ad7795 burning up

I've had several failures in a product line using the ad7795.  The ad7795 is being used to measure the resistance of RTD temperature sensors.  In a few of our boards the ad7795 burns up.  My technician is reporting burn marks on the top of the part and he is having to replace the ad7795.  Power to the part is 5V.  Both AVDD and DVDD are sourced from the same power supply (we're using ferrite beads to isolate AVDD and DVDD) and we're using 10uf and .1uf bypass caps on AVDD and DVDD.  I haven't found any information that indicates that this might be a problem or how to fix it.  I'm wondering if the part is susceptible to CMOS lock up.  Any suggestions?

Carl

  • hi cdb

    i have also been facing same issue with AD7795BRUZ , we are also getting ADC burn marks on chip on few board from site.

    Also we are using internal current sourcing for measuring PT100 sensor.

    Have you got any feedback or any reply from Analog Devices?

  • 0
    •  Analog Employees 
    on Jan 10, 2016 6:16 PM

    Hi,

          There are many possible causes of burning in the circuits and of the common are glitches which is a sudden spike of potential or current spike in a short period of time that exceeds the absolute max rating of the product. In your application, do you have switching that took place, like input switching? Are the circuits properly grounded both AGND and DGND? The two should be connected.

         If it is okay for you to share your schematics for us to look so we can have better understanding or give us a clue of the possible cause.

    Regards,

    Jonathan

  • Hi

    Thanks for your reply.

    My whole system is working on 12VDC and further i have used 5VDC regulator from Mornsun.

    I am using ADG1608 8:1 Mux to switch 16 Analog input , Also i have seperated digital and analog ground by Ferrite bead Wurth Elecktronik. Part no 742792012.

    Can you share mi your email id so that i can send my Schematic for your reference.

    But i have other design in which mux have not used even there too we have found same damage to ADC.

    Just common thing was that i have used internal dual 1mAmp current source for RTD measurement and in firmware i am switching on and off internal current source.  Does it make difference?

  • 0
    •  Analog Employees 
    on Jan 11, 2016 7:29 PM

    Hi Carl,

            For the ground, I was referring to the GND of the overall system. The AD7795 has AGND only, but digital circuits have on the system have different grounding. You may check the EVAL-AD7795 as a reference http://www.analog.com/media/en/technical-documentation/evaluation-documentation/EVAL-AD7795.pdf .

    Carl/ Sanjay,

           When does the Burn up encountered? Was it after power up? Does it occur specific time range? or Does it occur in an event like when starting conversion, connecting an input? Or Does it occur with no specific event or time?

    Regards,

    Jonathan

    jonathan.colao@analog.com

  • Jonathan,

    Your reference to multiple grounds is an odd statement since the AD7795 only has one ground pin (20).  I can only assume you're speaking in general terms for most ICs.

    We're not doing any switching on the analog inputs.  They're directly connected to PT100 sensors.  We're also using the internal 1ma current source.

    Harry,

    We'll test turn on and turn off.  Our design +5V on a power supply for digital and analog power.  The digital and analog power is isolated with ferrite beads.  I suppose we might be getting glitches, but I'm not certain what it could be from.

    Do you think putting 1N6373 TVSs on the power and NUP4201 voltage suppressors on the analog inputs would help?

    Carl