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.

ADS1278: Input channel frequently damaged.

Part Number: ADS1278
Other Parts Discussed in Thread: THS4541, , OPA209

Hi experts,

My customer reported that frequently their ADS1278 input channels damaged. Customer design is: vibration sensor -> OPA209 (+/-12V power) -> THS4541 (+5V power) -> ADS1278. All grounding, AGND and DGND are connected to a uniform GND plane. AVDD=5.0V, DVDD=1.8V, IOVDD=3.3V, no sequence control for power up/down. THS4541 is powered from same AVDD with ADS1278's AVDD. Strangely the front end op amp and THS4541 are always working properly, only input channels of ADS1278 damaged. Could you please help me what is the possible root cause that result damage to ADC inputs?

Thank you!

John

  • Hi John,

    How are they confirming that the input channel has been damaged? Are they able to measure a larger than normal input current with no valid output data for that channel? 

    I think most likely this problem may be caused by the startup conditions. If the +/-12V power supplies come up first and the 5V rail is briefly unpowered, the output of the THS4541 may be pulled up to whatever voltage the OPA209 is at. I would think that the resistors around the FDA would limit the current sufficiently not to cause damage to the ADC but if this transient spike occurs regularly with repeated power cycles it may be enough to damage the device over time. It would be best practice to include some protection circuitry at the input of the THS4541 since it is being driven from a source with a higher supply. Even if the THS4541 itself is not damaged, enough energy may be passed to the ADC to damage it.

    You might also just double check how the ground is poured around the ADS1278 DGND and AGND. If these ground pins get separated by more than 300mV or so it can damage the device, so best to verify one of them isn't coupling a switching transient or something of that nature.

    Best,

    Zak