• State Resolved
  • Locked Locked
  • Replies 2 replies
  • Subscribers 47 subscribers
  • Views 677 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

LMV344 unused and improperly terminated op-amps..

Amanda Ross95
Intellectual 2740 points
Other Parts Discussed in Thread: LMV344

Hello there,

We're currently working on a design and have received boards where the third and fourth op-amp inputs have been left floating and are unused. After reading this article by Todd Toporski, it is seeming like we may have to scrap the first iteration of boards. But then upon reading the summary here on the forums by Bruce Trump, we are wondering how bad the worst case scenario will be.

It is given that we will have increased quiescent current and with that probably a higher thermal output but we would like to know by how much? Would you be able to give a worst case estimate of how much the Iq could increase versus if we went with a unity gain follower termination?

Bruce's article says that internal noise interference from same-package op amps shouldn't be too much of a worry, is that true for the LMV344?

Any comments are appreciated. Thanks so much.

-Amanda

  • 0
    TI__Guru 52356 points

    Hello Amanda,

    The warning about increased supply current usually applies to bipolar devices. Since a bipolar transistor has real base currents, when "railed" the internal sources dump large currents into the bases of the transistors, saturating them and causing the larger supply currents. The current to drive the bases comes from the supply..so, obviously,  the supply current rises. Output stages are usually the biggest cause of the excess currents since they are the largest devices with the lowest betas. This is most noticeable on bipolar high power and/or high-speed devices.

    When the inputs are left unconnected, the outputs tend to slam against one of the rails ("comparator" action) - which is where the large currents occur.

    CMOS devices do not have large gate currents since they are voltage driven (but there are always exceptions). So you do not see as large of supply current rise with a CMOS device versus a bipolar device.

    The LMV344 is a CMOS device with a MOSFET output stage, so the supply current will not rise too much - maybe a few percent - but not enough to be dangerous.

    The danger of "chatter" is real. This is where the high impedance inputs "float" to random voltages due to the interplay of bias currents and leakages, and if the two inputs come close in potential (around the offset voltage), it can cause the input to start moving - usually slamming from one rail to the other, or slowly floating somewhere in between them.  In sensitive circuits (high impedance), the "slamming" output can inject noise into adjacent channels. In most circuits with <100k resistors, you probably would not even notice it.

    If the inputs float outside the common mode range, this simply cuts off the input stage and the channel flops and the output goes to some undetermined state. This should not affect adjacent channels as current designs take amp-to-amp isolation as one of the priorities.

    In short, the LMV344 will not be "damaged" by allowing the inputs to float - but it is not recommended due to the potential noise issues above.

    So in the future revision of your boards, the amps should be biased. But I don't think it will harm anything in the current boards.

    Regards,

  • 0 in reply to Paul Grohe
    TI__Intellectual 2740 points
    Thank you Paul for your detailed answer. I really appreciate your taking the time to explain things in the terms you did.

    Best Regards,
    -Amanda