• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 50 subscribers
  • Views 359 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.

LMH7322: LMH7322 and LM139A output transients caused by inputs being present before power up

Kannan Sockalingam
Intellectual 1490 points
Part Number: LMH7322
Other Parts Discussed in Thread: LM139A, , LM139, TLV7031, TLV7011

Guys, I am working with a customer and wondering if we have any information on the magnitude of possible transients that could be seen on the outputs, when device is powered OFF and there is voltages on the inputs (coming from an external source - out of our control). SAME question for LM139A as well. Basically looking to assess what could happen if the box is powered down to change test configurations without removing the input signals. 

For the LMH7322:

VCCI is +3.3V

VCCO is -0.8V

VEE is -3.3V

Input signal levels are 1.0V and 1.4V.

 

For the LM139A:

Supply Voltages are ±3.3V

Inverting input is set at half the + rail

Noninverting input is from another ground referenced +3.3V supply.

Output is pulled up to ground through 4.7k.

 

  • 0
    TI__Mastermind 49896 points
    Hello Kannan,

    I am not aware of any specific data (other than what is in the datasheet) on what the LMH7322 does during power-down. It does not have any POR-type circuitry.

    Is it possible to assert the LE pins beforehand? That may put the outputs in a known state. But what happens at power-off is undetermined. The matrix of test combinations explodes with the various supply sequencing and ramping and output loading. They would have to test it under their actual conditions. We cannot duplicate all possible conditions on the bench.

    The LM139A also does not have any POR type circuits, so the behavior is also indeterminate (though it is a much simpler device and possibly more predictable).

    Note the they are violating the input range of the LM139A. The specified input range is GND to 2V below V+ (over temp, which I assume is important since the LM139 is selected).

    Which translated to a split supply, would be an input range of -3.3V to +1.3V. The inverting input is 3.3/2= 1.65V. While this may work fine around room temp, at cold temps they may see some failures.

    I'm not clear on what voltage "Noninverting input is from another ground referenced +3.3V supply" is, but if it is ever greater than +1.3V, then that is also not-good..

    When the power drops, it also changes the input common mode (as the part sees it) in relation to the input voltage, so the inputs are not really "stable" during power down, and the part reacts as it does. How much "glitch" you see would most likely be proportional to the supply ramp-down rate and response time of the comparator (which is also changing). So it is not easy to determine what will happen. Every application is different.

    We do have some devices with POR that ensures a known output state, but that is mostly for power-up. It does not help much if the supply drops faster than the POR can react. One such open collector comparators with POR is the TLV7011, but it is 5.5V max, or the TLV7031, but it is slow (3us).

    They will just have to try it...And as any system that rely on unspecified operations of a device (luck), they may encounter issues in the future (change in process or vendors).
  • 0 in reply to Paul Grohe
    Prodigy 20 points

    Concerning the LM139 input range, the datasheet SLCS006U-Revised November 2018, section 7.6 note 3 says that the device will function properly if one input exceeds the common mode input range as long as the other input is within the common mode range.  

    Have I misunderstood the concern expressed in the above post, or the above paraphrased note?

    Please let me know,

    Thanks.

  • 0 in reply to Bill Moran
    TI__Mastermind 49896 points
    Hello Bill,

    Yes, you understood the note correctly. The simple input stage (Fig. 6) will "tilt" in the direction of the PNP input device that is still functioning (still within input range and pulling tail current), so the output is expected to be correct. But specs like offset, recovery time, response time & supply current will all be affected during that condition.

    The voltage on the non-inverting ("other") pin was not specified in the post. If both inputs are outside the input voltage range, the output could be faulty (despite what the Apps section says, I personally do not trust the "both inputs out" condition and do not encourage that condition - particularly over temperature).

    It is our job to point out when a specification is being violated and bring it to the customers attention, so that they may make an informed decision.
  • 0 in reply to Paul Grohe
    TI__Mastermind 23395 points
    Bill
    Since there has been no further question for a while, I am closing this post. If you have further questions, feel free to reply and the post will be re-opened.
    Chuck