broken?
2. They initialized the device with RESET terminal H/L after this phenomenon was happened. But the abnormal phenomenon was not disappeared.
    Which part of the function is reset by this reset pin ->L->H->L?
3. Does ADS8568 have any gain control block? If the answer is yes, can the gain control block be reset by RESET terminal L/H/L?

Your reply would be much appreciated.

Best regards,
Kazuya

  • 0
    Guru 20540 points

    Hello guys,

    Could you please give me any reply?

    Thank you and best regards,
    Kazuya. 

  • 0 in reply to Kazuya Nakai59
    TI__Guru 59876 points

    Hello Kazuya-san,

    It sounds like the ADS8568 is getting stuck in a type of "brown-out" condition. In this case, it appears to me that RESETn is having no effect on the analog circuitry from which the internal reference voltage is generated. Can you confirm this by measuring the REFIO pin after the brown-out condition? What is the result before and after RESETn is issued? 

    Regards,

    Ryan

  • 0 in reply to Ryan Andrews
    Guru 20540 points

    Hello Ryan,

    Thank you very much for your reply.

    I will check it with the customer and I will inform you as soon as I get it.

    Thank you again and best regards,
    Kazuya.

  • 0 in reply to Kazuya Nakai59
    Guru 20540 points

    Hello Ryan,

    Thank you for your supports and I'm sorry to be late my response.

    I got the following information from the customer.

    I asked them to measure output voltage of REFIO, REFAP, REFBP,REFCP and REFDP pins.
    As the result, in case of normal operation, all of these pins were 2.5V.
    But in case of the phenomenon occurring, all of these pins were 4.2V~5V.

    As the other information,

    1. Total number of the device has the phenomenon is 3 pcs at this moment.
    　(Denominator is unknown.)
    2. They didn't access Internal CONFIG register and other resisters.
        So all resisters are default value.

    I think VREF voltage is shift to upper side of target voltage(2.5V).
    Is it possible that this VREF shift is happened when ADS8568 is powered up with AVDD (5V) remaining at 0.5V or more?

    Do you know why this phenomenon happened?

    Thank you and best regards,
    Kazuya.  

        

      

  • 0 in reply to Kazuya Nakai59
    Guru 20540 points

    Hello Ryan,

    Thank you very much for your supports.

    Do you have any update for this phenomenon?

    Thank you again and best regards,
    Kazuya.

  • 0 in reply to Kazuya Nakai59
    TI__Guru 59876 points

    Hello Kazuya-san,

    REFIO and all REFyP pins = 2.5 V is expected for normal operation.

    From the original description, it sounds like this phenomenon only occurs when AVDD is turned off momentarily, but 5 V is restored before AVDD fully discharges. In that case, it is possible for part of the internal reference circuit to get stuck, but it's unclear exactly what is the root cause (i.e. DAC output, reference buffer, etc.). 

    One way to fully recover the device reliably is always to ensure complete power-down is achieved before enabling analog and digital supply voltages.

    If the customer detects REFIO ~= 4.5 V, another option is to try toggling the internal reference settings.

    Can you ask the customer to toggle the following bits:

    1. REF_EN (bit 15, internal reference enable) -> set this bit to 0b and then 1b (i.e. disable and then enable internal reference). Does this recover the REFIO voltage?
    2. REFBUF (bit 14, internal reference buffers disable) -> set this bit to 1b and then back to 0b (i.e. disable and then enable internal reference buffers). Does this recover the REFyP pins (i.e. buffer outputs)?

    Regards,

    Ryan

  • 0 in reply to Ryan Andrews
    Guru 20540 points

    Hello Ryan,

    Thank you very much for your update.

    They use ADS8568 with the following conditions.
    1. Hardware mode (/HW/SW pin = Low)
    2. Serial mode (/PAR/SER pin = High)
    3. Internal reference mode (REFEN//WR pin = High)
    4. No CONFIG register access (They don't access to the register)

    It is difficult for them to change the register setting at this moment.

    Do you have any other way to find the cause of this phenomenon?
    Also have you ever seen any phenomenon like this customer case?

    Thank you again and best regards,
    Kazuya. 

  • 0 in reply to Kazuya Nakai59
    TI__Guru 59876 points

    Hello Kazuya-san,

    Sorry, I missed the detail about hardware mode in your original post. In that case, when the high REFIO voltage is detected, can the customer bring the REFEN/nWR pin low and return it to high? Does the REFIO voltage recover properly?

    The cause of the phenomenon is the low AVDD voltage which stays approximately 500 mV above ground before returning to 5 V. When the ADC is powered down, all supplies should be fully discharged to ensure reliable operation at power up.

    Regards,

    Ryan

  • 0 in reply to Ryan Andrews
    Guru 20540 points

    Hello Ryan,

    Thank you very much for your strong supports.

    Please let me just make sure
    If the REFIO voltage recover properly when REFEN/nWR pin is set to low and return it to high again, is it possible to say cleary that the cause of the phenomenon is 500mV approximately remaining at AVDD before returning the voltage to 5V?

    Thank you again and best regards,
    Kazuya.

  • 0 in reply to Kazuya Nakai59
    TI__Guru 59876 points

    Hello Kazuya-san,

    I am only suggesting the REFEN/nWR pin test as a method to recover normal device operation when the issue is detected. It does not indicate the root cause. The best method to ensure reliable device operation is to fully power-down both analog and digital supplies. I am quite sure that the root cause is due to improper power-down (i.e. AVDD is not fully discharged).

    Regards,

    Ryan

  • 0 in reply to Ryan Andrews
    Guru 20540 points

    Hello Ryan,

    Thank you very much for your many supports.

    I understood what you want to say.
    I asked them to check whether the REFIO voltage recover properly when REFEN/nWR pin is set to low and return it to high again.
     Also could I ask you a few questions as the follows?

    Q1.
    Have you ever seen or heard this phenomenon on ADS8568 or other device?

    Q2.
    Is there any possibility that this phenomenon occurs with other ADS8568 chip though the AVDD remaining voltage might be different. 

    Thank you again and best regards,
    Kazuya.

  • 0 in reply to Kazuya Nakai59
    TI__Guru 59876 points

    Generally speaking, any ADC can experience problems when power supplies are not ramped or discharged properly. The types of issues that can result from this are wide ranging and will vary from one device to the next. There is no guarantee that the supply voltage threshold will be consistent across multiple ADS8568 devices.

    Regards,

    Ryan

  • 0 in reply to Ryan Andrews
    Guru 20540 points

    Hi Ryan,

    Thank you very much for your strong supports.

    I told your reply and the customer asked us the following questions. Because they have a plan to use ADS8568 over thousands for their products in the future.
    Could you please give me your reply?

    Q1. What is the voltage of full discharging in case of ADS8568?
    Is it OK to discharge less than 0.1V? Or 0.2V?

    Q2. You suggested to do the REFEN/nWR pin test as a method to recover normal device operation when the issue is detected.
    If this method works well with a few ADS8568, can TI say this method be used for all of ADS8568?
    Is there any theoretical reason for this method?
    Or is it an idea as a trial?

    Q3. Most of ADS8568 on our boards do not cause this phenomenon even if the power supply voltage remains on the board. What is the difference?

    Q4. Is there a possibility that this phenomenon is caused by an individual IC failure?

    Q5. Is it okay to recognize that this phenomenon is occurred when the power is turned on timing only
    and is not occur at other times (during continuous power supply)?

    Your reply would be much appreciated.

    Thank you again and best regards,
    Kazuya.

  • 0 in reply to Kazuya Nakai59
    TI__Guru 59876 points

    Hello Kazuya-san,

    What was the result of toggling the REFEN/nWR pin? Did the REFIO voltage recover properly? If the REFIO voltage recovers when AVDD discharges to less than 500 mV during power down, or after toggling REFEN/nWR pin when incorrect REFIO voltage is detected, then I would not assume the device is damaged.

    I would not expect this issue to occur with continuous power supplies during normal operation. If the application requires to power-down and restore power, I would suggest waiting until AVDD < 300 mV to give some margin. Based on your earlier post, the issue does not occur if the customer waits for AVDD < 500 mV, so to me this seems like a reliable approach. Unfortunately, we do not have specific characterization test data available to give a precise limit.

  • 0 in reply to Ryan Andrews
    Guru 20540 points

    Hello Ryan,

    Thank you very much for your prompt reply.

    They have not tried to toggle the REFEN/nWR pin yet.
    The customer has the following additional questions.
    Could you please give me your relay?

    Q1.
    If the REFIO voltage recovers after toggling REFEN/nWR pin, 
    can it be applied to all ICs which the same phenomenon occurs?
    Is this method a proven method?
    Or is it just trial    ?

    Q2.
    Is this REFIO incorrect voltage not occurred except the power up timing?

    Thank you again and best regards,
    Kazuya.

  • 0 in reply to Kazuya Nakai59
    TI__Guru 59876 points

    Hello Kazyua-san,

    Q1: just a trial

    Q2: I believe this is correct. I am not aware of the same phenomenon occurring for different reason.

    Regards,

    Ryan

  • 0 in reply to Ryan Andrews
    Guru 20540 points

    Hello Ryan,

    Thank you very much for your reply.

    Could I ask you one additional question as the below?

    Q.
    Have you ever seen or heard this REFIO incorrect output voltage phenomenon of ADS8568?

    Thank you again and best regards,
    Kazuya.      

  • 0 in reply to Kazuya Nakai59
    TI__Guru 59876 points

    Hello Kazuya-san,

    I have not come across this issue before on ADS8568.

    Regards,

    Ryan

  • 0 in reply to Ryan Andrews
    Guru 20540 points

    Hello Ryan,

    Thank you very much for your reply and I'm sorry to be late my response.

    I asked them to turn on after going down less than 0.3V.
    They will add any delay circuit for re-powering up.

    Thank you again and best regards,
    Kazuya.