• State
  • Locked Locked
  • Replies 5 replies
  • Subscribers 63 subscribers
  • Views 608 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

LP2996 exposed pad GDN connection required

Bertram Brummer
Intellectual 1906 points

Hi ,

my customer is using the LP2996 for DDR Vtt supply and experienced some problems in his application with the exposed pad.  They saw in a so called environental test that the Vtt was 200-300mV too low and  this happend only on boards where the exposed pad was by accident not soldered to GND.

When putting some pressure from top the Vtt voltage was ok.  Now the question is , whether this is  due to the missing GND connection on the thermal pad?  (GND PIN is connected.   If they meassure thermal pad to GND it looks there is not connection.

Can you please get back if the thermal pad requires  electrically a connection to GND and is it possible that the  lower Vtt voltage is due to this missing connection. If yes , please explain why.

Thanks, Bertram

  • 0
    TI__Mastermind 22610 points

    There is most likely no direct connection from the die to the exposed pad. (aka down-bond), but I'll need to get M&B diagram to be sure.
    This is more likely to be a thermal issue where device cools when the exposed pad is making better contact with the copper area under the exposed pad.

  • 0
    TI__Mastermind 22610 points
    Confirmed by the Product Line ... there is no down-bond from die to thermal pad.
    All ground current is through device ground pin. Thermal Pad can be electrically isolated, but should be connected to copper area under the pad to improve thermal performance.
  • 0 in reply to Donald Jones
    TI__Intellectual 1906 points

    Hi Donald,

    thanks for the quick response.  I guess the thermal pad has a connection to the substrat but if there is no GND plug in the design than it is clear and expected.  This although does not explain what the customer is experiencing in their application.

    If they have no solder connection than they see sometimes a  200-300mV lower Vtt than expected. But if they push the device down  so that it get's connected this delta is gone.

    I asked whether they  see any impact when using coling spray and there is no change.  This means it is no thermal issue. Any idea why my customer is seeing that.  What additonal test's could be done to find the root cause ?

    In the internet I found some articles about substrate noise and that this can get prevented by usig GND plugs. Could this be something which could cause the Vtt shift, like some shift happening when  the substrate floats?

    Thanks in advance

  • 0 in reply to Bertram Brummer
    TI__Mastermind 22610 points

    "... If they have no solder connection than they see sometimes a  200-300mV lower Vtt ..."

    Sometimes? Not always?

    I'm fairly sure that a 200-300mV shift in Vtt is not a noise issue, and is not related to 'GND plugs'.

    If applying pressure to the device causes a voltage shift then this is physical issue ... package stress,  poor solder, cracked copper trace, etc, etc.

    I have alerted the product line to this thread as it appears to require some technical expertise from them that would be specific to this device.

  • 0 in reply to Donald Jones
    TI__Intellectual 1906 points

    Hi Donald,

    thanks for your support so far. I agree to involve the PL on that, because the customer is also telling me that so amount of boards they have seen that is not very big like 2-3 but also that after pressign down the device several times it looks that  Vout is ok which sounds to me more like a solder issue.

    Whom should I contact in the PL on that?

    If you wish we could have this discussion outside e2e and close this thread.  You can easily find me in the TI internal directory

    best regards, Bertram Brummer