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.

ADS1174: One channel not measuring after long term operation

Part Number: ADS1174
Other Parts Discussed in Thread: THS4524

Hello,

I am carrying out current measurements using Rogowski coils, one Operational amplifier THS4524, and one ADC ADS1174. 
When installed, the device is working 24/7 without interruption. 
Few devices are now having a similar issue: after months of operation without issue, I can now see that one of the channels is not working anymore (not always the same channel). It measures 0 when it should not. 
Since the devices are installed outside, I first thought the issue could come from the temperature or humidity, but I ran tests and measurements without getting any results.

Unfortunately, every time one of those devices is dismounted and brought back to us, it is working fine again.

So we were wondering if the issue could maybe appear on one of those components with long term operation and if it could disappear when powered off and on again. 

Have you already had feedback on those components that could relate to our problem?

Thank you in advance,

Best regards,

  • Hi Matthieu,

    Thank you for your post and sorry to hear you are experiencing such issues.

    We have heard from some customers that this has been an issue in the past. A recent similar example can be found in this thread. Namely, the root cause appears to be a sudden transient in the power supplies, especially ground. The restriction between AGND and DGND is +/-300mV maximum. Although no one has specifically observed a violation of this spec at the time of a channel failure, the solution has generally been to improve the connection in the PCB layout and place the short as close to the ADC as possible. Our EVM ties the thermal pad of the ADS1x7x device to AGND, but in theory DGND should work just as well. The idea is to connect the thermal pad to the lowest potential on the board.

    So just to clarify a few details:

    1. Does one device have repeated failures after returning to the field? If so, does the same channel fail again?
    2. When you say "the device is dismounted," are you referring to the ADC or to the system as a whole? I thought for a moment you might be de-soldering and re-soldering the ADC to the PCB.
    3. Does powering the unit on/off resolve the issue? In the past, customers have claimed that a failed channel would remain permanently dead. Is that what you have noticed as well?

    Best Regards,

  • Hello Ryan,

    Thank you for your quick answer.

    Thank you for linking this other thread, it really is similar indeed. This issue with the AGND and DGND could be the cause. It is happening on an old version of the layout (prototype) which could have AGND/DGND issues. We have greatly improved the GND part of our layout in the newer version (mostly for EMC reasons).

    To answer your questions:

    1. None of the defect devices have returned to the field.
    2. I was referring to the system as a whole. Sorry for this misleading word :)
    3. Yes, actually. The devices that we have brought back in our lab for tests are now working as expected: all channels are working.

    Best regards,

  • Hello Matthieu,

    Hopefully the new ground layout is working better for you then!

    If your units continue to be returned and you find that the ADS1174 devices begin to fail permanently, you can contact the distributor that sold them to you and submit a Failure Analysis request with them.

    Best Regards,