• State
  • Locked Locked
  • Replies 5 replies
  • Subscribers 64 subscribers
  • Views 894 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.

LM2662: Latch-Up if "LV" is left floating

Wayne Thayer47
Prodigy 30 points
Part Number: LM2662
Other Parts Discussed in Thread: MAX660

We are evaluating the LM2662 as a direct replacement for the CAT661 DC/DC converter. We have identified a source of latch-up that is not on the datasheet. A careful review of all documents available on the TI website has nothing about this issue.

Latch-up will occur on this part if an input power interruption occurs which causes the input voltage to be in the range 0.3V-1.3V. If power is re-applied while the voltage is in this range, the device will consume approximately 300mA of current (input rises to about 1.5V). IF the supply providing the input power to the LM2662 does not have over 500mA of capacity available, then the device will remain in the latched state.

For the circuit studied, we used an LM2931AM-5.0 to provide the input to the LM2662, with 300uF of capacitance in that net. 47uF was used for both the switching CAP and the output CAP. We tested the current required to get it out of the latched state by providing another 5V supply to the 300 uF capacitor bank with a diode, and then added resistance in series with the diode until it would not bring the LM2662 out of latch-up.

This condition is not tolerable for our product, so a circuit board modification will be necessary to hold LV to GND.

If you agree with these findings, please update your technical documents on this part so others don't waste their time or create a device that could have unreliable operation.

Thanks!

  • 0
    TI__Guru* 76065 points
    Thank you for the info. In the datasheet Pin Function table, and in Section 9.2.1.1, it clearly recommends that the LV pin must be grounded for supply voltage <3.5V. It seems a more explicit note on emphasizing such requirement would help. Again, thank you.
  • 0 in reply to Youhao Xi
    Prodigy 30 points
    You misunderstand. We followed the datasheet precisely: The power supply is 5V provided from an LM2931AM-5.0.

    The part was used entirely within the datasheet guidelines. You do not need a more explicit note about LV and low supply voltages. You need to place a warning about "latch-up" under intermittent supply conditions. ANY "latch-up" in a TI product should be taken very seriously, not blown off with a suggestion that the datasheet was not followed. The fact is that this part is dangerous to use if there is any chance of a fast supply cycle or intermittent supply.

    Please re-consider updating the datasheet to emphasize the possibility of "latch-up" if LV is not tied to GND.
  • 0 in reply to Wayne Thayer47
    TI__Guru* 76065 points

    Thank you and you are right.  Happy Thanksgiving!

  • 0 in reply to Youhao Xi
    Prodigy 30 points

    Thank you for your re-consideration.

    Note the MAX660 apparently has the same problem!

    Should we have an expectation that TI will do something about this? (Like immediately changing the documentation and consider repairing the design?)

  • 0 in reply to Wayne Thayer47
    TI__Guru* 76065 points
    We will discuss this internally and see what is the best way to go. Again, thank you.