ADG1434 ESD part recommendation


I am using the ADG1434 quad switch and we are having what appear to be ESD failures. Using a curve tracer, we can see 20-25uA @ 0.5V between pin 1 and VSS on two different failed parts. We have one more failed part that is closer to a dead short between pin 1 and VSS. Pin 1 was where the ESD event (+/-8kV direct) occurred.

We were using the 0402ESDA-MLP made by Cooper Bussmann, which have a working voltage of 30V and a clamp voltage of 35V. They are performing to their datasheet during our tests. The ADG1434 is biased at +/-16V and they fail more often when they are ON (ie. the switches are biased), as opposed to OFF (VDD = VSS = 0V). I am assuming that I need to find an ESD diode that is bidirectional, Working voltage of 15V (to meet our EIA_232_F spec), with a clamp voltage around 18V-19V so that we don't violate the abs. max Vdd to Vss of 35V. I'm having trouble finding something like this in a reasonable size package (0402 or 0603).

What is the ESD rating on the ADG1434? Any other event types that could cause a latch up that would damage the ADG1434?

Thanks for your time!


  • 0
    •  Analog Employees 
    on Apr 4, 2012 4:48 PM

    Hi Jeff,

    The ADG1434 has a 2kV HBM ESD rating.

    Are you using the TSSOP or LFCSP package, is pin 1 IN1 or Drain?  Where did you apply the 8kV pulse and where have you the ESD protection?  Are you testing sytem to IEC61000-4-2 standard?  Yes, you will need to clamp to 18-19V level so as not to violate the max ratings.

    The main cause of latch-up is if your analog or digital inputs exceed your power supply rails or if you get large glitch or transient spike on your inputs.

    Another part that might be worth considering is the ADG5434.

    This part has 8kV HBM ESD, its latch-up proof and is also on a higher voltage process so using existing diode that clamps to 35V would be fine.

    If using TSSOP package then the ADG1434 and ADG1534 are pin-pin compatible.



  • Thanks for the quick reply.

    We are using the LFCSP package, so pin 1 is D1.

    We applied the pulse at the connector input, which then goes to though 4 inches of 10 mil trace, another connector interface at which the ESD diode is placed 100 mils from. There is another 2 inches of 10 mil trace and then the input to the switch. We were very conscious of ESD diode layout during part placement. I don't know that it can improve much. We may have to deal with the latch-up another way or move to the more protected part.